Re: Add loop script

[Timo Mikkolainen]

Aaand as some peoples' email programs filter out scripts, here's the
script in link form:
http://koti.welho.com/tmikkola/scripts/addLoop.js

2008/7/9 Timo Mikkolainen <[EMAIL PROTECTED]>:
> Cleaned up the code a bit...
>
> 2008/7/9 Timo Mikkolainen <[EMAIL PROTECTED]>:
>> Here's a simple add loop script for SDS objects. Your selection must
>> consist of SDS objects only and you have to select at least one edge
>> for it to work.
>> place it in Realsoft3D/scripting/.
>>
>

Re: Add loop script

[Timo Mikkolainen]

Cleaned up the code a bit...

2008/7/9 Timo Mikkolainen <[EMAIL PROTECTED]>:
> Here's a simple add loop script for SDS objects. Your selection must
> consist of SDS objects only and you have to select at least one edge
> for it to work.
> place it in Realsoft3D/scripting/.
>


addLoop.js
Description: JavaScript source

Add loop script

[Timo Mikkolainen]

Here's a simple add loop script for SDS objects. Your selection must
consist of SDS objects only and you have to select at least one edge
for it to work.
place it in Realsoft3D/scripting/.


addLoop.js
Description: JavaScript source

Re: foresr

[Timo Mikkolainen]

Impressive!
You should run the trees through Zaug's Real RangeZ! to get some variance =)

2008/5/21 Arjo Rozendaal <[EMAIL PROTECTED]>:
> Hi,
>
>
>
> Nice trees Mark!
>
>
>
> Here's my wip. There are 22500 trees in this scene
>
>
>
> www.xs4all.nl/~joly/temp/forest.jpg
>
>
>
> It's nothing much yet, just experimenting.
>
>
>
> I chose a completely different approach to the tree. It has no individual
> leaves but rather a texture of complete twigs with leaves on a polygon.
>
> After creating the tree I exported it as obj to minimize the number of faces
> in Polytrans.
>
> This way I got a tree containing about 3000 triangles. This is much easier
> to handle, doesn't require so much memory.
>
> Of course it's not as nice as real twigs, but the low memory usage has some
> obvious advantages.
>
>
>
> The tree I've been working on is not yet good enough, when I think it's
> worth sharing I'll post it.
>
> I have to find out what shape is best for the leaf object. It should not be
> a simple square polygon but it should mimic the bunch of twigs with the
> smallest amount of polygons possible. That should hide the cardboard effect
> a bit.
>
>
>
> Arjo.
>
>

Re: service pack 1 release appears? :-)

[Timo Mikkolainen]

Looks like the SP1 release is imminent (unless we find major
showstopper bugs from the release candidate).

2008/4/29 Frank Brübach <[EMAIL PROTECTED]>:
> dear rs friends, a short question: the service pack 1 of rs 3d V6 does 
> appears end of this month (april) or the next one? :-) because I have got 
> some strange problems with the rs 5.1 version (SP4) during my last weekly 
> works with realsoft renderings and javascript buildings I would like to work 
> with a stabil version 6:-) with full functionally power:-D
>
>  best regards, good luck for finding all silly bugs, bye, Frank Brübach
>
>  by the way: this forum is still alive ?
>  ___
>  Jetzt neu! Schützen Sie Ihren PC mit McAfee und WEB.DE. 30 Tage
>  kostenlos testen. http://www.pc-sicherheit.web.de/startseite/?mc=00
>
>
>

Re: question about my vsl shader :-) rs-forum.the-final.com !

[Timo Mikkolainen]

2008/4/22 Frank Brübach <[EMAIL PROTECTED]>:
> Hello again... first of all: See the good work!!! 
> "http://rs-forum.the-final.com": GOOD LOOKING FORUM and GREAT WORK TO 
> MATTHIAS and BORIS! :-) to change the realsoft 3ddart.org forum. Hope this 
> site cannot cracked by spam mail and crazy people, they are trying to damage 
> this important realsoft forum. -
>
>  - Next little picture of my little work in progress... glass and 
> watertextures with vsl and perhaps some day "ice" too:-)
>  The rendering times crease to 4:02 min., four different vsl textures, but 
> the aim should be to get it into one vsl texture, if possible...
>
>  Question 1) the rendering times will reduce handling with only "one vsl" 
> texture combining all properties?
I'd imagine so, especially if you can cut down some operations that
the materials share.
>
>  Question 2) the vsl wizard shows only "color", "transparency", "bump map" 
> and so on... Why it isn't possible to get all properties in the "vsl wizard" 
> of the objects? Perhaps a "tree structure" will make sense! ? ;-)
That's what you see in the advanced mode. There's only so much you can
do with the wizards.
>
>  -> So you can find perhaps "optical thickness" and "refraction" and perhaps 
> make new entries for the object properties ?
Time to delve in to VSL then:
Surface Properties
Surface:Transparency = (1,0,0)
Surface: Optical Thickness = 0,3

This'll set transparency to pure red and the optical thickness to a
completely arbitrary value of 0,3.
>
>  bye, frankolino, servus and sunny "springtime" has began here in germany :-D
>
>
>  -Ursprüngliche Nachricht-
>  Von: user-list@light.realsoft3d.com
>  Gesendet: 17.04.08 18:39:56
>  An: user-list@light.realsoft3d.com
>  Betreff: Re: first "ice-water"-picture and open question for an ice shader 
> cooperation:-)
>
>  I have forgotten: the first picture with "ice-water"-effects as attachment, 
> very simple vsl, o.k.:-) I have combined two textures, but the effect ist 
> good! only the rendering times are increasing (bad!);-)  bye, servus, Frank 
> Brübach
>
>
>  -Ursprüngliche Nachricht-
>  Von: user-list@light.realsoft3d.com
>  Gesendet: 17.04.08 18:08:23
>  An: user-list@light.realsoft3d.com
>  Betreff: Re: open question for an ice shader cooperation:-)
>
>  Hi Timo :-) Every help is welcome:-)))  I invite you:-D
>
>  Yes, it will be a strong work to get a good "ice shader" or similar 
> water-ice products.
>
>  - I have some ideas for the roughness, bumpmap, perhaps even fog and 
> noise... the problem is only for me to understand this complex possibilities 
> for a good looking ice shader with e.g. "air bubbles" (your advice is good!) 
> or something with fresnel, scratches and bump map things and so on...
>
>  - I have started with the glass material... (matt glass), after that I will 
> manage a special look for "roughness" and "cold-freezing" roughness... I will 
> have a certain look to the examples from the rs folder (vsl or material)...
>
>  - creating such vsl things and learn with the shader language, for me it's 
> not so easy to handle with the vsl stuff (yes, very amazing!) but the vsl 
> power is damned great, the input/output channels are really wide spread;-) 
> must laugh...
>
>  yes, If you have some ideas, bring it to the forum :-) I am working forword 
> tomorrow with the vsl shader and bring the next results here... best regards, 
> servus, Frank
>
>  "ice-shader" made of this: color, transparency, refraction, fresnel, fog, 
> bumpmap/ noise...
>
>  good evening:-)
>
>  -Ursprüngliche Nachricht-
>  Von: user-list@light.realsoft3d.com
>  Gesendet: 17.04.08 14:10:27
>  An: user-list@light.realsoft3d.com
>  Betreff: Re: open question for an ice shader cooperation:-)
>
>  A good start on all shaders is to think what the qualities of the
>  material are (i.e. what makes ice look like ice). The glass-like parts
>  are easy enough, you can pretty much just use a glass shader as a base
>  (with a tweaked color and optical thickness). The harder parts are the
>  surface features and air bubbles inside the ice. The surface features
>  can either be modeled or done with some creative bumpmapping (there's
>  some bump materials in the startup project that fit the bill). The air
>  bubbles are harder though: you can either model them or just cheat and
>  add a high frequency noise to the bump.
>  I'll see if I can give you some pointers once I get home.
>
>  On 15/04/2008, Frank Brübach <[EMAIL PROTECTED]> wrote:
>  > Dear RS friends, an open question:-)
>  >
>  >  "Perhaps somebody has fun to build with me an "ice shader" with the vsl 
> editor... I have started with a simple glas textur (from the rs manual, vsl 
> chapter), I have attached the example as a rar file above...
>  >
>  >  would be very nice to work with other realsoft guys to learn with the vsl 
> editor and get some interesting new vsl shader for ice, water and so on... "
>  >
>  >  servus, best regards, Frankolino

Re: How can I control 1D particles Size, using an animated texture?

[Timo Mikkolainen]

I don't see how that would make any difference.
The optimal solution would be to make the size channel accessible in VSL.

On 18/04/2008, Beg-inner <[EMAIL PROTECTED]> wrote:
> Hi Timo again..
>
>  Would things work, if the particle specific channels, like 'Size' in my
> case, was implemented so they could show up in the 'Render Channel' popup
> menu that becomes activated when 'Pointwise' is checked.. ?
>
>  I mean, if I in the Geometry Tab of the Particle, set the 'Color' to
> pointwise, and in its 'Render Channel' popup chose 'Size'
>  If this could be done, would a Material mapped texture onto a Particlecloud
> then affect the 'Color' Channel in the Particle and that in turn add its
> value to the pointwise 'Size' ?
>  Or am I just on totally wrong track... ?=( as usual.. rofl..
>  If this had worked.. it would be far easier than using JS or..hmm hehe..
>
>  Maybe a bad example by me, since the texture should only control the size
> of the particles.. but not their color=).. hmm
>
>
>  Thx in advance
>
>  Take Care
>  Best Regards
>  SteFan Gustafsson ( Beg-inner )
>  A Proud Owner and User of Real3D and Realsoft3D..
>
>
>
> >
> > Well, you can't access particle size from VSL (afaik), so the only way
> > I can think of is to assign the material to a custom channel in the
> > particles and then use JS to copy that value to the particle size.
> > Alternatively you could try to use a choreograph instead of js.
> > There's a few probs though: I don't know if either way can access the
> > value and pointwise attributes in JS might still be a bit
> > broken/incomplete.
> >
> > On 28/02/2008, Beg-inner <[EMAIL PROTECTED]> wrote:
> >
> > >
> > >
> > > Hi all..
> > >
> > > Does anyone know how to make particles pointwise Size value to be
> controlled
> > > by an (animated) texture?
> > > AFAIK, the attributes values can be held in a channel and be used in
> > > materials to control/influence channels like Color and so on.. to make a
> > > material that change color depending on the particle size.. but I cant
> see
> > > how to make it work the other way around.. Change size (or other
> Particle
> > > specific channels) by a color ... or `?
> > > (Do one need to use JS to get info back n  forth between a texture and
> the
> > > Pointwise Size attribute ?)
> > >
> > > I am going to use these controlled particles inside a Metaball level, to
> > > create cool animated Metaball shapes.. =)
> > >
> > > Thx in advance..
> > >
> > > Take Care
> > > Best Regards
> > > SteFan Gustafsson ( Beg-inner )
> > > A Proud Owner and User of Real3D and Realsoft3D..
> > >
> >
> >
> >
> > --
> > No virus found in this incoming message.
> > Checked by AVG Free Edition.
> > Version: 7.5.516 / Virus Database: 269.22.13/1375 - Release Date:
> 4/12/2008 11:32 AM
> >
> >
> >
>

Re: How can I control 1D particles Size, using an animated texture?

[Timo Mikkolainen]

TBH I can't remember... we did do some digging with Zaug some months
ago, will have to see if either of us has some code handy (if we ever
managed to get it working...).

On 18/04/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Hi Timo,
>
>  how do you access custom channels via JS :-?
>
>
>  Matthias
>
>
>
>  ----- Original Message -
>  From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>  To: 
>  Sent: Thursday, April 17, 2008 2:17 PM
>  Subject: Re: How can I control 1D particles Size, using an animated texture?
>
>
>  > Well, you can't access particle size from VSL (afaik), so the only way
>  > I can think of is to assign the material to a custom channel in the
>  > particles and then use JS to copy that value to the particle size.
>  > Alternatively you could try to use a choreograph instead of js.
>  > There's a few probs though: I don't know if either way can access the
>  > value and pointwise attributes in JS might still be a bit
>  > broken/incomplete.
>  >
>  > On 28/02/2008, Beg-inner <[EMAIL PROTECTED]> wrote:
>  > >
>  > >
>  > > Hi all..
>  > >
>  > > Does anyone know how to make particles pointwise Size value to be 
> controlled
>  > > by an (animated) texture?
>  > > AFAIK, the attributes values can be held in a channel and be used in
>  > > materials to control/influence channels like Color and so on.. to make a
>  > > material that change color depending on the particle size.. but I cant 
> see
>  > > how to make it work the other way around.. Change size (or other Particle
>  > > specific channels) by a color ... or `?
>  > > (Do one need to use JS to get info back n  forth between a texture and 
> the
>  > > Pointwise Size attribute ?)
>  > >
>  > > I am going to use these controlled particles inside a Metaball level, to
>  > > create cool animated Metaball shapes.. =)
>  > >
>  > > Thx in advance..
>  > >
>  > > Take Care
>  > > Best Regards
>  > > SteFan Gustafsson ( Beg-inner )
>  > > A Proud Owner and User of Real3D and Realsoft3D..
>  >
>

Re: How can I control 1D particles Size, using an animated texture?

[Timo Mikkolainen]

Well, you can't access particle size from VSL (afaik), so the only way
I can think of is to assign the material to a custom channel in the
particles and then use JS to copy that value to the particle size.
Alternatively you could try to use a choreograph instead of js.
There's a few probs though: I don't know if either way can access the
value and pointwise attributes in JS might still be a bit
broken/incomplete.

On 28/02/2008, Beg-inner <[EMAIL PROTECTED]> wrote:
>
>
> Hi all..
>
> Does anyone know how to make particles pointwise Size value to be controlled
> by an (animated) texture?
> AFAIK, the attributes values can be held in a channel and be used in
> materials to control/influence channels like Color and so on.. to make a
> material that change color depending on the particle size.. but I cant see
> how to make it work the other way around.. Change size (or other Particle
> specific channels) by a color ... or `?
> (Do one need to use JS to get info back n  forth between a texture and the
> Pointwise Size attribute ?)
>
> I am going to use these controlled particles inside a Metaball level, to
> create cool animated Metaball shapes.. =)
>
> Thx in advance..
>
> Take Care
> Best Regards
> SteFan Gustafsson ( Beg-inner )
> A Proud Owner and User of Real3D and Realsoft3D..

Re: open question for an ice shader cooperation:-)

[Timo Mikkolainen]

A good start on all shaders is to think what the qualities of the
material are (i.e. what makes ice look like ice). The glass-like parts
are easy enough, you can pretty much just use a glass shader as a base
(with a tweaked color and optical thickness). The harder parts are the
surface features and air bubbles inside the ice. The surface features
can either be modeled or done with some creative bumpmapping (there's
some bump materials in the startup project that fit the bill). The air
bubbles are harder though: you can either model them or just cheat and
add a high frequency noise to the bump.
I'll see if I can give you some pointers once I get home.

On 15/04/2008, Frank Brübach <[EMAIL PROTECTED]> wrote:
> Dear RS friends, an open question:-)
>
>  "Perhaps somebody has fun to build with me an "ice shader" with the vsl 
> editor... I have started with a simple glas textur (from the rs manual, vsl 
> chapter), I have attached the example as a rar file above...
>
>  would be very nice to work with other realsoft guys to learn with the vsl 
> editor and get some interesting new vsl shader for ice, water and so on... "
>
>  servus, best regards, Frankolino
>
>  ...I have such things in mind: Refraction, Relief (Bump Map), Refraction, 
> Fresnel and more...
>  __
>  Bis 50 MB Dateianhänge? Kein Problem!
>  http://freemail.web.de/club/landingpage.htm/?mc=025556
>
>
>

Re: contest site down

[Timo Mikkolainen]

http://www.forummatrix.org seems like a good place to start. There's a
nice wizard for selecting a forum sw there.

On 14/04/2008, Boris Jahn <[EMAIL PROTECTED]> wrote:
>
>
> Hi Jan,
> yes I've got a backup of the forum. I hope that it'll be possible to export
> it to any other forum software but I guess it will be. Also no pictures will
> be lost. Also the contest will be running on. The actual topic is
> "Photorealism".  Sorry ... no panic ... but there are days >:(
> First of all the gallery will be back and then I'll search for a forum
> solution. First of all I've to find a new software to get rid of this
> cracking attacks..
>
>
> Bye
> Boris
>
>
>
>
> do you have a backup of the forum? all these threads contained very useful
> information and many nice RS images. I tried to visit 3ddart through a web
> archive (http://web.archive.org/web/*/3ddart.org) but there
> is only the September 2005 as the latest backup...
>
> Jan
>
>
>

Re: .OBJ IMPORT/EXPORT - Buggy ?

[Timo Mikkolainen]

See http://people.scs.fsu.edu/~burkardt/data/mtl/mtl.html and
http://local.wasp.uwa.edu.au/~pbourke/dataformats/mtl/
Basically VSL is out of the scope, V3 materials are pretty close to optimal.
Some subset of VSL might be doable though, as in textures/constants
mapped to color, illumination, specular sharpness/brightness,
transparency, bump and optical thickness. Reflection maps are
supported as well, but there's no set way to do them in RS.
MTL does have some support for procedural textures as well, but I
doubt if it'd be of any use to implement them.

On 07/04/2008, Arjo Rozendaal <[EMAIL PROTECTED]> wrote:
> Oh ok, I'm sorry. I guess this is not a bug but a shortcoming of the
>  exporter.
>  I think the Obj exporter should export .mtl files to make it possible to
>  export material definitions.
>  Yesterday I saw the same complaint on another forum. Photoshop CS3 needs the
>  mtl files to be able to show textured objects.
>  I've no idea what's supported by mtl files, or if it's possible to create
>  them with VSL code as a source.
>
>
>  Arjo.
>
>  > -Oorspronkelijk bericht-
>  > Van: [EMAIL PROTECTED] [mailto:owner-
>  > [EMAIL PROTECTED] Namens studio
>
> > Verzonden: maandag 7 april 2008 9:28
>
> > Aan: user-list@light.realsoft3d.com
>
> > Onderwerp: Re: .OBJ IMPORT/EXPORT - Buggy ?
>
> >
>  > > Hi Garry,
>  > > The Obj. Exporter is not broken. But you must set UV coords to "per
>  > face uv
>  > > cords".
>  > > I don't understand why this is not the default setting though
>  > > Arjo.
>  >
>
> > Hi Arjo :
>  >
>  > I did ! I followed the example in this (very nice!) Flash Video .
>
> >
>  > >> http://www.nic.fi/~vesamesk/texture_export.htm
>  >
>
> > Just look at the .gif attached to previous mail and you will
>  > see how RS does not have the proper VSL shader coding . I
>  > circled the missing part in red .
>  >
>  >  Maybe the .OBJ Importer is the culprit ? The Exporter does
>  > create a folder 'root/textures' , but the texture residing in
>  > that folder does not get applied via VSL when imported .
>  >
>  > Import the .OBJ I attached and you will see what I mean .
>  >
>  >  Seriously , something is broken there , I think . Perhaps
>  > Vesa used the much-sought-after V6 Demo for his tutorial ?
>  >
>  > garry
>
> >
>  >
>  > >> Bernie :
>  > >>
>  > >>   Is the .OBJ Import/Export broken for SDS Objects with Per Face
>  > >> textures ? (for instance , created by UV-Image)
>  > >>
>  > >>   I tried the same settings Vesa uses in his tree-export Flash
>  > tutorial
>  > >> , but with a simple object it will import geometry , but no texture
>  > >> file included . (see .GIF)
>  > >> http://www.nic.fi/~vesamesk/texture_export.htm
>  > >>
>  > >>
>  > >> Here's the original simple SDS Sphere (50KB link below) ,
>  > >> http://www.studiodynamics.net/uv/UV_image_tut_proj.zip
>  > >>
>  > >> Attached are the Exported .OBJ object/texture after the SDS object
>  > has
>  > >> been Face-Mapped via UV-Image .
>  > >>
>  > >> Not even a shred of info in the Manual about how-to import/export
>  > >> _Anything . No info about the .OBJ loader .
>  > >>
>  > >>   Seems odd to me that Vesa would create a tutorial about exporting
>  > SDS
>  > >> Face-Mapped trees , when the .OBJ Loader or Exporter is broken .
>  > >>
>  > >>   Maybe I'm doing something wrong ?
>  > >>
>  > >> Also , the imported .OBJ will not show up in the UV Editor .
>  > >>
>  > >> Thanks In Advance
>  > >>
>  > >> studio
>  > >
>
>

Re: look at this!

[Timo Mikkolainen]

Strictly speaking, it's photon mapping with a Java applet. It's nearly
as powerful (and in some cases even more so) as a native application.

On 17/03/2008, Gunnar Radeloff <[EMAIL PROTECTED]> wrote:
> Hi folks,
>
>  this is very frustrating! global illumination in a website...
>
>  http://www.cc.gatech.edu/~phlosoft/photon/
>
>
>  gunnar
>

Re: proper UV coordinates

[Timo Mikkolainen]

All cases of 'incomplete' or missing documentation should be listed
somewhere... there's so many cases that there's no way we can remember
them all =) The beta site has a breakdown of all the new features
which the beta testers can evaluate and comment on, maybe we need a
similar site/tool for the manual?

On 06/03/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> "proper UV coordinates"
>
>  are UVs which are usable for your operation :-?
>
>  It's not possible to say what is "proper" in your case.
>  Only that the faces in the UV-Space should not overlap :-?
>
>  Same QT can be: What is a good wheel?
>  OK, round.
>  Material ?
>  Size ?
>  Design ?
>
>  Do you need a manual, or paper ?
>
>  Garry on the toilet, who is shouting: Paaaper!
>  Nope, no please at the end :-p
>  Hmmm, good start for a new 3 Stooges story :-?
>  Let's have fun
>
>  Matthias
>
>
>  - Original Message -
>  From: "studio" <[EMAIL PROTECTED]>
>  To: 
>  Sent: Thursday, March 06, 2008 11:35 AM
>  Subject: Re: The RSV6 Viewer ... and Texture Baking ... solved for SDS !
>
>
>  > > Damn,
>  > > price is going to Frank (12 min faster ;-),
>  > > if it's solved now :-?
>  > > Matthias
>  >
>  > Hi Matt :
>  >
>  >   Several solutions . Hopefully it will ultimately be solved
>  > by RS in a future service pack .
>  >
>  >   As far as the 'UV-Image' Tool , I started a tutorial today
>  > about Realtime illumination mapping , got part way into it
>  > and had to give up . The attached message was a small clue ,
>  > but in the end I ran out of energy searching for more clues .
>  >
>  >   I know this is what you have been making an effort to point
>  > out recently , but I was trying to point out something diff-
>  > erent , at the same time , about realtime rendering bugs .
>  >
>  >   For Analytic it's a simple tutorial , but for SDS it is
>  > simply not do-able with V6.01.03 , unless we go into the
>  > world of workaround 'A' . (and maybe even workaround 'B')
>  >
>  >  Anyway , started to search the manual for info on this "proper
>  > UV coordinates"  predicament and gave up . It would have been
>  > very nice if a link had been placed there in the manual to wisk
>  > a user away to that section where one could explore this import-
>  > ant area of "proper UV coordinates" for SDS . (oh well)
>  >
>  >   A 'key word' Reference Manual search proved fruitless . I'm
>  > not saying the info is , or isn't in the manual , but this
>  > business of searching page-by-page is very time consuming .
>  >
>  > Perhaps more easily accesible info will make it's way into
>  > an expanded User Manual in the future ...
>  >
>  > love
>  > Daddy
>  >
>

Re: The RSV6 Viewer ... and Texture Baking

[Timo Mikkolainen]

Iirc you can set a map as an illumination map in the object
properties. I'd imagine this would show up in realtime as well. No
idea what (if anything) it uses for map/uv coords though.

On 28/02/2008, Mark Heuymans <[EMAIL PROTECTED]> wrote:
> >
>  > Yes, it's worth a try with RS. Baking illumination can work well, I used
>  > this trick for ceilings in GI scenes.
>
>
>
> Big problem: RS's OGL realtime engine ignores the Illumination channel in
>  VSL materials. How to make the illumaps show up in RS's own realtime engine?
>  Is there a walkaround? All the more because shadows would have to be baked
>  too, the RS engine doesn't do shadows, even though my graphics card does.
>
>  I'd love to say RS can do great realtime walkthroughs with pre-baked
>  illumination - would be great for architectural presentations - but no, it's
>  all about rendering. Baking can still be very useful for rendering btw. (but
>  tricky!)
>
>
>  -Mark
>
>

Re: website

[Timo Mikkolainen]

It's more likely to be a font size issue, jQuery is extensively tested
on all browsers. Try pressing ctrl + a few times, it causes the
sections to overlap. Nothing much you can do about it except changing
to a fluid layout (i.e. a layout that resizes on one or both axis
depending on content size).

On 07/02/2008, Frank Bueters <[EMAIL PROTECTED]> wrote:
> Hello Zaug,
>
> > Frank,
> >
> > Nice site and nice 3D, the boat images are especially well done.
>
> Thank you. Thanks for the kind comments of others too.
>
>
>
> > Perhaps it is just my browser settings, but some of the text
> > seems to get overlayed by other text and sub-headers.
>
> the website uses the jquery library, could that be it? In IE, Firefox and
> Safari, both on windows and mac all looked well when I tested.
>
>
> regards,
>
> Frank Bueters
>
>
>

Re: Mocap from video

[Timo Mikkolainen]

I think it's somewhat doable already, it's just a matter of combining
different technologies.
The biggest drawbacks of a single camera setup are occlusion and
missing depth information. Depth information can already be derived
from a single camera video rather well, but occlusion is still a hard
problem. There's no 'clean' way to get around it, all you can do is
quesstimate.
A similar problem I've been thinking of is tracking a (computer)
player's head (or eyes) to get the same functionality that TrackIR
has. It's a fairly controlled subset of the
single-camera-motion-tracking problem and would be an excellent start:
It would enable tracking with a single webcam, no additional HW
needed.

On 04/02/2008, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> Look at http://www.motion-capture.de/
>
> Videotrack is a motion capture software for movements.
>
> I'd like to see a software that captures movements from video and compines
> it to a character. Like "Shadow-Chaser", but the software would compere
> the movements - as well as the person.
> link ->
> http://www.5star-shareware.com/Windows/Hobby/Sport/shadow-chaser-screenshot.html
>
> I read some research papers and it might be possible in near future, but
> at the moment there are a lot of difficulties. (Problems in model-based
> approaches that use single video, but with extra information it might be
> done - also low cost. High end mocap solutions are working ofcourse.)
>
> Hannu
>
>
>
>

Re: HDR plugin

[Timo Mikkolainen]

http://home.vasteras2.net/henriknyqvist/hdr/
(took a while to find this, could everyone check that their sites'
links point to the right page?)

On 19/01/2008, Frank Bueters <[EMAIL PROTECTED]> wrote:
> Hi,
>
> Does anybody know what has happenend to Henrik Nyqvist's HDR plugin?
> http://home.swipnet.se/supersulo/r3hdr.zip is pretty dead. Perhaps someone
> will be so kind to mail r3hdr.zip to me?
>
> Kind regards,
>
> Frank Bueters
>
>
>

Re: How to add an 'accordion' client window like MiscTools?

[Timo Mikkolainen]

Classic mistake, forgot the link:
http://koti.welho.com/tmikkola/ColorPanel.tar.bz2

On 17/01/2008, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> Here's my 'color panel' in case someone wants to play around with it.
> Basically it's just a color palette, you can drag&drop colors from and
> to there: dragging a color to the view window will assign it to the
> selected object, dragging it over a color widget will set that assign
> the color to it. Just follow Matthias' steps except for the saving
> part.
>
> On 17/01/2008, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > Yep, it worked, although I had to edit the saved file a bit so it had
> > a different id than the gui item it was saved from.
> >
> > On 17/01/2008, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > > Sounds like a dirty enough trick to work ;) Will test in a few hours.
> > >
> > > On 17/01/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > > Maybe the simplest way:
> > > > Create a simple Available GUI Objects->Window->Client
> > > > (drag'nDrop it under the Select Window)
> > > > Then use the RMB over a "accordeon" menu (let's say "Advanced),
> > > > select save. Save it as "Test"
> > > > Use the RMB over your new Client Window and select "Replce".
> > > > Replace it with "Test".
> > > > Use RMB over the doubled Window, rename it and Put new tools in :-?
> > > >
> > > > Hope this was helpy,
> > > > Matthias
> > > > - Original Message -
> > > > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > > > To: 
> > > > Sent: Wednesday, January 16, 2008 10:40 PM
> > > > Subject: How to add an 'accordion' client window like MiscTools?
> > > >
> > > >
> > > > > I'd like to add one more client window under the select window, but no
> > > > > matter what I do, I can't make one with a horizontal 'header' (like in
> > > > > MiscTools). If someone knows how to do this, could you provide step by
> > > > > step instructions?
> > > > >
> > > >
> > >
> >
>

Re: How to add an 'accordion' client window like MiscTools?

[Timo Mikkolainen]

Here's my 'color panel' in case someone wants to play around with it.
Basically it's just a color palette, you can drag&drop colors from and
to there: dragging a color to the view window will assign it to the
selected object, dragging it over a color widget will set that assign
the color to it. Just follow Matthias' steps except for the saving
part.

On 17/01/2008, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> Yep, it worked, although I had to edit the saved file a bit so it had
> a different id than the gui item it was saved from.
>
> On 17/01/2008, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > Sounds like a dirty enough trick to work ;) Will test in a few hours.
> >
> > On 17/01/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > Maybe the simplest way:
> > > Create a simple Available GUI Objects->Window->Client
> > > (drag'nDrop it under the Select Window)
> > > Then use the RMB over a "accordeon" menu (let's say "Advanced),
> > > select save. Save it as "Test"
> > > Use the RMB over your new Client Window and select "Replce".
> > > Replace it with "Test".
> > > Use RMB over the doubled Window, rename it and Put new tools in :-?
> > >
> > > Hope this was helpy,
> > > Matthias
> > > - Original Message -
> > > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > > To: 
> > > Sent: Wednesday, January 16, 2008 10:40 PM
> > > Subject: How to add an 'accordion' client window like MiscTools?
> > >
> > >
> > > > I'd like to add one more client window under the select window, but no
> > > > matter what I do, I can't make one with a horizontal 'header' (like in
> > > > MiscTools). If someone knows how to do this, could you provide step by
> > > > step instructions?
> > > >
> > >
> >
>

Re: How to add an 'accordion' client window like MiscTools?

[Timo Mikkolainen]

Yep, it worked, although I had to edit the saved file a bit so it had
a different id than the gui item it was saved from.

On 17/01/2008, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> Sounds like a dirty enough trick to work ;) Will test in a few hours.
>
> On 17/01/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > Maybe the simplest way:
> > Create a simple Available GUI Objects->Window->Client
> > (drag'nDrop it under the Select Window)
> > Then use the RMB over a "accordeon" menu (let's say "Advanced),
> > select save. Save it as "Test"
> > Use the RMB over your new Client Window and select "Replce".
> > Replace it with "Test".
> > Use RMB over the doubled Window, rename it and Put new tools in :-?
> >
> > Hope this was helpy,
> > Matthias
> > - Original Message -
> > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > To: 
> > Sent: Wednesday, January 16, 2008 10:40 PM
> > Subject: How to add an 'accordion' client window like MiscTools?
> >
> >
> > > I'd like to add one more client window under the select window, but no
> > > matter what I do, I can't make one with a horizontal 'header' (like in
> > > MiscTools). If someone knows how to do this, could you provide step by
> > > step instructions?
> > >
> >
>

Re: How to add an 'accordion' client window like MiscTools?

[Timo Mikkolainen]

Sounds like a dirty enough trick to work ;) Will test in a few hours.

On 17/01/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Maybe the simplest way:
> Create a simple Available GUI Objects->Window->Client
> (drag'nDrop it under the Select Window)
> Then use the RMB over a "accordeon" menu (let's say "Advanced),
> select save. Save it as "Test"
> Use the RMB over your new Client Window and select "Replce".
> Replace it with "Test".
> Use RMB over the doubled Window, rename it and Put new tools in :-?
>
> Hope this was helpy,
> Matthias
> - Original Message -
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Wednesday, January 16, 2008 10:40 PM
> Subject: How to add an 'accordion' client window like MiscTools?
>
>
> > I'd like to add one more client window under the select window, but no
> > matter what I do, I can't make one with a horizontal 'header' (like in
> > MiscTools). If someone knows how to do this, could you provide step by
> > step instructions?
> >
>

Re: How to add an 'accordion' client window like MiscTools?

[Timo Mikkolainen]

V6. IIRC I did manage to do this in V5 a long time ago.

On 17/01/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> V6 or V5?
>
> Matthias
> - Original Message -----
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Wednesday, January 16, 2008 10:40 PM
> Subject: How to add an 'accordion' client window like MiscTools?
>
>
> > I'd like to add one more client window under the select window, but no
> > matter what I do, I can't make one with a horizontal 'header' (like in
> > MiscTools). If someone knows how to do this, could you provide step by
> > step instructions?
> >
>

How to add an 'accordion' client window like MiscTools?

[Timo Mikkolainen]

I'd like to add one more client window under the select window, but no
matter what I do, I can't make one with a horizontal 'header' (like in
MiscTools). If someone knows how to do this, could you provide step by
step instructions?

Re: HDRI

[Timo Mikkolainen]

Naturally, I just mean that the ratio between the images should be 0.5.

On 12/01/2008, Frank Bueters <[EMAIL PROTECTED]> wrote:
>
> > Yep, that would work very well. Just halve the illumination
> > between every image (as in 1, 0.5, 0.25, 0.125, 0.0625...).
>
> I'm trying.
> It should not only be halved, it should also be raised I think.
>
> Regards,
>
> Frank Bueters
>
>

Re: HDRI

[Timo Mikkolainen]

Yep, that would work very well. Just halve the illumination between
every image (as in 1, 0.5, 0.25, 0.125, 0.0625...).

On 11/01/2008, Frank Bueters <[EMAIL PROTECTED]> wrote:
>
> > I'm quite sure the same thing could be achieved with some VSL
> > trickery, but the shader it'd need would be rather
> > complicated. The best option really would be to make the
> > UVImage tool better (image format selection, proper shader
> > evaluation).
>
> Hi Timo,
>
> I'd be quite happy if the hdr image has to be assembeld outside rs3d.
> I'm just not sure that a collection of images with varying illumination
> levels would make a hdr image.
> The usual, photographers way is to take a number of pictures with different
> exposure levels (exposure bracketing) and assemble those in PS or HDRshop.
> Would varying illumination levels be the same?
>
> Proper shader evaluation would be nice too but not of (my) first concern.
>
>
> Frank Bueters
>
>

Re: Antwort: VR pano's

[Timo Mikkolainen]

I'm quite sure the same thing could be achieved with some VSL
trickery, but the shader it'd need would be rather complicated. The
best option really would be to make the UVImage tool better (image
format selection, proper shader evaluation).

On 10/01/2008, Frank Bueters <[EMAIL PROTECTED]> wrote:
>
> >
> > >> The camera/mirror sphere/HDRshop method can handle every advanced
> > >> scene illumination (should even work with V6 GI) but it's
> > very laborious...
> > >> Mark
> > >
> > >  The "camera/mirror sphere/HDRshop method" ? I'd be
> > guessing you are
> > > talking about using RS to capture HDRi images , in the same way one
> > > would do so , in real life ? However , it is just a guess .
> > >
> >
> >
> > 100% correct!
> >
> > >  The "very laborious" is also a bit of a mystery to me . Was this
> > > method covered here and I missed it's journey into the void, or is
> > > this considered to be 'common knowledge' , somehow ?
> > >
> > > thanks in advance
> > >
> > > garry
> >
> >
> > Here's a brief description:
> >
> > - make a small analytical sphere with mirror material (just
> > color=0 and
> > reflection=1)
> > - put a camera exactly underneath it and zoom in on the sphere
> > - put these in a level, and place it at a strategic position
> > in the scene and render. Make the camera Current if you
> > render to a file.
> >
> > I uploaded an example scene with such a setup:
> > http://www.athanor3d.com/pub/pano_volfog.zip
> >
> > Next, the HDRShop part. I noticed that they went commercial,
> > how about USD 400 for a single license and USD 16500 for a
> > 100-seat license!
> > http://projects.ict.usc.edu/graphics/HDRShop/
> > Version 1 is still free however, for non-commercial use.
> >
> > OK, here's the procedure:
> > - load a fisheye image rendered with the above method
> > - set Select->Draw Options to Circle
> > - draw a circle as accurately as possible around the sphere's edge
> > - crop the image (Image->Crop)
> > - Now the fun part: open the Image->Panorama->Panoramic
> > Transformations dialog
> > - Source image should be the loaded image, Format = Mirrored Ball
> > - click the Arbitrary Rotation button, then Settings: set
> > X-axis to 90 degrees
> > - Destination image: new image, Format = Latitude/Longitude
> > - OK and save!
> >
> > Now you have a similar projection to the UVimage evaluated
> > one but with full shading.
> >
> > I applied an image generated this way in RS to view it
> > http://www.athanor3d.com/pub/OGLpanoviewer.zip
> > Higher quality can be achieved with higher resolutions, but
> > the bottleneck is the 1024 max OGL texture resolution, for
> > rendering it can be much higher, of course.
> >
> >
> > hope this helps,
> > Mark
>
> Nice one Mark, I just went through the whole sequence. Had to adjust the cam
> for the mirroring bal, it's aspect is at 0, has to be 1, but that's ok.
> The clear advantage above the UVimage method is that a full rendered image
> is the result, with atmospheric effects and all.
> The drawback however is that the resulting image covers only half the
> sphere, where the UVimage method evaluates the full scene. I suppose that
> could be overcome by a second rendering, and merge the two images in an
> image editor.
> Don't know what I like best, depends on the use I suppose.
>
> One of the uses that you have mentionend is 360 degree HDR images for GI
> rendering.
> The UVimage method evaluates the illumination values from the reflective
> sphere and puts them in a BMP image. If we would/could scale the
> illumination values for the chrome material in a number of steps (7 or so)
> and assemble those in HDRshop? Would that make a HDR image?
>
> Regards,
>
> Frank Bueters
>
>
>
>

Re: VR pano's

[Timo Mikkolainen]

I actually wrote a small tutorial on this in the wiki:
http://rs3dwiki.the-final.info/index.php?title=Rendering_Environment_Maps
I didn't realize that you can just use a reflective material as the
evaluation basically looks straight at the surface =)
One fun project would be to create a material that does the same thing
without the sphere (using a camera backplane as a 'raycatcher' like my
DOF material): the good thing about such a solution is that you can
render to any format that RS supports (unlike material evaluation, I
think I already have an idea on the bugmanager for that).

On 08/01/2008, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Now my really last mail:
>
> Best results was with analtic sphere,
> chrome material without specular
> and high sample count <10 ...
>
> Testimage:
> http://the-final.com/eval.jpg
>
>
> Matthias
>
>
>
> - Original Message -
> From: "Frank Bueters" <[EMAIL PROTECTED]>
> To: 
> Sent: Tuesday, January 08, 2008 12:35 PM
> Subject: VR pano's
>
>
> > Hello Realfriends,
> >
> > Does anybody have any experience in making VR environments with RS3D? I mean
> > those panorama's mapped to cubes, cylinders or spheres, like in Quicktime
> > VR.
> >
> > I remember someone on this list once described the whole proces with some
> > free-, downloadable software and a script for the camera in RS3D (V4?).
> > Can't reftrieve this information and perhaps it's all very obsolete.
> >
> > So, I guess my question is, can anyone recommend a program or a tool to
> > create VR worlds from 3D?
> >
> > Kind regards,
> >
> > Frank Bueters
> >
> >
>

Re: Javascript, one more math function

[Timo Mikkolainen]

Use load instead of include, it reloads every time. Include is done
only once (i.e. it won't reload if you run the script again).

On 30/12/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> My favourite (used most often the good old print):
>
> // script begin
> var a = 1;
> var b = 2;
>
> print ("My first variable: "+a+"\n"+
> "var b on next line: "+b+"\n"+
> " and Pi is: "+Math.PI
> );
> // script end
>
> Textstrings in ""
> a + to add the next parts, maybe a variable
> and \n for linebreaks...
> Maybe somebody likes it, too ;-)
>
> But why is the "include" not working in all circumstances :-?
> Would be really great to work in external Editors, while
> checking the result in RS in play-animation-loop :-)
> The advance would be, too, if a bad script
> crashes RS, the script is not lost in total :-?
> And external Editors have syntax-highlighting...
> No more copy and paste...
> Reuse of scripts would be easier, too.
>
> Matthias
>
>
> > Matthias,
> > lol, Thankz for the explain, but what happened is that I tried to insert
> > the greek letter but it did not make it through the encoding; thus the
> > "?"  : D
> > I use it much and just did not want it to be left out of the list.
> >
> > Also, from fre-ber is some more handy math stuuf from the site you
> > linked earlier:
> >
> > http://www.javascripter.net/faq/arithmet.htm#top
> >
> > Cheerz,
> > Zaug
> >
>

Re: Javascript

[Timo Mikkolainen]

http://www.w3schools.com/js/js_obj_math.asp

On 25/12/2007, Alasdair <[EMAIL PROTECTED]> wrote:
> Hi Timo
> in my javascript book Math.random() is just random() so Ive got to guess
> which category each command may be referred to under realsoft3d
> Alasdair
>
> - Original Message -
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Tuesday, December 25, 2007 2:15 PM
> Subject: Re: Javascript
>
>
> > Basic stuff like Math.random() is the same as in your browser's JS, so
> > you can take a look at general JS tutorials for that.
> >
> > On 25/12/2007, Alasdair <[EMAIL PROTECTED]> wrote:
> >>
> >>
> >> Hi everybody
> >> I know you are all busy getting into v6 (havent learnt v5 enough & cant
> >> afford it yet) but can anybody help me please?
> >> Alasdair
> >>
> >>
> >> - Original Message -
> >> From: Alasdair
> >> To: user-list@light.realsoft3d.com
> >> Sent: Monday, December 24, 2007 5:19 PM
> >> Subject: Javascript
> >>
> >>
> >> where can I find a list of the commands such as Math.random() ?
> >> Alasdair
> >>
> >>
> >>  
> >>
> >>
> >> No virus found in this incoming message.
> >> Checked by AVG Free Edition.
> >> Version: 7.5.516 / Virus Database: 269.17.7/1194 - Release Date:
> >> 23/12/2007
> >> 17:27
> >>
> >>
> >>
> >
> >
> > --
> > No virus found in this incoming message.
> > Checked by AVG Free Edition.
> > Version: 7.5.516 / Virus Database: 269.17.8/1195 - Release Date:
> > 24/12/2007 11:19
> >
> >
>
>

Re: Javascript

[Timo Mikkolainen]

Basic stuff like Math.random() is the same as in your browser's JS, so
you can take a look at general JS tutorials for that.

On 25/12/2007, Alasdair <[EMAIL PROTECTED]> wrote:
>
>
> Hi everybody
> I know you are all busy getting into v6 (havent learnt v5 enough & cant
> afford it yet) but can anybody help me please?
> Alasdair
>
>
> - Original Message -
> From: Alasdair
> To: user-list@light.realsoft3d.com
> Sent: Monday, December 24, 2007 5:19 PM
> Subject: Javascript
>
>
> where can I find a list of the commands such as Math.random() ?
> Alasdair
>
>
>  
>
>
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.5.516 / Virus Database: 269.17.7/1194 - Release Date: 23/12/2007
> 17:27
>
>
>

Re: Is there a way to access sds handle selection lists though JS?

[Timo Mikkolainen]

I tried adding them to r3subdiv.js myself:
R3SUBDIVA_FaceSelection = 1235506;
function GetR3SUBDIVA_FaceSelection() {
  return R3Get(this.r3obj, R3SUBDIVA_FaceSelection, R3TID_INTEGER,
R3TNF_ARRAY);
}

R3SUBDIVA_EdgeSelection = 1235507;
function GetR3SUBDIVA_EdgeSelection() {
  return R3Get(this.r3obj, R3SUBDIVA_EdgeSelection, R3TID_INTEGER,
R3TNF_ARRAY);
}

R3SUBDIVA_VertexSelection = 1235508;
function GetR3SUBDIVA_VertexSelection() {
  return R3Get(this.r3obj, R3SUBDIVA_VertexSelection, R3TID_INTEGER,
R3TNF_ARRAY);
}


   this.GetFaceSelection=GetR3SUBDIVA_FaceSelection;
   this.GetEdgeSelection=GetR3SUBDIVA_EdgeSelection;
   this.GetVertexSelection=GetR3SUBDIVA_VertexSelection;

they do return something, but I have no idea what... most likely
pointers or something equally useful.

On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> Aand none of the R3SUBDIVA types are defined in js, so meh.
>
> On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > (Boring day at work...)
> > I looked through the mailing list archives and found something on this 
> > issue:
> > http://www.mail-archive.com/user-list@light.realsoft3d.com/msg03616.html
> > It looks like all kinds of array related stuff is busted in JS. I'm
> > not sure if this is even fixed in V6 yet (at least bug 2097 is still
> > open and unconfirmed). I'll test all this when I get home.
> >
> > On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > > Looking at the Low Level API section in the reference manual, I'd
> > > quess the correct call would actually be something like:
> > > var faceSelection = R3Get(sdsObject, R3SUBDIVA_FaceSelection,
> > > R3TID_INTEGER, R3TNS_ARRAY);
> > > I'm not sure of the typeid and array flag, they might very well be
> > > wrong. It might also need some additional flags (like R3TNF_READONLY
> > > and R3TNF_BYVALUE).
> > >
> > > On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > > > Looking at r3subdiv.h, it does have these:
> > > > R3SUBDIVA_FaceSelection,
> > > > R3SUBDIVA_EdgeSelection,
> > > > R3SUBDIVA_VertexSelection
> > > > that apparently hold the selection data.
> > > > I'd imagine these can be (somehow) accessed with R3GetAttrs(), but I
> > > > haven't tested it yet.
> > > > I could only find examples of this on the SDK side, I don't even know
> > > > if the function exists on the JS side. If it does, I have no idea what
> > > > to pass as params (using the SDK it would be something like
> > > > R3GetAttrs(sdsObject, R3SUBDIVA_FaceSelection, &faceSelection,
> > > > R3RA_Done);).
> > > >
> > > > On 16/12/2007, Juha Sunnari <[EMAIL PROTECTED]> wrote:
> > > > > Eh, nevermind, further investigation reveals that there are methods in
> > > > > r3prim.js
> > > > > called SELECTNEXT, SELECTUP, SELECTPREV, etc...
> > > > >
> > > > > Juha Sunnari kirjoitti:
> > > > > > That's what I would like to know, too. In the sdk it looks like
> > > > > > adjacency information for edges
> > > > > > and faces are stored in R3EDGELISTNODE and R3FACEPEREDGENODE.
> > > > > > structs defined in r3subdiv.h. However I haven't really looked into
> > > > > > how to manipulate these in js or
> > > > > > if it is even possible at the moment. Maybe the RS folks could give
> > > > > > some information on this. :)
> > > > > >
> > > > > > Timo Mikkolainen kirjoitti:
> > > > > >> I'm trying to create a generic function library for SDS 
> > > > > >> manipulation,
> > > > > >> starting with modifying the handle selection (e.g. selecting an 
> > > > > >> edge
> > > > > >> loop or -ring, selecting connected faces etc.). Going through all
> > > > > >> handles and checking which ones are selected is slow and hack-ish 
> > > > > >> and
> > > > > >> I can't access the selection order. I take it that the selected
> > > > > >> handles (with the correct order) are stored as tags in the sds 
> > > > > >> object,
> > > > > >> but I don't know how to access them through JS, any ideas?
> > > > >
> > > >
> > >
> >
>

Re: Is there a way to access sds handle selection lists though JS?

[Timo Mikkolainen]

Aand none of the R3SUBDIVA types are defined in js, so meh.

On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> (Boring day at work...)
> I looked through the mailing list archives and found something on this issue:
> http://www.mail-archive.com/user-list@light.realsoft3d.com/msg03616.html
> It looks like all kinds of array related stuff is busted in JS. I'm
> not sure if this is even fixed in V6 yet (at least bug 2097 is still
> open and unconfirmed). I'll test all this when I get home.
>
> On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > Looking at the Low Level API section in the reference manual, I'd
> > quess the correct call would actually be something like:
> > var faceSelection = R3Get(sdsObject, R3SUBDIVA_FaceSelection,
> > R3TID_INTEGER, R3TNS_ARRAY);
> > I'm not sure of the typeid and array flag, they might very well be
> > wrong. It might also need some additional flags (like R3TNF_READONLY
> > and R3TNF_BYVALUE).
> >
> > On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > > Looking at r3subdiv.h, it does have these:
> > > R3SUBDIVA_FaceSelection,
> > > R3SUBDIVA_EdgeSelection,
> > > R3SUBDIVA_VertexSelection
> > > that apparently hold the selection data.
> > > I'd imagine these can be (somehow) accessed with R3GetAttrs(), but I
> > > haven't tested it yet.
> > > I could only find examples of this on the SDK side, I don't even know
> > > if the function exists on the JS side. If it does, I have no idea what
> > > to pass as params (using the SDK it would be something like
> > > R3GetAttrs(sdsObject, R3SUBDIVA_FaceSelection, &faceSelection,
> > > R3RA_Done);).
> > >
> > > On 16/12/2007, Juha Sunnari <[EMAIL PROTECTED]> wrote:
> > > > Eh, nevermind, further investigation reveals that there are methods in
> > > > r3prim.js
> > > > called SELECTNEXT, SELECTUP, SELECTPREV, etc...
> > > >
> > > > Juha Sunnari kirjoitti:
> > > > > That's what I would like to know, too. In the sdk it looks like
> > > > > adjacency information for edges
> > > > > and faces are stored in R3EDGELISTNODE and R3FACEPEREDGENODE.
> > > > > structs defined in r3subdiv.h. However I haven't really looked into
> > > > > how to manipulate these in js or
> > > > > if it is even possible at the moment. Maybe the RS folks could give
> > > > > some information on this. :)
> > > > >
> > > > > Timo Mikkolainen kirjoitti:
> > > > >> I'm trying to create a generic function library for SDS manipulation,
> > > > >> starting with modifying the handle selection (e.g. selecting an edge
> > > > >> loop or -ring, selecting connected faces etc.). Going through all
> > > > >> handles and checking which ones are selected is slow and hack-ish and
> > > > >> I can't access the selection order. I take it that the selected
> > > > >> handles (with the correct order) are stored as tags in the sds 
> > > > >> object,
> > > > >> but I don't know how to access them through JS, any ideas?
> > > >
> > >
> >
>

Re: Is there a way to access sds handle selection lists though JS?

[Timo Mikkolainen]

(Boring day at work...)
I looked through the mailing list archives and found something on this issue:
http://www.mail-archive.com/user-list@light.realsoft3d.com/msg03616.html
It looks like all kinds of array related stuff is busted in JS. I'm
not sure if this is even fixed in V6 yet (at least bug 2097 is still
open and unconfirmed). I'll test all this when I get home.

On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> Looking at the Low Level API section in the reference manual, I'd
> quess the correct call would actually be something like:
> var faceSelection = R3Get(sdsObject, R3SUBDIVA_FaceSelection,
> R3TID_INTEGER, R3TNS_ARRAY);
> I'm not sure of the typeid and array flag, they might very well be
> wrong. It might also need some additional flags (like R3TNF_READONLY
> and R3TNF_BYVALUE).
>
> On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> > Looking at r3subdiv.h, it does have these:
> > R3SUBDIVA_FaceSelection,
> > R3SUBDIVA_EdgeSelection,
> > R3SUBDIVA_VertexSelection
> > that apparently hold the selection data.
> > I'd imagine these can be (somehow) accessed with R3GetAttrs(), but I
> > haven't tested it yet.
> > I could only find examples of this on the SDK side, I don't even know
> > if the function exists on the JS side. If it does, I have no idea what
> > to pass as params (using the SDK it would be something like
> > R3GetAttrs(sdsObject, R3SUBDIVA_FaceSelection, &faceSelection,
> > R3RA_Done);).
> >
> > On 16/12/2007, Juha Sunnari <[EMAIL PROTECTED]> wrote:
> > > Eh, nevermind, further investigation reveals that there are methods in
> > > r3prim.js
> > > called SELECTNEXT, SELECTUP, SELECTPREV, etc...
> > >
> > > Juha Sunnari kirjoitti:
> > > > That's what I would like to know, too. In the sdk it looks like
> > > > adjacency information for edges
> > > > and faces are stored in R3EDGELISTNODE and R3FACEPEREDGENODE.
> > > > structs defined in r3subdiv.h. However I haven't really looked into
> > > > how to manipulate these in js or
> > > > if it is even possible at the moment. Maybe the RS folks could give
> > > > some information on this. :)
> > > >
> > > > Timo Mikkolainen kirjoitti:
> > > >> I'm trying to create a generic function library for SDS manipulation,
> > > >> starting with modifying the handle selection (e.g. selecting an edge
> > > >> loop or -ring, selecting connected faces etc.). Going through all
> > > >> handles and checking which ones are selected is slow and hack-ish and
> > > >> I can't access the selection order. I take it that the selected
> > > >> handles (with the correct order) are stored as tags in the sds object,
> > > >> but I don't know how to access them through JS, any ideas?
> > >
> >
>

Re: Is there a way to access sds handle selection lists though JS?

[Timo Mikkolainen]

Looking at the Low Level API section in the reference manual, I'd
quess the correct call would actually be something like:
var faceSelection = R3Get(sdsObject, R3SUBDIVA_FaceSelection,
R3TID_INTEGER, R3TNS_ARRAY);
I'm not sure of the typeid and array flag, they might very well be
wrong. It might also need some additional flags (like R3TNF_READONLY
and R3TNF_BYVALUE).

On 17/12/2007, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> Looking at r3subdiv.h, it does have these:
> R3SUBDIVA_FaceSelection,
> R3SUBDIVA_EdgeSelection,
> R3SUBDIVA_VertexSelection
> that apparently hold the selection data.
> I'd imagine these can be (somehow) accessed with R3GetAttrs(), but I
> haven't tested it yet.
> I could only find examples of this on the SDK side, I don't even know
> if the function exists on the JS side. If it does, I have no idea what
> to pass as params (using the SDK it would be something like
> R3GetAttrs(sdsObject, R3SUBDIVA_FaceSelection, &faceSelection,
> R3RA_Done);).
>
> On 16/12/2007, Juha Sunnari <[EMAIL PROTECTED]> wrote:
> > Eh, nevermind, further investigation reveals that there are methods in
> > r3prim.js
> > called SELECTNEXT, SELECTUP, SELECTPREV, etc...
> >
> > Juha Sunnari kirjoitti:
> > > That's what I would like to know, too. In the sdk it looks like
> > > adjacency information for edges
> > > and faces are stored in R3EDGELISTNODE and R3FACEPEREDGENODE.
> > > structs defined in r3subdiv.h. However I haven't really looked into
> > > how to manipulate these in js or
> > > if it is even possible at the moment. Maybe the RS folks could give
> > > some information on this. :)
> > >
> > > Timo Mikkolainen kirjoitti:
> > >> I'm trying to create a generic function library for SDS manipulation,
> > >> starting with modifying the handle selection (e.g. selecting an edge
> > >> loop or -ring, selecting connected faces etc.). Going through all
> > >> handles and checking which ones are selected is slow and hack-ish and
> > >> I can't access the selection order. I take it that the selected
> > >> handles (with the correct order) are stored as tags in the sds object,
> > >> but I don't know how to access them through JS, any ideas?
> >
>

Re: Is there a way to access sds handle selection lists though JS?

[Timo Mikkolainen]

Looking at r3subdiv.h, it does have these:
R3SUBDIVA_FaceSelection,
R3SUBDIVA_EdgeSelection,
R3SUBDIVA_VertexSelection
that apparently hold the selection data.
I'd imagine these can be (somehow) accessed with R3GetAttrs(), but I
haven't tested it yet.
I could only find examples of this on the SDK side, I don't even know
if the function exists on the JS side. If it does, I have no idea what
to pass as params (using the SDK it would be something like
R3GetAttrs(sdsObject, R3SUBDIVA_FaceSelection, &faceSelection,
R3RA_Done);).

On 16/12/2007, Juha Sunnari <[EMAIL PROTECTED]> wrote:
> Eh, nevermind, further investigation reveals that there are methods in
> r3prim.js
> called SELECTNEXT, SELECTUP, SELECTPREV, etc...
>
> Juha Sunnari kirjoitti:
> > That's what I would like to know, too. In the sdk it looks like
> > adjacency information for edges
> > and faces are stored in R3EDGELISTNODE and R3FACEPEREDGENODE.
> > structs defined in r3subdiv.h. However I haven't really looked into
> > how to manipulate these in js or
> > if it is even possible at the moment. Maybe the RS folks could give
> > some information on this. :)
> >
> > Timo Mikkolainen kirjoitti:
> >> I'm trying to create a generic function library for SDS manipulation,
> >> starting with modifying the handle selection (e.g. selecting an edge
> >> loop or -ring, selecting connected faces etc.). Going through all
> >> handles and checking which ones are selected is slow and hack-ish and
> >> I can't access the selection order. I take it that the selected
> >> handles (with the correct order) are stored as tags in the sds object,
> >> but I don't know how to access them through JS, any ideas?
>

Is there a way to access sds handle selection lists though JS?

[Timo Mikkolainen]

I'm trying to create a generic function library for SDS manipulation,
starting with modifying the handle selection (e.g. selecting an edge
loop or -ring, selecting connected faces etc.). Going through all
handles and checking which ones are selected is slow and hack-ish and
I can't access the selection order. I take it that the selected
handles (with the correct order) are stored as tags in the sds object,
but I don't know how to access them through JS, any ideas?

Re: V6 - (Ambient Occlusion)

[Timo Mikkolainen]

I'll make a V6 version of AO (with the end-userified gui we did with
Arjo) once I get the time. I was supposed to do it last weekend but
I've been too busy (sorry Jyrki).
What comes to the speed issue, the V5 version is indeed as slow if not
even slower than one bounce GI. The V6 version should be significantly
faster as it doesn't do unneeded target surface material evaluations.
The Arnold renders used GI, no AO there.

On 02/12/2007, studio <[EMAIL PROTECTED]> wrote:
> Hi Frank :
>
>   Not sure if you caught the thread ~ 1 year ago , when Arjo sent
> the list his reworked version of Timo's AO ? That shader is on
> the RS.info site (turn down the AA for best results) .
> http://realsoft.info/files/projects/AmbientOcclusion.zm9
>
>   Thanks for the link . Perhaps you're right about AO being an
> excellent realtime tool for lighting . It appears Realsoft
> likes it too , and has made speed improvements etc in V6 .
>
>   Looking forward to trying them out .
>
> garry
>
>
> > studio wrote:
> > >  Always been a big fan of AO (Ambient Occlusion) so am looking
> > > forward to V6 and it's included AO enhancements . Not aware of
> > > anyone here using AO in their current RS renders , but perhaps
> > > that will change with V6 .
> > >
> > >  Haven't really been keeping current with AO technically how-
> > > ever , but looking back on the original Arnold renderings I
> > > can't help but wonder if AO wasn't a big ingredient and a
> > > large factor in the success of that quantum leap .
> > >
> > > http://www.3dluvr.com/marcosss/morearni/index.html
> > >
> > >   I always thought it was a pure Gi implimentation and so was
> > > always blown away , but perhaps I assumed wrong and was simply
> > > too 'newbie' to realise just how much a significant factor AO
> > > was in those renders ?
> > >
> > > garry
> > >
> > I must admit to have never looked too into AO in any great detail,
> > however recently I have seen quite a few references to it as a real time
> > technique for rendering realistic lighting effects onto models. A simple
> > example can be seen here
> > http://www.ozone3d.net/tutorials/ambient_occlusion.php . I wonder if
> > these are the starting steps of some of the final progressions of the
> > real time graphics technologies into the areas of more traditionally
> > processed rendering.
> >
> > Frank "The Rookie" Dodd
> >
>

Re: How to use loop through with js?

[Timo Mikkolainen]

Sorry for the late reply, I haven't touched js for a while.
I'm building an sds function library atm, once I get the basics
functioning I'll see if I can magic up a working version of the add
loop function.

On 19/10/2007, K-UDA <[EMAIL PROTECTED]> wrote:
> Hello, Timo.
> I added a script.
> However, it is still unfinished.
> I wish this helps you.
> Best Regards
> K-UDA
>
> - Original Message -
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Wednesday, October 17, 2007 5:00 AM
> Subject: How to use loop through with js?
>
>
> > I'm trying to make a simple-ish script that adds an edge loop to the
> > selected object.
> > The only part missing is the edge 'ring' selection, i.e.
> > ++loop tool. I'm having a hard time finding any
> > references to this in the js files. Apparently it should be in
> > sdsplus, but things there look mighty empty.
>
>

Re: Field evaluator in action

[Timo Mikkolainen]

I looked at the docs, which say the arg should be "Number, address of
result value". That doesn't help much... anyways: the method doesn't
change the value of the given arg and doesn't return anything, so it
seems to be broken. Or maybe there's a magic way of passing a pointer
in js =P

On 18/10/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Play the animation, hope something is going on then.
>
> Open the choregraphy window and modify the:
>  one-wheel-driver <- script inside
> path
> pcurve
> min curve chor, to adjust the back and forward movement.
>
> If you modify the pcurve nurbs you have to set manually a new
> length :-?
>
> Open Properties->Script tab of:
> one-wheel-driver <- script inside
>
> and modify "len" value.
>
> If somebody knows how to use this on a nurbs curve:
>
> len = Self.MEASURELENGTH();
> and avoid this:
> error("MEASURELENGTH() needs 1 parameters");
>
> What the ..., is the needed parameter :-?
> Example is welcome :-)
>
> feel free to ask ;-)
>
> Matthias
>

How to use loop through with js?

[Timo Mikkolainen]

I'm trying to make a simple-ish script that adds an edge loop to the
selected object.
The only part missing is the edge 'ring' selection, i.e.
++loop tool. I'm having a hard time finding any
references to this in the js files. Apparently it should be in
sdsplus, but things there look mighty empty.

Re: getting closer

[Timo Mikkolainen]

So you applied it to two spheres around the lights, interesting =) I
applied it straight to the visor, but your way actually looks better.

On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> My fault, I'm (sometimes) a little bit stupid.
>
> Files updated (screenshot+*.r3d)
>
> http://the-final.com/robo_walker/
>
> Both effects combined.
> Maybe you can tell me if I've assigned
> you're material right :-?
>
> Looks like an LED with some shine to me :-)
>
> Many thanks again,
> Matthias
>
>
> - Original Message -
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Friday, September 28, 2007 12:57 PM
> Subject: Re: getting closer
>
>
> > The project I attached in the last mail should work.
> >
> > On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > Maybe I'm doing something wrong :-?
> > >
> > > Is it possible to set up a simple scene with your
> > > LED shader in action ?
> > >
> > > Matthias
> > >
> > > - Original Message -
> > > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > > To: 
> > > Sent: Friday, September 28, 2007 11:24 AM
> > > Subject: Re: getting closer
> > >
> > >
> > > > I took a look at the project file and fixed the material a bit: it now
> > > > uses a curve to control the sharpness and it works with shadow casting
> > > > lights. Place the 'eye' lights inside the visor to get the proper
> > > > effect. You might have to boost the light brightness quite a bit, I
> > > > had them at 5. Also add a bit of green and even a bit less blue to the
> > > > light color to get the orange-ish highlights you see on photos. You
> > > > might have to tweak the light falloffs a bit to get a proper effect
> > > > (the 'eye' that's farther from the camera tends to look too bright).
> > > >
> > > > On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > > > Many thanks for the shader :-)
> > > > >
> > > > > Please have a look here:
> > > > >
> > > > > http://the-final.com/robo_walker/
> > > > >
> > > > > some images and the scene file....
> > > > > The screenshot.jpg shows nearly what I'm
> > > > > looking for, but it should be a little bit more "glow".
> > > > >
> > > > > Maybe to much reflection?
> > > > >
> > > > > Only found the reality images in the folder...
> > > > >
> > > > > It's a combination of reality and "how I think"
> > > > > what it should look like. Can't really explain it.
> > > > > But my todays result is not exactly what I'm looking for :-?
> > > > >
> > > > > Matthias
> > > > >
> > > > > - Original Message -
> > > > > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > > > > To: 
> > > > > Sent: Friday, September 28, 2007 9:41 AM
> > > > > Subject: Re: getting closer
> > > > >
> > > > >
> > > > > > A photo of the effect you're after might help. Here's a quick 
> > > > > > cookup;
> > > > > > don't set Sharpness over 10, the effect will disappear.
> > > > > >
> > > > > > On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > > > > > Has anybody a good idea/shader
> > > > > > > to simulate LEDs behind black glass?
> > > > > > >
> > > > > > > www.the-final.com/b_daempfer.jpg
> > > > > > >
> > > > > > > Matthias
> > > > > > > - Original Message -
> > > > > > > From: "Mark Heuymans" <[EMAIL PROTECTED]>
> > > > > > > To: 
> > > > > > > Sent: Thursday, September 27, 2007 8:14 PM
> > > > > > > Subject: Re: getting closer
> > > > > > >
> > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > > www.the-final.com/b_daempfer.jpg
> > > > > > > > >
> > > > > > > > > Matthias
> > > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > Matthias,
> > > > > > > >
> > > > > > > > Great work again and thanks for the info!
> > > > > > > >
> > > > > > > > cheers,
> > > > > > > > Mark H
> > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>

Re: getting closer

[Timo Mikkolainen]

The project I attached in the last mail should work.

On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Maybe I'm doing something wrong :-?
>
> Is it possible to set up a simple scene with your
> LED shader in action ?
>
> Matthias
>
> ----- Original Message -
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Friday, September 28, 2007 11:24 AM
> Subject: Re: getting closer
>
>
> > I took a look at the project file and fixed the material a bit: it now
> > uses a curve to control the sharpness and it works with shadow casting
> > lights. Place the 'eye' lights inside the visor to get the proper
> > effect. You might have to boost the light brightness quite a bit, I
> > had them at 5. Also add a bit of green and even a bit less blue to the
> > light color to get the orange-ish highlights you see on photos. You
> > might have to tweak the light falloffs a bit to get a proper effect
> > (the 'eye' that's farther from the camera tends to look too bright).
> >
> > On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > Many thanks for the shader :-)
> > >
> > > Please have a look here:
> > >
> > > http://the-final.com/robo_walker/
> > >
> > > some images and the scene file
> > > The screenshot.jpg shows nearly what I'm
> > > looking for, but it should be a little bit more "glow".
> > >
> > > Maybe to much reflection?
> > >
> > > Only found the reality images in the folder...
> > >
> > > It's a combination of reality and "how I think"
> > > what it should look like. Can't really explain it.
> > > But my todays result is not exactly what I'm looking for :-?
> > >
> > > Matthias
> > >
> > > - Original Message -
> > > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > > To: 
> > > Sent: Friday, September 28, 2007 9:41 AM
> > > Subject: Re: getting closer
> > >
> > >
> > > > A photo of the effect you're after might help. Here's a quick cookup;
> > > > don't set Sharpness over 10, the effect will disappear.
> > > >
> > > > On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > > > Has anybody a good idea/shader
> > > > > to simulate LEDs behind black glass?
> > > > >
> > > > > www.the-final.com/b_daempfer.jpg
> > > > >
> > > > > Matthias
> > > > > - Original Message -
> > > > > From: "Mark Heuymans" <[EMAIL PROTECTED]>
> > > > > To: 
> > > > > Sent: Thursday, September 27, 2007 8:14 PM
> > > > > Subject: Re: getting closer
> > > > >
> > > > >
> > > > > >
> > > > > >
> > > > > > > www.the-final.com/b_daempfer.jpg
> > > > > > >
> > > > > > > Matthias
> > > > > > >
> > > > > >
> > > > > >
> > > > > > Matthias,
> > > > > >
> > > > > > Great work again and thanks for the info!
> > > > > >
> > > > > > cheers,
> > > > > > Mark H
> > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>

Re: getting closer

[Timo Mikkolainen]

I took a look at the project file and fixed the material a bit: it now
uses a curve to control the sharpness and it works with shadow casting
lights. Place the 'eye' lights inside the visor to get the proper
effect. You might have to boost the light brightness quite a bit, I
had them at 5. Also add a bit of green and even a bit less blue to the
light color to get the orange-ish highlights you see on photos. You
might have to tweak the light falloffs a bit to get a proper effect
(the 'eye' that's farther from the camera tends to look too bright).

On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Many thanks for the shader :-)
>
> Please have a look here:
>
> http://the-final.com/robo_walker/
>
> some images and the scene file
> The screenshot.jpg shows nearly what I'm
> looking for, but it should be a little bit more "glow".
>
> Maybe to much reflection?
>
> Only found the reality images in the folder...
>
> It's a combination of reality and "how I think"
> what it should look like. Can't really explain it.
> But my todays result is not exactly what I'm looking for :-?
>
> Matthias
>
> - Original Message -
> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> To: 
> Sent: Friday, September 28, 2007 9:41 AM
> Subject: Re: getting closer
>
>
> > A photo of the effect you're after might help. Here's a quick cookup;
> > don't set Sharpness over 10, the effect will disappear.
> >
> > On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> > > Has anybody a good idea/shader
> > > to simulate LEDs behind black glass?
> > >
> > > www.the-final.com/b_daempfer.jpg
> > >
> > > Matthias
> > > - Original Message -
> > > From: "Mark Heuymans" <[EMAIL PROTECTED]>
> > > To: 
> > > Sent: Thursday, September 27, 2007 8:14 PM
> > > Subject: Re: getting closer
> > >
> > >
> > > >
> > > >
> > > > > www.the-final.com/b_daempfer.jpg
> > > > >
> > > > > Matthias
> > > > >
> > > >
> > > >
> > > > Matthias,
> > > >
> > > > Great work again and thanks for the info!
> > > >
> > > > cheers,
> > > > Mark H
> > > >
> > > >
> > >
> >
>


LedGlass.r3d
Description: Binary data

Re: getting closer

[Timo Mikkolainen]

A photo of the effect you're after might help. Here's a quick cookup;
don't set Sharpness over 10, the effect will disappear.

On 28/09/2007, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Has anybody a good idea/shader
> to simulate LEDs behind black glass?
>
> www.the-final.com/b_daempfer.jpg
>
> Matthias
> - Original Message -
> From: "Mark Heuymans" <[EMAIL PROTECTED]>
> To: 
> Sent: Thursday, September 27, 2007 8:14 PM
> Subject: Re: getting closer
>
>
> >
> >
> > > www.the-final.com/b_daempfer.jpg
> > >
> > > Matthias
> > >
> >
> >
> > Matthias,
> >
> > Great work again and thanks for the info!
> >
> > cheers,
> > Mark H
> >
> >
>


LedGlass.r3d
Description: Binary data

Re: External Standalone Render Apps and REALSOFT, maybe this one ....

[Timo Mikkolainen]

The problem is that most of these are polygon renderers and even those
that can handle csg stuff would take a huge amount of work on the
material side.

On 07/08/07, Aidan O Driscoll <[EMAIL PROTECTED]> wrote:
> Hi Jean-Sebastien,
>
> As I have discovered from my travels with Blender their are numerous
> OPEN SOURCE Free Standalone Render apps of different flavours:
>
> http://www.indigorenderer.com/joomla/
> http://sunflow.sourceforge.net
>
> AND for those who favour RENDERMAN have a look here:
>
> http://www.3delight.com/en/
> http://www.aqsis.org/xoops/modules/news/
> http://www.cs.utexas.edu/~okan/Pixie/pixie.htm
> http://jrman.sourceforge.net/
>
> All of these - opensource also. 3DElight is free ish, but looks real good.
>
> Yip - Jean, if some of our RS Coders could do some plugins for RS [
> as is the case with Blender ] it would be excellent .. Hint Hint :)
> Primarily, I stress because these apps are FREEE & Open Source.
>
> Cheers
> Aidan
>
>
>
> At 14:04 06/08/2007, you wrote:
> >Aidan O Driscoll wrote:
> >>Hi Guys,
> >>
> >>Been using Blender [ v2.44 ] on and off over the last few months.
> >>One thing that Blender has is access to a lot of FRE EXTERNAL
> >>standalone Render Apps. These include INDIGO, SUNFLOW and PIXIE [
> >>Renderman ],  [ google the words to find the sites ] but the one
> >>for here is KERKYTHEA:
> >>
> >>http://www.kerkythea.net/joomla/
> >>
> >>Their is a special build of Blender which has Kerkythea integrated,
> >>but today I launched Kerkythea itself to find, in the File, Open
> >>menu it can import OBJ etc. So this obviously gives an in for
> >>REALSOFT. Model in RS, export as OBJ and then open in Kerkythea for
> >>an alternative render [ as one may do with Maxwell etc, but at huge
> >>cost as opposed to this being free ]. This standalone render app
> >>has GI implementation etc, in fact read this:
> >>
> >>
> >>/Supported Rendering Techniques
> >> - Classic Ray Tracing
> >> - Path Tracing (Kajiya)
> >> - Bidirectional Path Tracing (Veach & Guibas)
> >> - Metropolis Light Transport (Kelemen, Kalos et al.)
> >> - Photon Mapping (Jensen) [mesh maps, photon maps, final
> >> gathering, irradiance caching, caustics]
> >> - Diffuse Interreflection (Ward)
> >> - Depth Rendering
> >> - Mask Rendering
> >> - Clay Rendering
> >>
> >>/as per features page here:
> >>
> >>http://www.kerkythea.net/joomla/index.php?option=com_content&task=view&id=14&Itemid=48
> >>
> >>
> >>
> >>This one of course is for all you doodlers. Maybe some fun. AND I
> >>stress, this mail is not set up  for comparison reasons OR to have
> >>a stab at RS, just may be of interest to some.
> >>
> >>It has its own materials workshop, lights, cameras etc.
> >>
> >>Cheers
> >>Aidan
> >I wish Realsoft could be used as a  command line standalone renderer
> >with a RIB interface.
> >
> >Jean-Sebastien Perron
> >www.NeuroWorld.ws
> >
> >
> >
> >--
> >No virus found in this incoming message.
> >Checked by AVG Free Edition. Version: 7.5.476 / Virus Database:
> >269.11.8/940 - Release Date: 06/08/2007 16:53
>
>

Re: Script to make curve.

[Timo Mikkolainen]

Hi K-UDA,
the script is great, I have no idea how you managed to do half of the
stuff in it =)
The video tutorial is nice too, it quickly shows the ins and outs of the script.

On 25/07/07, K-UDA <[EMAIL PROTECTED]> wrote:

Hello List.
I wrote a script to create a curve.
This script creates the nurbs curve on the chosen SDS surface.
I wish this script helps the work of your modeling.
Download from the following page.
http://www2s.biglobe.ne.jp/~k-uda/INDEX-S1.htm
The simple direction for uses.
1.Select the SDS surface.
2.Click "Create curve" button.
3.Select the nurbs curve in the SelectWindow.
4.Click "Bend curve" button.
Best regards,
K-UDA

Re: Re: Zbrush3+Realsoft3D .... and an alternative!

[Timo Mikkolainen]

It shouldn't be too hard; the only problem I can think of is that the
normal maps' normal might be created in relation to the smoothed
normal and I don't know how good/accessible that is in RS.

On 05/06/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:

Hi Timo and all,

>>From: Timo Mikkolainen <[EMAIL PROTECTED]>

>>You'd be better off using 16 or 32 bit >>bump/displacement maps: they
>>don't need a shifty material to be used in RS and, >>as the name says,they 
can be used for displacement.


Does RS support normal maps at all? You mention a "shifty material." I've seen 
the rgb 'normal' images RS creates during the rendering process (gi etc) and it seems 
like it wouldn't take too much of hack for RS to support them. Thots?

Thanks,

cm

Re: Zbrush3+Realsoft3D .... and an alternative!

[Timo Mikkolainen]

You'd be better off using 16 or 32 bit bump/displacement maps: they
don't need a shifty material to be used in RS and, as the name says,
they can be used for displacement. I just don't know if the mentioned
programs can create displacement maps.
For using highpoly models in RS,  import them as trisets. Trisets
render considerably faster than sds poly models and are lighter all
around. As for the difference between bump and normal maps, a bump map
has actual height information in it, while a normal map only has the
generated normal. So in essence, you can create a normal map from a
bump map, but you can't create a bump map from a normal map.

On 05/06/07, Aidan O Driscoll <[EMAIL PROTECTED]> wrote:

Hi Daniel,

Thanks for the reply. Yes - later - after sending this mail and some
discussion at the IRC - the conclusion arrived at is that Normal
Maps, as you said, are glorified bump maps, but the difference is
Bump maps use greyscale to generate the bump levels while Normal maps
use RGB. I must try and investigate the XNORMAL app in between
Blender/Silo and Realsoft. What Xnormal seems to do is create one of
these Normal maps by taking two obj versions of the mesh. A low poly
version and the high poly version. From these it creates a normal
map. As to how this works or understanding same - me dont understand yet :)

So again - why is this all of interest? The above is used
considerably in the gaming industry it seems. AND for RS at the
moment which [ I think ] has problems shifting large poly counts, its
a way of getting sculpted models into an RS scene with low overhead.
AND for free as both Xnormal and Blender ARE Free.

Cheers
Aidan



At 05:37 05/06/2007, you wrote:

>I'm no guru on this, but a normal map is a texture that defines a
>surface normal via color.
>A surface normal is the vector (a three dimensional direction) that
>points away from a surface. In rendering you use this vector to
>evaluate how much light any given area on an object reflects, and in
>which direction. By mapping a normal map to a surface you can
>artificially influence these normal vectors, and thus give an object a
>lot more detail in the rendering phase, that is not really there in
>geometry. Thus it renders fairly quickly, too. I may be wrong on this,
>but to me a normal map is a shortcut from a displacement map, since
>the later must be evaluated to generate normals, while the first
>already defines them by itself...
>hope that helped explain this technology some more.
>
>Daniel
>
>On 6/4/07, Aidan O Driscoll <[EMAIL PROTECTED]> wrote:
>>AND - Just came across this:
>>
>>http://sv3.3dbuzz.com/vbforum/showthread.php?s=&threadid=77037
>>
>
>
>
>--
>No virus found in this incoming message.
>Checked by AVG Free Edition. Version: 7.5.472 / Virus Database:
>269.8.9/832 - Release Date: 04/06/2007 18:43

Re: Second Life Issue: Having a RGB-output representing UV-space´s XYZ

[Timo Mikkolainen]

XYZ in what coordinate system?
If it's global coordinates, just make a material with
surface:color=surface:coordinates
, apply it to the model with default mapping and use map2obj to
extract/bake the color channel. Alternatively you can use parallel
mapping to define the coordinate system and use
surface:color=surface:map coords in the material.

On 25/05/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:

Lol, as soon as I found out about the upgrade of SL I thought about
this myself. Unfortunately, I never had time to really think before I
was, once again, lost inside. However, I do think that this shouldn't
be too difficult and I will try something when I manage to re-surface.
If anyone else is willing to make an attempt - I would be happy. :)

Regards,
Fredrik Bergholtz

On 24/05/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> Hello,
>
> I have a question. Can anybody tell me an easy way to convert the uv-space
> of a mesh into an TGA-output in which RGB represents XYZ of the mesh?
>
> Thank you for your answer!
> Albert
>
> --
> GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
> Alle Infos und kostenlose Anmeldung: http://www.gmx.net/de/go/freemail
>

Re: Array handling in JS

[Timo Mikkolainen]

I'd assume they read this list :)

On 28/03/07, K-UDA <[EMAIL PROTECTED]> wrote:

Hello, Timo.
Did you ask Realsoft about it ?
Best regards,
K-UDA

- Original Message -----
From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
To: 
Sent: Friday, March 23, 2007 3:10 PM
Subject: Re: Array handling in JS


> Nope, no progress at all.
>
> On 22/03/07, Juha Sunnari <[EMAIL PROTECTED]> wrote:
>> Did anyone solve this, or is it a bug? I tried to change the function
>> definition of for example GetInputTypes to:
>>
>> function GetR3MATPRPA_InputTypes() {
>>   return R3Get(this.r3obj, R3MATPRPA_InputTypes, R3TID_INTEGER,
>> R3TNF_ARRAY | R3TNF_READONLY | R3TNF_BYVALUE);
>> }
>>
>> and a couple of others by changing the combination of those flags, but
>> didn't get it to work.
>>
>> Timo Mikkolainen wrote:
>> > In the linux version everything gets printed to stdout (commandline),
>> > so I see everything properly. The actual popup is just an annoyance,
>> > if anything I'd like to get rid of it.
>> > When you look at the printed arrays, they look like pointers (as in,
>> > their values are close to eachother and changing the vsl object's
>> > input doesn't change the number).
>> >
>> > On 12/02/07, K-UDA <[EMAIL PROTECTED]> wrote:
>> >> Hello, Timo.
>> >> You are looking at the result of "line 554 print("");".
>> >> Please watch "realsoft\logs\Realsoft 3D.log".
>> >> Best regards,
>> >>
>> >> K-UDA
>> >> - Original Message -
>> >> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> >> To: 
>> >> Sent: Monday, February 12, 2007 9:42 AM
>> >> Subject: Re: Array handling in JS
>> >>
>> >>
>> >> > http://koti.welho.com/tmikkola/scripts/testMat.js
>> >> >
>> >> > See printObjectProperties(vslObject)
>> >> >
>> >> > On 11/02/07, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
>> >> >> Hi Timo,
>> >> >>
>> >> >> is it possible to post a snippet?
>> >> >>
>> >> >> Matthias
>> >> >>
>> >> >>
>> >> >> - Original Message -
>> >> >> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> >> >> To: 
>> >> >> Sent: Saturday, February 10, 2007 12:47 PM
>> >> >> Subject: Array handling in JS
>> >> >>
>> >> >>
>> >> >> > I'm trying to read the input channels of a vsl object in JS, but
>> >> >> > I
>> >> >> > don't know how to read the 'arrays' the JS functions return. I've
>> >> >> > tried foo[0], foo.length, foo.head and whatnot; nothing works.
>> >> >> > The
>> >> >> > arrays are returned by GetInputTypes(), GetInputChannels(), etc
>> >> >> > in
>> >> >> > r3matprp.js.

Re: Array handling in JS

[Timo Mikkolainen]

Nope, no progress at all.

On 22/03/07, Juha Sunnari <[EMAIL PROTECTED]> wrote:

Did anyone solve this, or is it a bug? I tried to change the function
definition of for example GetInputTypes to:

function GetR3MATPRPA_InputTypes() {
  return R3Get(this.r3obj, R3MATPRPA_InputTypes, R3TID_INTEGER,
R3TNF_ARRAY | R3TNF_READONLY | R3TNF_BYVALUE);
}

and a couple of others by changing the combination of those flags, but
didn't get it to work.

Timo Mikkolainen wrote:
> In the linux version everything gets printed to stdout (commandline),
> so I see everything properly. The actual popup is just an annoyance,
> if anything I'd like to get rid of it.
> When you look at the printed arrays, they look like pointers (as in,
> their values are close to eachother and changing the vsl object's
> input doesn't change the number).
>
> On 12/02/07, K-UDA <[EMAIL PROTECTED]> wrote:
>> Hello, Timo.
>> You are looking at the result of "line 554 print("");".
>> Please watch "realsoft\logs\Realsoft 3D.log".
>> Best regards,
>>
>> K-UDA
>> - Original Message -
>> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> To: 
>> Sent: Monday, February 12, 2007 9:42 AM
>> Subject: Re: Array handling in JS
>>
>>
>> > http://koti.welho.com/tmikkola/scripts/testMat.js
>> >
>> > See printObjectProperties(vslObject)
>> >
>> > On 11/02/07, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
>> >> Hi Timo,
>> >>
>> >> is it possible to post a snippet?
>> >>
>> >> Matthias
>> >>
>> >>
>> >> - Original Message -
>> >> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> >> To: 
>> >> Sent: Saturday, February 10, 2007 12:47 PM
>> >> Subject: Array handling in JS
>> >>
>> >>
>> >> > I'm trying to read the input channels of a vsl object in JS, but I
>> >> > don't know how to read the 'arrays' the JS functions return. I've
>> >> > tried foo[0], foo.length, foo.head and whatnot; nothing works. The
>> >> > arrays are returned by GetInputTypes(), GetInputChannels(), etc in
>> >> > r3matprp.js.
>> >> >
>> >
>

Re: tip about fog, skydome and DOF

[Timo Mikkolainen]

Map fetches the map coordinates from the given world coordinates. It's
mostly useful if you change the evaluated point by overwriting
coordinates and still want to get correct map coordinates. The only
'real-life' example I can think of is the shadowfog material in
tutorprojects->rendering->smokyroom.r3d. It should have a Map
coords=map(coordinates) after Coordinates+=multiply(dpos,Ray) for it
to have correct map coordinates.

On 07/03/07, [EMAIL PROTECTED]
<[EMAIL PROTECTED]> wrote:



The map object seems to be one of the biggest mysteries it arrived.
Maybe someone could explain excactly what it does, what it "replaces", and
how to use it properly?

Regards
Karl


-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Mark Heuymans
Sent: Wednesday, March 07, 2007 4:16 PM
To: user-list@light.realsoft3d.com
Subject: Re: tip about fog, skydome and DOF


Hi Arjo,
I often use a squashed sphere as fog object, with a parallel fog mapping.
And instead of using a constant turbidity, I use a curve with Mapcoords.Z as
input, resulting in a gradient atmosphere: turbidity is decreasing with
height. I'll have a look at it again, will upload a sample project later...
and how to implement the mysterious Map object?? (quote "Map object computes
the mapping coordinate projection from the given parameter." - yeah, that
explains it all ;)

-Mark H


Yes I prefer this type of fog too instead of the post one.
From your answer I understand you use a sphere object for the fog?
I use the atmosphere object (in the lights tab, the most right object).
That one doesn't have a size.


PS volumetric fog without volume shadows is fast, I prefer that to post
effect fog because reflections etc are correct.

 This e-mail and any attachment are confidential and may be privileged or
otherwise protected from disclosure. It is solely intended for the person(s)
named above. If you are not the intended recipient, any reading, use,
disclosure, copying or distribution of all or parts of this e-mail or
associated attachments is strictly prohibited. If you are not an intended
recipient, please notify the sender immediately by replying to this message
or by telephone and delete this e-mail and any attachments permanently from
your system.

Re: Possible particle bug

[Timo Mikkolainen]

Hmh, I quess the angle setting requires a reference direction (which
the surface provides).

On 18/02/07, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:

It does affect, if constructor is set to "distribute over surface".
See attached file.

Matthias

- Original Message -----
From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
To: 
Sent: Sunday, February 18, 2007 5:29 PM
Subject: Re: Possible particle bug


> Yep, it has no effect.
>
> On 18/02/07, Chris Sellars <[EMAIL PROTECTED]> wrote:
> > Hello Timo,
> >
> > Try the properties window for the particles and open the geometry tab.
> > Select angle, then try changing the value. When I do this, nothing happens.
> > Also try using vertex paint, to paint a new angle value onto the particles,
> > this also does not work for me.
> >
> > Thanks,
> >
> > Chris
> >
> >
> > - Original Message -
> > From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
> > To: 
> > Sent: Sunday, February 18, 2007 2:28 PM
> > Subject: Re: Possible particle bug
> >
> >
> > > Pointwise editing seems to work fine and I can rotate a particle (i.e.
> > > it shows rotation controls) by selecting both of its points.
> > >
> > > On 18/02/07, Chris Sellars <[EMAIL PROTECTED]> wrote:
> > > > Can anyone confirm a bug for me before I report it.
> > > > I used to be able to edit the angle of 2D particles (eg hairs), both
> > > > generally and pointwise.
> > > > Now nothing happens when I edit the angle values.
> > > > This is spoiling the hairstyles of my characters.
> > > >
> > > > Thanks,
> > > >
> > > > Chris Sellars
> > > > Morsel Animation
> > > >
> > > >
> > >
> > >
> > > --
> > > No virus found in this incoming message.
> > > Checked by AVG Free Edition.
> > > Version: 7.5.441 / Virus Database: 268.18.1/691 - Release Date: 17/02/2007
> > 17:06
> > >
> > >
> >
> >
>

Re: Possible particle bug

[Timo Mikkolainen]

Yep, it has no effect.

On 18/02/07, Chris Sellars <[EMAIL PROTECTED]> wrote:

Hello Timo,

Try the properties window for the particles and open the geometry tab.
Select angle, then try changing the value. When I do this, nothing happens.
Also try using vertex paint, to paint a new angle value onto the particles,
this also does not work for me.

Thanks,

Chris


- Original Message -
From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
To: 
Sent: Sunday, February 18, 2007 2:28 PM
Subject: Re: Possible particle bug


> Pointwise editing seems to work fine and I can rotate a particle (i.e.
> it shows rotation controls) by selecting both of its points.
>
> On 18/02/07, Chris Sellars <[EMAIL PROTECTED]> wrote:
> > Can anyone confirm a bug for me before I report it.
> > I used to be able to edit the angle of 2D particles (eg hairs), both
> > generally and pointwise.
> > Now nothing happens when I edit the angle values.
> > This is spoiling the hairstyles of my characters.
> >
> > Thanks,
> >
> > Chris Sellars
> > Morsel Animation
> >
> >
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.5.441 / Virus Database: 268.18.1/691 - Release Date: 17/02/2007
17:06
>
>

Re: Possible particle bug

[Timo Mikkolainen]

Pointwise editing seems to work fine and I can rotate a particle (i.e.
it shows rotation controls) by selecting both of its points.

On 18/02/07, Chris Sellars <[EMAIL PROTECTED]> wrote:

Can anyone confirm a bug for me before I report it.
I used to be able to edit the angle of 2D particles (eg hairs), both
generally and pointwise.
Now nothing happens when I edit the angle values.
This is spoiling the hairstyles of my characters.

Thanks,

Chris Sellars
Morsel Animation

Re: Array handling in JS

[Timo Mikkolainen]

Yep, I noticed this too... I think the value is just a pointer to the
array. If this is so, it's probably impossible to access the array in
js.

On 17/02/07, K-UDA <[EMAIL PROTECTED]> wrote:

Hello, Timo.
I changed an input channel and tested it.
The return value was the same.
I duplicated a material object and tested it.
The return value was different.
You had better ask to Realsoft about it.
I am sorry that I cannot help you.
Best regards,
K-UDA

- Original Message -----
From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
To: 
Sent: Monday, February 12, 2007 1:32 PM
Subject: Re: Array handling in JS


> In the linux version everything gets printed to stdout (commandline),
> so I see everything properly. The actual popup is just an annoyance,
> if anything I'd like to get rid of it.
> When you look at the printed arrays, they look like pointers (as in,
> their values are close to eachother and changing the vsl object's
> input doesn't change the number).
>
> On 12/02/07, K-UDA <[EMAIL PROTECTED]> wrote:
>> Hello, Timo.
>> You are looking at the result of "line 554 print("");".
>> Please watch "realsoft\logs\Realsoft 3D.log".
>> Best regards,
>>
>> K-UDA
>> - Original Message -
>> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> To: 
>> Sent: Monday, February 12, 2007 9:42 AM
>> Subject: Re: Array handling in JS
>>
>>
>> > http://koti.welho.com/tmikkola/scripts/testMat.js
>> >
>> > See printObjectProperties(vslObject)
>> >
>> > On 11/02/07, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
>> >> Hi Timo,
>> >>
>> >> is it possible to post a snippet?
>> >>
>> >> Matthias
>> >>
>> >>
>> >> - Original Message -
>> >> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> >> To: 
>> >> Sent: Saturday, February 10, 2007 12:47 PM
>> >> Subject: Array handling in JS
>> >>
>> >>
>> >> > I'm trying to read the input channels of a vsl object in JS, but I
>> >> > don't know how to read the 'arrays' the JS functions return. I've
>> >> > tried foo[0], foo.length, foo.head and whatnot; nothing works. The
>> >> > arrays are returned by GetInputTypes(), GetInputChannels(), etc in
>> >> > r3matprp.js.
>> >> >
>> >>
>> >
>> >
>>
>
>

Re: relocate face from SDS UV set

[Timo Mikkolainen]

The easiest way I found was to delete the face(s) from one UVset,
create a new UVset from that face(s) only and merge it with the other
UVset.

On 13/02/07, Arjo Rozendaal <[EMAIL PROTECTED]> wrote:

Hi,

Don't bother about this question anymore. Solved it in another way.
Something else went wrong with the UVsets, so I used facewise mapping
instead.

Arjo.

> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] Behalf Of Arjo Rozendaal
> Sent: dinsdag 13 februari 2007 12:05
> To: Reallist
> Subject: relocate face from SDS UV set
>
>
> Hi,
>
> I devided a sds object in different UV sets. I made a mistake. There's one
> face that I added to the wrong set. Is there a way to move a face from one
> set to another?
>
> Arjo.
>
>

Re: Array handling in JS

[Timo Mikkolainen]

In the linux version everything gets printed to stdout (commandline),
so I see everything properly. The actual popup is just an annoyance,
if anything I'd like to get rid of it.
When you look at the printed arrays, they look like pointers (as in,
their values are close to eachother and changing the vsl object's
input doesn't change the number).

On 12/02/07, K-UDA <[EMAIL PROTECTED]> wrote:

Hello, Timo.
You are looking at the result of "line 554 print("");".
Please watch "realsoft\logs\Realsoft 3D.log".
Best regards,

K-UDA
- Original Message -
From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
To: 
Sent: Monday, February 12, 2007 9:42 AM
Subject: Re: Array handling in JS


> http://koti.welho.com/tmikkola/scripts/testMat.js
>
> See printObjectProperties(vslObject)
>
> On 11/02/07, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
>> Hi Timo,
>>
>> is it possible to post a snippet?
>>
>> Matthias
>>
>>
>> - Original Message -
>> From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
>> To: 
>> Sent: Saturday, February 10, 2007 12:47 PM
>> Subject: Array handling in JS
>>
>>
>> > I'm trying to read the input channels of a vsl object in JS, but I
>> > don't know how to read the 'arrays' the JS functions return. I've
>> > tried foo[0], foo.length, foo.head and whatnot; nothing works. The
>> > arrays are returned by GetInputTypes(), GetInputChannels(), etc in
>> > r3matprp.js.
>> >
>>
>
>

Re: Array handling in JS

[Timo Mikkolainen]

http://koti.welho.com/tmikkola/scripts/testMat.js

See printObjectProperties(vslObject)

On 11/02/07, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:

Hi Timo,

is it possible to post a snippet?

Matthias


- Original Message -
From: "Timo Mikkolainen" <[EMAIL PROTECTED]>
To: 
Sent: Saturday, February 10, 2007 12:47 PM
Subject: Array handling in JS


> I'm trying to read the input channels of a vsl object in JS, but I
> don't know how to read the 'arrays' the JS functions return. I've
> tried foo[0], foo.length, foo.head and whatnot; nothing works. The
> arrays are returned by GetInputTypes(), GetInputChannels(), etc in
> r3matprp.js.
>

Array handling in JS

[Timo Mikkolainen]

I'm trying to read the input channels of a vsl object in JS, but I
don't know how to read the 'arrays' the JS functions return. I've
tried foo[0], foo.length, foo.head and whatnot; nothing works. The
arrays are returned by GetInputTypes(), GetInputChannels(), etc in
r3matprp.js.

Re: Real DOF hack

[Timo Mikkolainen]

On 07/02/07, [EMAIL PROTECTED]
<[EMAIL PROTECTED]> wrote:

Hi

Haven't checked your shader yet, but I'm considering installing the demo
at work just for that :-)

Realistic DOF is very interresting, and should be built into the system
as an optional method.


I agree.



However, it *MUST* be optional; realistic or quickfake, and not even the
default option. I even think the quickfake one should be made to better
handle reflections and/or when going through foggy materials if
possible. I've had limited success with quickfake DOF reflections (copy+
distance trick), and no success at all within foggy volumes. Don't think
highlights would ever be captured correctly in post, as usually only the
8bit image is available to postprocess.


It might be possible, to some extent, to copy the volume's distance to
a custom channel and then add it to the distance channel in post just
before DOF. I haven't tried this yet.



Reason that it should be optional, is naturally the extreme hit of doing
realistic DOF, with rendertimes multiplying quickly.


Funny enough, but if you use RS post dof with .5 post scaling, my DOF
is just as fast with simple scenes. I haven't tried it with complex
scenes, but I think the post DOF will get faster.



With my computer setup, I'm actually very happy of using postprocessed
DOF (not realsoft's implementation though, I consider that only a
preview).

As for CoC, isn't that pretty much preset depending on the
cameratype/sensorsize you're trying to simulate? My Nikon D80 dslr has a
CoC of 0.02, compact cameras would be around 0.002-0.005, and medium to
large format has larger than dslr CoCs. Wouldn't it be enough just to
set CoC as a constant? It doesn't vary with camaera and lens settings as
such afaik.


CoC might've been the wrong term... what I mean is the 'blur' amount.
This is somehow calculated from aperture size and distance from lens
to film.



Having started abit with photography recently, controlling DOF depending
on subject is of great importance. Having two lenses with 7 curved
blades, I now see the need to increase that to 9 blades. My 18-200VR
lens is one superb lens, but draws horrible bokeh in my view. So now I
need a good portrait lens just to get nice bokeh (how out of focus
objects are "rendered" by the lens) effect :-)


Note that you can control bokeh/blur form with a texture =) it's a bit
slow, but the effect is nice. You can even use colored textures,
although the total 'contribution' is normalized to 1,1,1.



DOF (fake or realistic looking) is one of CGI's biggest faults, but I
didn't realize it until I started shooting myself. In real life,
*getting* all in focus can be a bitch, in CGI it's the default setup and
cost less rendertime wise. I see little use of DOF in the competision
images.

I recently did some tests trying to create a quickefake version of a
Tilt lens. Easily done in RS3D. Instead of the usual method of making a
separate channel measure distance from camera to subject, I mapped it
using a default parallel map in some odd angle instead of perpendicular
to the camera (controlling the focus plane). Results were pretty odd :-)
Looking into a field of flowers, I could have the top of all flowers
into focus, gradually falling out of focus toward the ground.


Sounds like fun, I should try that =)



Just some real world examples of this:
http://www.nikonians.org/html/resources/nikon_articles/nikkor/special_pu
rpose/85_PC/falguera.html
http://www.naturfotograf.com/28pc.html
And the most obvious effect of them all :-)
http://users.gsinet.net/pjwhite/35tstest3f8full.jpg

Have you tried your shader doing tilt lens work?

I'm not quite sure how this should be implemented... would it suffice
if I added a 'blur' channel to the plane ? This way you could change
the blur amount locally.


Just my 2-300 cents ;-)

Regards
Karl

Re: Real DOF hack

[Timo Mikkolainen]

From http://www.blender.org/documentation/htmlI/x12981.html


The Depth of Field filter is a 2D filter, i.e. a post processing
technique, and as such, has advantages and disadvantages. It uses the
rendered image, plus a Z Buffer (which tells the filter how far away
each pixel is from the camera) to figure out which pixels are blurred
or not blurred.

Because its a 2D effect it has the advantage of being extremely quick.
However there are a few disadvantages:

Reflections are not blurred correctly. If you look at a reflection,
you'll notice that the reflection's blur is based on the distance from
the camera of the reflection plane, not the object in the reflection.

Because the DOF is done on a 2D image, rather than a 3D scene, the
blur cannot know what is behind any given object, therefore often the
edges of an extremely blurred object in the foreground will look
smudgy or dirty.

On 06/02/07, Arjo Rozendaal <[EMAIL PROTECTED]> wrote:

Hi Timo,

What does this material do exactly? I mean what's wrong with the regular
DOF?

Arjo.

> -Original Message-
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] Behalf Of Timo Mikkolainen
> Sent: dinsdag 6 februari 2007 16:27
> To: user-list@light.realsoft3d.com
> Subject: Real DOF hack
>
>
> I posted this in the forum a while ago, but I thought I'd post it here
> as well: I made a little material hack that does real DOF:
> http://koti.welho.com/tmikkola/materials/DOF.zip
> http://koti.welho.com/tmikkola/materials/DOF.txt
>
> Read the readme for short instructions.
> Naturally it won't work with post-blurred GI as it doesn't even copy
> depth to the image...
> I'd love to make a script that adds and choreographs the material to a
> user selected camera, but I have no idea how to calculate the CoC
> values from the camera properties.
>

Real DOF hack

[Timo Mikkolainen]

I posted this in the forum a while ago, but I thought I'd post it here
as well: I made a little material hack that does real DOF:
http://koti.welho.com/tmikkola/materials/DOF.zip
http://koti.welho.com/tmikkola/materials/DOF.txt

Read the readme for short instructions.
Naturally it won't work with post-blurred GI as it doesn't even copy
depth to the image...
I'd love to make a script that adds and choreographs the material to a
user selected camera, but I have no idea how to calculate the CoC
values from the camera properties.

Re: fakeShadows

[Timo Mikkolainen]

Ok, should work now: http://koti.welho.com/tmikkola/materials/fakeShadows.zip

On 31/01/07, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:

Looks like there's a small bug, the shader doesn't flip the normal
towards the camera ray before it checks if it's facing the light ray.
I'll fix it when I get home.

On 31/01/07, Frank Bueters <[EMAIL PROTECTED]> wrote:
>
> Hi Timo,
>
> Either I'm confused or there's more to it than just add one shader to the
> scene and the other to the light source. Can't get it to produce
> (soft)shadows. I've attached a basic setup. Would you mind to have a quick
> glance at it and tell me what I do wrong?
>
> Regards,
>
> Frank Bueters
>
>
> > There's two shaders in the project, add the other one to the
> > scene (or all parts you want to receive shadows) and the
> > other one to the light you want to cast the shadows. You can
> > also turn shadows off in the light properties. I'll add a
> > readme once I have the time.
> >
> > On 31/01/07, Frank Bueters <[EMAIL PROTECTED]> wrote:
> > > Hello real users,
> > >
> > > Does anyone know how to put Timo's fakeShadows to work? Or
> > is it being
> > > explained somewhere? Tried to add it to the light source,
> > to objects,
> > > to the root... nothing happens.
> > >
> > > Regards,
> > >
> > > Frank Bueters
> > >
> > >
> > >
> > >
> > >
>
>

Re: fakeShadows

[Timo Mikkolainen]

Looks like there's a small bug, the shader doesn't flip the normal
towards the camera ray before it checks if it's facing the light ray.
I'll fix it when I get home.

On 31/01/07, Frank Bueters <[EMAIL PROTECTED]> wrote:


Hi Timo,

Either I'm confused or there's more to it than just add one shader to the
scene and the other to the light source. Can't get it to produce
(soft)shadows. I've attached a basic setup. Would you mind to have a quick
glance at it and tell me what I do wrong?

Regards,

Frank Bueters


> There's two shaders in the project, add the other one to the
> scene (or all parts you want to receive shadows) and the
> other one to the light you want to cast the shadows. You can
> also turn shadows off in the light properties. I'll add a
> readme once I have the time.
>
> On 31/01/07, Frank Bueters <[EMAIL PROTECTED]> wrote:
> > Hello real users,
> >
> > Does anyone know how to put Timo's fakeShadows to work? Or
> is it being
> > explained somewhere? Tried to add it to the light source,
> to objects,
> > to the root... nothing happens.
> >
> > Regards,
> >
> > Frank Bueters
> >
> >
> >
> >
> >

Re: fakeShadows

[Timo Mikkolainen]

There's two shaders in the project, add the other one to the scene (or
all parts you want to receive shadows) and the other one to the light
you want to cast the shadows. You can also turn shadows off in the
light properties. I'll add a readme once I have the time.

On 31/01/07, Frank Bueters <[EMAIL PROTECTED]> wrote:

Hello real users,

Does anyone know how to put Timo's fakeShadows to work? Or is it being
explained somewhere? Tried to add it to the light source, to objects, to the
root... nothing happens.

Regards,

Frank Bueters

Re: UV and geometry unwrapping

[Timo Mikkolainen]

I haven't found a proper way to do this. I really wish RS would
implement a proper unwrapping system where you can just define the UV
seams and press 'Unwrap'.

On 03/01/07, Chris Sellars <[EMAIL PROTECTED]> wrote:

Happy New Year to everyone,

Does anyone know of a facility on Realsoft to Unwrap a UV Set straight to a
2D flat plane. The existing tool set seems to do a very gradual unwrap, but
this takes forever if you wish to flatten a UV set completely.
(I am trying to unwrap the UV sets of various areas of a character
accurately, so that all faces remain undistorted when compared to the actual
SDS, but are layed flat onto a 2D plane for texturing).

Any help much appreciated.

Thanks,

Chris Sellars
Morsel Animation

Re: My computer took a holiday...

[Timo Mikkolainen]

Meh, bought myself a new PSU as an early present and I'm up and running again :)

On 22/12/06, studio <[EMAIL PROTECTED]> wrote:

> My computer made a nice 'pop' sound and stopped working (surprise).
> Apparently the psu is busted... anyways, I'll be pretty much out of
> the action for the rest of the year.

Sorry To Hear , But ... :

  What Action ?

  Cheers anyway & hope you find time to enjoy the
Christmas holidays in other ways .

studio

My computer took a holiday...

[Timo Mikkolainen]

My computer made a nice 'pop' sound and stopped working (surprise).
Apparently the psu is busted... anyways, I'll be pretty much out of
the action for the rest of the year.

Re: SV: An SSS [ Sub Surface Scattering ] problems

[Timo Mikkolainen]

Yerp, using the GI post effect/blur would be nice, but it just has too
many problems. The final version will have its own post effect, which
also makes it possible to texture the SSS effect. I'll try to finish
it this week: I'll put it on my site and mail it to the list then.

On 11/12/06, Aidan O Driscoll <[EMAIL PROTECTED]> wrote:

Hi Guys,

Uppz, TIMO answered so disregard my last mail.

Cheers Timo, seems this shader is of good interest, great stuff

Thanks again
Aidan

At 13:43 11/12/2006, you wrote:
>Hi Vesa,
>
>When I posted the SSSish Shader [ on behalf of TIMO ] I did not mention
>this - "don't set the "subsurface color" to 1,1,1."
>
>I seem to remember TIMO mentioning this, but in passing, so forgot same.
>
>Would it be possible for you to edit the SSSish Shader with any changes
>you mention OR any 'upgrades' you see useful [ With plain english headers
>on the sliders and tooltips if possible :) ]
>
>I reckon Timo wont mind,
>
>Cheers
>Aidan
>
>At 08:28 11/12/2006, you wrote:
>>"urn:schemas-microsoft-com:vml" xmlns:o =
>>"urn:schemas-microsoft-com:office:office" xmlns:w =
>>"urn:schemas-microsoft-com:office:word" xmlns:st1 =
>>"urn:schemas-microsoft-com:office:smarttags">
>>Hello,
>>
>> >I think I have found the source of the strange colouring, don't set the
>> "subsurface color" to 1,1,1.
>>
>>I tried the example you sent and I can confirm that the face of the
>>snowman got unexpected bluish tone.
>>
>>It appears that GI shader and SSS shader are not totally compatible. ´The
>>SSS shader computes a diffuse channel as follows:
>>
>> if(unshaded)
>> sbuf = raytracer(..)
>> sbuf /= Copy(color)
>> diffuse += multiply(multiplier, color)
>>
>>When you apply this computation to the orange 'carrot nose' color, you
>>will get a cyan tone. While it is probably OK for the SSS shader, it
>>colors the GI in a wrong way. I would expect that if you create new
>>channels that are private to SSS and change the channel references from
>>the SSS shader to the new channels (and change the post effects
>>respectively, if SSS needs them), the problem disappears.
>>
>>I hope this helps!
>>
>>
>>Kind regards,
>>
>>Vesa
>>No virus found in this incoming message.
>>Checked by AVG Free Edition.
>>Version: 7.5.432 / Virus Database: 268.15.15/581 - Release Date:
>>09/12/2006 15:41
>No virus found in this incoming message.
>Checked by AVG Free Edition.
>Version: 7.5.432 / Virus Database: 268.15.15/581 - Release Date:
>09/12/2006 15:41

Re: SV: An SSS [ Sub Surface Scattering ] problems

[Timo Mikkolainen]

The tie-in with the GI post effect and diffuse channel are only
temporary, I'll separate SSS from GI in the final version.

On 11/12/06, Vesa Meskanen <[EMAIL PROTECTED]> wrote:



Hello,

>I think I have found the source of the strange colouring, don't set the
"subsurface color" to 1,1,1.

I tried the example you sent and I can confirm that the face of the snowman
got unexpected bluish tone.

It appears that GI shader and SSS shader are not totally compatible. ´The
SSS shader computes a diffuse channel as follows:

if(unshaded)
sbuf = raytracer(..)
sbuf /= Copy(color)
diffuse += multiply(multiplier, color)

When you apply this computation to the orange 'carrot nose' color, you will
get a cyan tone. While it is probably OK for the SSS shader, it colors the
GI in a wrong way. I would expect that if you create new channels that are
private to SSS and change the channel references from the SSS shader to the
new channels (and change the post effects respectively, if SSS needs them),
the problem disappears.

I hope this helps!


Kind regards,

Vesa

Re: RS with SP3 crashing when closed after session ... anyone else?

[Timo Mikkolainen]

I have a similar problem on linux, RS gets stuck with an Xlib:
unexpected async reply error.

On 27/11/06, Aidan O Driscoll <[EMAIL PROTECTED]> wrote:

Hi Guys,

Anyone else having this problem -

Since installing SP3 - after a session with RS and when closing the app, I
get the WINDOWS HAS ENCOUNTERED blah blah - error box. This happens pretty
much every time. We were talking about this at the IRC and others have had
the same issue.

Any ideas?

Cheers
Aidan

Re: SSS test subject

[Timo Mikkolainen]

On 23/11/06, Mark Heuymans <[EMAIL PROTECTED]> wrote:

At 18:26 23-11-2006, you wrote:
>All procedurals. It's just the std wrinkle material, some cell bumps
>and a simple noise+if for the veins. I used vertex channels for
>controlling the bumps and SSS color.


Works nicely. Procedurals can be great: no tiling, no lost
textures... it just takes some time to come up with something
interesting  It's amazing what's possible by combining noise and
cells, I've spent whole evenings playing with these (and cellular,
scratches and bricks of course). Any plans to animate it or place it
in its natural habitat?


Natural habitat maybe, but I'd rather not animate it: the render times
are too long.




>We came to the conlusion that these things live in shallow water and
>grab on to your gonads when you swim there. The purple things have
>strong neurotoxins and acids in them, so basically it immobilizes (or
>kills) you and then melts and eats your nether regions.


Interesting species, a perfect present for the neighbours!

-Mark H

Animation as 3D texture

[Timo Mikkolainen]

I'm trying to use an animation as a 3d texture, but I can't think of a
way to use map coords' z axis as an input instead of time. I even
tried to choreograph the current frame index with map z as input, but
it didn't work. Any ideas?

Re: SSS test subject

[Timo Mikkolainen]

All procedurals. It's just the std wrinkle material, some cell bumps
and a simple noise+if for the veins. I used vertex channels for
controlling the bumps and SSS color.
We came to the conlusion that these things live in shallow water and
grab on to your gonads when you swim there. The purple things have
strong neurotoxins and acids in them, so basically it immobilizes (or
kills) you and then melts and eats your nether regions.

On 23/11/06, Mark Heuymans <[EMAIL PROTECTED]> wrote:

At 02:57 22-11-2006, you wrote:
>As I don't like animal testing, I decided to test my SSS shader on a
>plant...ish thing:
>http://koti.welho.com/tmikkola/Gallery/grabby2.jpg
>GI, SSS, DOF, Soft light, etc. Took forever to render, added noise and
>lens distortion in Gimp.


Great work Timo, you pushed the renderer to its limits!
I'm wondering if the texture is procedural or bitmap?

And: where can I get one, (s)he looks so cute, and what do you feed them? ;)

thanks for showing,
-Mark

SSS test subject

[Timo Mikkolainen]

As I don't like animal testing, I decided to test my SSS shader on a
plant...ish thing:
http://koti.welho.com/tmikkola/Gallery/grabby2.jpg
GI, SSS, DOF, Soft light, etc. Took forever to render, added noise and
lens distortion in Gimp.

Re: Ambient occlusion updated.

[Timo Mikkolainen]

On 13/11/06, Mark Heuymans <[EMAIL PROTECTED]> wrote:


>>I like the AO effect, except for the render times. Why not go for
>>full GI if it's only a little bit longer to render?
>This will be fixed (atleast to some extent) in V6: the RayTracer VSL
>object can evaluate only the ray distance, which should make AO quite
>a bit faster.


That's nice to hear. I want V6!
Is it possible to implement some kind of adaptive raycount, i.e.
before starting the real tracing, the Raytracer object looks around a
bit for close objects or close surface curvatures and determines
raycount from there. For this first step, no materials would have to
be evaluated, just geometry. Basically, what I mean is: use high
raycount/low blur where it's needed near details and corners, low
raycount/high blur in the middle of large areas. The Blur effect
would have to be adaptive too - difficult...


Some sort of adaptive sampling would be nice, but if I remember
correctly, the RayTracer object already does this with higher
RayCounts. This doesn't affect the blur kernel size though. What
you're describing is pretty close to how photon mapping works, which
would be nice to have in RS.


tnx,
Mark H

Re: Ambient occlusion updated.

[Timo Mikkolainen]

On 12/11/06, Mark Heuymans <[EMAIL PROTECTED]> wrote:


> >  Mainly , I noticed that your "AO_Effects" Post-Pro-
> > cessing Effect seems to have no effect . Is this a
> > small oversite , or am I just missing something .
>
>This is the original posteffect created by Timo. It's there when you want to
>use the original shaders, also created by Timo. Should have written that in
>the readme. Makes me think, that it might be wiser to save my contributions
>on my own webspace and create a link in the wiki. As I can't change the
>files as they are on the Realsoft info site.
>
>Arjo.


Hi Arjo,
Thanks for the efforts and the sharing!
I like the AO effect, except for the render times. Why not go for
full GI if it's only a little bit longer to render?

This will be fixed (atleast to some extent) in V6: the RayTracer VSL
object can evaluate only the ray distance, which should make AO quite
a bit faster.


The effect of a single large special light that surrounds the scene
(a hollow cylinder without bottom cap works well) can mimic the AO
effect pretty well but again, render times can pose a problem if you
set lightsource quality high enough to get rid of the noise. But that
way you avoid glass and fog problems.

Aaah! I tried your project on a single monitor setup and couldn't get
to the Properties window; the pre-SP3 solution didn't work any more
(delete some files in Documents and Settings\). Can someone
refresh my memory on how to solve this without hassling with monitors?

tnx
-Mark H

Re: Ambient occlusion updated.

[Timo Mikkolainen]

I'll try it once I get home. I already have new version of AO pending;
I'll release it once I finish my new site layout.

On 09/11/06, Arjo Rozendaal <[EMAIL PROTECTED]> wrote:

Hi,

I made a little change in Timo's ambient occlusion shader. Now you can
combine ambient occlusion with normal shading.
I attached the shader with a little readme to show what it does.
Maybe Vesa can put it on the Realsoft info site?

Arjo.

Re: displacement animation

[Timo Mikkolainen]

On 08/11/06, Mark Heuymans <[EMAIL PROTECTED]> wrote:


>Incredible nice shader and lighting.
>The animation feel like I was on medication overdose.
>Where did you get that beautiful displacement map?
>
>Jean-Sebastien Perron
>www.neuroworld.ws

Thanks; I used World Machine for primary output and then converted to
HDR with HDRShop to get the map. This is much better than an 8-bit
texture like bmp because of higher precision and reduced banding.
There's still some banding visible but it's not really distracting.


Have you tried using Gradient X and Gradient Y in the texture VSL object?
This might remove the banding.


The limit for World Machine is 8192x8192 so that'll be the next step
(plus extra RAM) ;)

Vesa: does the Surface normal apply to a displaced surface? In other
words, do slope-dependent materials work?


-Mark H

Re: displacement animation

[Timo Mikkolainen]

On 08/11/06, Mark Heuymans <[EMAIL PROTECTED]> wrote:

 >Hello Mark
 >
 >> Vesa: does the Surface normal apply to a displaced surface? In other
 >> words, do slope-dependent materials work?
 >
 >Yes, they do.  Slope related features should be computed in the Surface
 >Propertes shader using Bump Normal, not Surface Normal.
 >
 >Thanks for the beautiful animation! I hope you do not mind if I put  it into
 >our web gallery?
 >
 >
 >Best regards,
 >
 >Vesa


Of course, I'd be honoured!

Next, I'll add a flowmap (a simulated mapping of deposits, output of
World Machine), and maybe get rid of the camera motion.
Too bad I can't animate the actual erosion process: that would
require a different heightmap for every frame, and automatic
processing by HDRShop... makes me think of the Arexx days.


You might be able to batch process the format change with pfstools:
http://www.mpi-inf.mpg.de/resources/pfstools/



Jean-Sebastien: now that I think of it, no wonder it felt to you like
a medication overdose; I've been working on hospital animations for
two months. Apparently I can't get it out of my system :)

Here's an idea that mimics Vue's Ecosystem: is it possible (or easy
to implement) to paste objects on a surface according to a VSL
channel? A kind of Scope, but scope of objects' distribution (or 3d
particles). Paint hairs or trees by VSL! Can it be done?


tnx,
Mark H

Re: displacement animation/distribute objects

[Timo Mikkolainen]

On 09/11/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:

Hi Jean-Sebastien,

> When they released Vue5, they never offer upgrade prices for Vue4 users.

that's not true. E-on has send me an e-mail with an upgrade  ;-)

> But in the end it is easy to replicate by manual copy/placing :
> -copy a couples of object and place them.
> -Group all the previous objects
> -copy and place the previous group 5 times
> -Group everything again
> -Copy the new group and place it many times
> -etc...
> -I swear you eyes will never see the difference.

Did you recognize the distribute-tool in RS
(see attached picture)
I know that it will not with rendertime displacement.


Maybe we could use a companion (or extension) for the NURBS displace
tool that displaces an object (or its points) based on a channel
value, not just a bitmap.



Matthias

Re: Service Pack 3

[Timo Mikkolainen]

Automated texture generation would be great, including shadow maps
from raytraced shadows: this'd make soft shadows more usable with GI.
And once we go that way, why not do irradiance caching as well ;)

On 02/11/06, David Coombes <[EMAIL PROTECTED]> wrote:

> I just finished a huge architectural project, rendered 8000 frames at
> PAL resolution (on 4 computers), with pretty high AA settings. It's
> just a matter of getting a little help in CPU power... as long as
> render times stay below 10 minutes per frame it's manageable for
production.
> It's even possible to throw in some slow VSL shaders, or the
> occasional area lightsource. Know how to optimize your scenes! Just
> don't combine notorious CPU-hungry features like GI, volumetric
> shadows, special lights.
> The one thing I really fear is GI animation; the tutorial in the
> manual (with temporal sampling) is really intimidating... I usually
> bake an illumination map into the ceilings, this works very well.

For performance enhancements, I think there's a lot of tricks out there. eg.
Volumetric smoke is a pig to raytrace, but there's been some work on
realtime volumetric smoke for games and it looks fantastic. Whatever sort of
a hack it is, it would be a useful addition. I'm thinking perhaps the engine
could remain the default raytracer, but with various plugin render modes
that are drawn and composited in as needed. eg. You could have some scenery
perhaps with a characters, and select the scenery and set it's render mode
to Scanline. This would then be rendered on a GPU perhaps, and the
characters raytraced into the scene.

For animation, it might be possible to have the software perform some smart
or automated precalculations. For something like VSL noise, a texture map is
far, far faster. So how's about per object or per material txture creation,
where a texture is created at a specified resolution and that's used on the
object? At the moment these sorts of optimisations need to be handled
manually, rendering out textures with the associated problems that's had.
Adding 10 minutes at the start of render to create textures and work from
those would speed things up dramatically. Similarly for GI, precalculated
Spherical Harmonics could work in a lot of cases to provide fast and useable
GI after an initial calculation phase to set it all up.

David Coombes
[EMAIL PROTECTED]

...
- Original Message -
From: "Mark Heuymans" <[EMAIL PROTECTED]>
To: 
Sent: Thursday, November 02, 2006 4:34 PM
Subject: Re: Service Pack 3


>
> >>And thanks for charging nothing again - RS still offers most bang
> >>for your buck ;)
> >>
> >>
> >>-Mark Heuymans
> >>
> >Normally a ServicePack or an update is only to solve bugs.
> >It's really nice to them to add new features for free.
> >Thankyou very much.
> >
> >Realsoft has alway been the most powerful software.
> >The price is high, but for what you get... it's worth it.
> >
> >Jean-Sebastien Perron
> >www.neuroworld.ws
>
>
>
> Absolutely, especially because with a single license you can set up
> an unlimited render farm. No $$$ per CPU :)
>
> I mailed to this list a few times, but it seems that they don't
> arrive if I mail from my new computer through a shared internet
connection...
>
> About the raytracing discussion: yes, as always there are pros and
> cons but I've always liked it and will to continue to like it! With
> today's fast dual-core computers, raytracing is as useful as ever. If
> I look back at my Amiga 32MHz days, I'm in heaven now. That doesn't
> mean I wouldn't welcome some GPU help of course.
> I just finished a huge architectural project, rendered 8000 frames at
> PAL resolution (on 4 computers), with pretty high AA settings. It's
> just a matter of getting a little help in CPU power... as long as
> render times stay below 10 minutes per frame it's manageable for
production.
> It's even possible to throw in some slow VSL shaders, or the
> occasional area lightsource. Know how to optimize your scenes! Just
> don't combine notorious CPU-hungry features like GI, volumetric
> shadows, special lights.
> The one thing I really fear is GI animation; the tutorial in the
> manual (with temporal sampling) is really intimidating... I usually
> bake an illumination map into the ceilings, this works very well.
>
>
> Happy rendering,
> Mark H
>
>
>
>

Re: I think it's time for Realsoft to go from Raytracing to Scanline rendering with 3D videocards and or consoles

[Timo Mikkolainen]

Implementing scanline rendering in RS would be extremely hard, as most
of the geometry in RS is not polygon based. You also have to
understand that stuff like GI in realtime engines is made possible by
lengthy preprocessing like spherical harmonics, normal maps, etc.
Using the GPU for 'general' FP calculations would be possible and
probably even advantageous, but it's a pain to implement it in a
general enough way.
If you really want to have scanline rendering in RS, give a nice chunk
of money to Mirco so he'll continue work on his RM4RS3D thingie.
PS3 would make a nice renderslave though (it even runs Linux).

On 01/11/06, Jean-Sebastien Perron <[EMAIL PROTECTED]> wrote:

With PS3 and Xbox 360 doing fully lighted scene with shaders and
millions of polygons at 1920*1080 60fps

this is just for example :
http://www.gamespot.com/ps3/action/insomniacshooter/screenindex.html
http://www.gamespot.com/xbox360/action/bioshock/screenindex.html

Is it time for Realsoft to buy a PS3 or 360 devkit and start working on
a 3D application dedicated to those wonderful hardware?
They both have Hard Disk and internet connection

I may be wrong, but I think ScanlineRendering technique have won over
Raytracing in Quality now? or just equal.
ScanlineRendering now is 1 000 000 times faster than raytracing.
Raytracing is difficult to implement to a hardware pipeline.

PS3 and 360 are the new Amiga.
PC is too slow behind and if more powerfull it's still 900% more expensive.

Consoles even at 600$ are at the price level of a regular PC 3D
videocard alone.

This is valid for Realsoft as for any other 3D application.

Jean-Sebastien Perron
www.neuroworld.ws

Re: The last tutorial on NeuroWorld

[Timo Mikkolainen]

Wiki allows quite a lot of html markup, links are the only thing you
must wiki-fy.

On 20/10/06, Daniel Richter <[EMAIL PROTECTED]> wrote:

What I never undestood about the wiki is why it doesn't just allow
people to insert HTML.. which many poeple know really well... yet
another language to learn and only for a very narrow focus..

my 2 cents.

Daniel


On 10/20/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Hi Jean-Sebastien,
>
> to:
> A last question, How do I put my tutorials links on RSWiki?
>
> First create an account:
>
> http://rs3dwiki.the-final.info/index.php?title=Special:Userlogin
>
> Go to the desired Page, check first the sandbox to play:
>
> http://rs3dwiki.the-final.info/index.php?title=Sandbox
>
> use there the "edit" links at the articles.
>
> Use the Preview instead of saving the site to watch the output.
> The wiki has a versioning function. Every saved work will be hold in the
> database.
>
> Try the following text for your links in the sandbox:
> BEGIN-TRY
>
> =Your 1st Level Headline =
> ==Your 2nd Level Headline ==
> ===Your 3rd Level Headline ===
>
> Here is my Text with the Links:
>
> * 
[http://www.neuroworld.ws/tutorials/jsp_tutorial_multichoregraphy_control/jsp_tutorial_multichoregraphy_control.htm]
 Top Tutorial.
>
> * 
[http://www.neuroworld.ws/tutorials/jsp_tutorial_multichoregraphy_control/jsp_tutorial_multichoregraphy_control.htm]
 Top Tutorial.
>
> * 
[http://www.neuroworld.ws/tutorials/jsp_tutorial_multichoregraphy_control/jsp_tutorial_multichoregraphy_control.htm]
 Top Tutorial.
>
>
> END-TRY It's really easy,he.
> This should give a bullet-List (the stars in front of the links will be 
converted to bulletlist ;-) with your tutorial links.
>
> If you've played enough, don't save and go directly to the:
>
> http://rs3dwiki.the-final.info/index.php?title=Special:Upload
>
> Upload a nice picture for the tutorials (try to remember the imagename ;-)
> Then go here:
>
> http://rs3dwiki.the-final.info/index.php?title=Offsite_Tutorials
>
> Put this at the end of the "edit Textbox" to create a new tablerow with one 
column:
>
>
> |-
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> |}
>
>
> or this  to create a new tablerow with two columns ():
>
>
> |-
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial1|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial2|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> |}
>
>
> or this  to create two new tablerows with two columns ():
>
>
> |-
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial1|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial2|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> |-
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial3|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> 
{{AnnotatedLink|imgwidth=100|image=jp-imagename.jpg|title=JP-Tutorial4|description=Some
 JP-Tutorials on Neuroworld, September 20th
> 2006|link=http://www.neuroworld.ws/tutorials.htm}}
> |}
>
>
>
>
> Hope this was helpy,
> Matthias
>
>
>
>
>
>

Re: RS-Wiki is available :-?

[Timo Mikkolainen]

Big thanks from me too.

On 18/10/06, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:

Good job. Thanks. :)

Regards,
Fredrik Bergholtz


On 18/10/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Solved.
>
> Matthias
> - Original Message -
> Wrom: QTIPWIGYOKSTTZRCLBDXRQBGJSNBOHMKHJYFMYXOEAIJJP
> To: 
> Sent: Tuesday, October 17, 2006 9:30 PM
> Subject: RS-Wiki currently not available :-?
>
>
> > Freber has informed me that the wiki
> > is currently not available.
> > I'll call the provider as fast as possible to
> > see what's the error.
> >
> > Matthias
> >
>

Re: Editing Realsoft Wiki is a knightmare and a waist of time!

[Timo Mikkolainen]

On 18/10/06, Jean-Sebastien Perron <[EMAIL PROTECTED]> wrote:

There is no way I will put my tutorials there.

Don't put them there then.

I don't have 3 hours to loose understanding html

There's no Wiki markup editors (like WYSIWYG html editors) that I know
of, so this is a somewhat valid complaint. For most tutorials you only
need image tags and lists though.


Why everything that is designed to be free is overly complicated???
Thunderbird, Blender, Linux...

More control usually brings more complexity. I wouldn't call
Thunderbird hard to use though. Blender's UI design is somewhat
unorthodox (never bothered to learn it myself), but as you're not
paying for it, they don't need to listen to your complaints. Linux is
a complex thing (as are all operating systems), fortunately newer
distros are closing in on other desktop OS's by hiding the complexity
behind a simplified user interface. As a general note, free software
is created (mostly) by coders for coders and that shows: coders are
not UI designers. There has been some movement towards better UI
design in the OSS world though, like http://openusability.org


Jean-Sebastien Perron
www.neuroworld.ws

Re: Where can I find a tutorial about a landscape material?

[Timo Mikkolainen]

One more version, in this one the 'up' direction is the mapping object's z axis.

On 05/10/06, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:

Here's another version with curve control and noise.

On 05/10/06, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:
> The attached material renders 'slope' to surface illumination.
>
> On 04/10/06, Jean-Sebastien Perron <[EMAIL PROTECTED]> wrote:
> >
> >
> >
> >
> > Very impressive material.
> >
> >
> >
> > Sorry, but I can't understand this material, I am not a nuclear physicist.
> >
> >
> >
> > I would like to know just how to detect the slope to create a scope
> > material.
> >
> >
> >
> > Jean-Sebastien Perron
> >
> > www.neuroworld.ws
> >
> >
> >
> >
> >
> >  
> >
> >
> > De : [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED] De la part de Matthias Kappenberg
> >  Envoyé : Wednesday, October 04, 2006 7:44 AM
> >  À : user-list@light.realsoft3d.com
> >  Objet : Re: Where can I find a tutorial about a landscape material?
> >
> >
> >
> >
> >
> > something like this?
> >
> >
> > A very simple mountain.
> >
> >
> > shameless stolen some dust-material and used for snow :-)
> >
> >
> > (Was Karls "dusty jar" or something like this)
> >
> >
> > Rest is a simple point-translator, and some color materials.
> >
> >
> > Render with reasonably quality, drag'n drop the scene cam in a viewport.
> >
> >
> >
> >
> >
> > Hope it's helpy,
> >
> >
> > Matthias
> >
> >
> >
> > - Original Message -
> >
> >
> > From: [EMAIL PROTECTED]
> >
> >
> > To: user-list@light.realsoft3d.com
> >
> >
> > Sent: Wednesday, October 04, 2006 8:21 AM
> >
> >
> > Subject: RE: Where can I find a tutorial about a landscape material?
> >
> >
> >
> >
> >
> > Hi
> >
> >
> >
> >
> >
> > What excactly are you looking for? Define "landscape material". Do you want
> > a texture that can mimic a landscape texture, or a material that can let you
> > choose between different types of landscape texture coverage (low flat, low
> > slopy, high flat, high slopy etc)?
> >
> >
> >
> >
> >
> > For textures, use stock texture and repaint them to suit your needs.
> >
> >
> >
> >
> >
> > For scope selectors, import your mesh into World Builder and create
> > scopemaps from there. I'm sure it can be done
> >
> >
> > Can "easily" be done in RS3D, but not instant preview and great and quick
> > flexibility. VSL coding needed.
> >
> >
> > -Original Message-
> >  From: [EMAIL PROTECTED]
> > [mailto:[EMAIL PROTECTED] On Behalf Of Jean-Sebastien Perron
> >  Sent: Monday, October 02, 2006 3:09 PM
> >  To: user-list@light.realsoft3d.com
> >  Subject: Where can I find a tutorial about a landscape material?
> >
> > Where can I find a tutorial about a landscape material?
> >
> >
> >
> > Jean-Sebastien Perron
> >
> > www.neuroworld.ws
> >
> >
> >
> >
> >  This e-mail and any attachment are confidential and may be privileged or
> > otherwise protected from disclosure. It is solely intended for the person(s)
> > named above. If you are not the intended recipient, any reading, use,
> > disclosure, copying or distribution of all or parts of this e-mail or
> > associated attachments is strictly prohibited. If you are not an intended
> > recipient, please notify the sender immediately by replying to this message
> > or by telephone and delete this e-mail and any attachments permanently from
> > your system.
>
>
>





Slope3.r3d
Description: Binary data

Re: Where can I find a tutorial about a landscape material?

[Timo Mikkolainen]

Here's another version with curve control and noise.

On 05/10/06, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:

The attached material renders 'slope' to surface illumination.

On 04/10/06, Jean-Sebastien Perron <[EMAIL PROTECTED]> wrote:
>
>
>
>
> Very impressive material.
>
>
>
> Sorry, but I can't understand this material, I am not a nuclear physicist.
>
>
>
> I would like to know just how to detect the slope to create a scope
> material.
>
>
>
> Jean-Sebastien Perron
>
> www.neuroworld.ws
>
>
>
>
>
>  
>
>
> De : [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] De la part de Matthias Kappenberg
>  Envoyé : Wednesday, October 04, 2006 7:44 AM
>  À : user-list@light.realsoft3d.com
>  Objet : Re: Where can I find a tutorial about a landscape material?
>
>
>
>
>
> something like this?
>
>
> A very simple mountain.
>
>
> shameless stolen some dust-material and used for snow :-)
>
>
> (Was Karls "dusty jar" or something like this)
>
>
> Rest is a simple point-translator, and some color materials.
>
>
> Render with reasonably quality, drag'n drop the scene cam in a viewport.
>
>
>
>
>
> Hope it's helpy,
>
>
> Matthias
>
>
>
> - Original Message -
>
>
> From: [EMAIL PROTECTED]
>
>
> To: user-list@light.realsoft3d.com
>
>
> Sent: Wednesday, October 04, 2006 8:21 AM
>
>
> Subject: RE: Where can I find a tutorial about a landscape material?
>
>
>
>
>
> Hi
>
>
>
>
>
> What excactly are you looking for? Define "landscape material". Do you want
> a texture that can mimic a landscape texture, or a material that can let you
> choose between different types of landscape texture coverage (low flat, low
> slopy, high flat, high slopy etc)?
>
>
>
>
>
> For textures, use stock texture and repaint them to suit your needs.
>
>
>
>
>
> For scope selectors, import your mesh into World Builder and create
> scopemaps from there. I'm sure it can be done
>
>
> Can "easily" be done in RS3D, but not instant preview and great and quick
> flexibility. VSL coding needed.
>
>
> -Original Message-
>  From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED] On Behalf Of Jean-Sebastien Perron
>  Sent: Monday, October 02, 2006 3:09 PM
>  To: user-list@light.realsoft3d.com
>  Subject: Where can I find a tutorial about a landscape material?
>
> Where can I find a tutorial about a landscape material?
>
>
>
> Jean-Sebastien Perron
>
> www.neuroworld.ws
>
>
>
>
>  This e-mail and any attachment are confidential and may be privileged or
> otherwise protected from disclosure. It is solely intended for the person(s)
> named above. If you are not the intended recipient, any reading, use,
> disclosure, copying or distribution of all or parts of this e-mail or
> associated attachments is strictly prohibited. If you are not an intended
> recipient, please notify the sender immediately by replying to this message
> or by telephone and delete this e-mail and any attachments permanently from
> your system.





Slope2.r3d
Description: Binary data

Re: Where can I find a tutorial about a landscape material?

[Timo Mikkolainen]

The attached material renders 'slope' to surface illumination.

On 04/10/06, Jean-Sebastien Perron <[EMAIL PROTECTED]> wrote:





Very impressive material.



Sorry, but I can't understand this material, I am not a nuclear physicist.



I would like to know just how to detect the slope to create a scope
material.



Jean-Sebastien Perron

www.neuroworld.ws





 


De : [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] De la part de Matthias Kappenberg
 Envoyé : Wednesday, October 04, 2006 7:44 AM
 À : user-list@light.realsoft3d.com
 Objet : Re: Where can I find a tutorial about a landscape material?





something like this?


A very simple mountain.


shameless stolen some dust-material and used for snow :-)


(Was Karls "dusty jar" or something like this)


Rest is a simple point-translator, and some color materials.


Render with reasonably quality, drag'n drop the scene cam in a viewport.





Hope it's helpy,


Matthias



- Original Message -


From: [EMAIL PROTECTED]


To: user-list@light.realsoft3d.com


Sent: Wednesday, October 04, 2006 8:21 AM


Subject: RE: Where can I find a tutorial about a landscape material?





Hi





What excactly are you looking for? Define "landscape material". Do you want
a texture that can mimic a landscape texture, or a material that can let you
choose between different types of landscape texture coverage (low flat, low
slopy, high flat, high slopy etc)?





For textures, use stock texture and repaint them to suit your needs.





For scope selectors, import your mesh into World Builder and create
scopemaps from there. I'm sure it can be done


Can "easily" be done in RS3D, but not instant preview and great and quick
flexibility. VSL coding needed.


-Original Message-
 From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Jean-Sebastien Perron
 Sent: Monday, October 02, 2006 3:09 PM
 To: user-list@light.realsoft3d.com
 Subject: Where can I find a tutorial about a landscape material?

Where can I find a tutorial about a landscape material?



Jean-Sebastien Perron

www.neuroworld.ws




 This e-mail and any attachment are confidential and may be privileged or
otherwise protected from disclosure. It is solely intended for the person(s)
named above. If you are not the intended recipient, any reading, use,
disclosure, copying or distribution of all or parts of this e-mail or
associated attachments is strictly prohibited. If you are not an intended
recipient, please notify the sender immediately by replying to this message
or by telephone and delete this e-mail and any attachments permanently from
your system.


Slope.r3d
Description: Binary data

Re: Ok, OK I will do it, only because I can.

[Timo Mikkolainen]

Isn't this exaclty what the wiki is for? Just create a page for thequestion (as 
in an empty page for the tutorial) and people will createthe tutorial if they 
feel up to it. It's centralized and everyone canedit it.
On 20/09/06, Jean-Seastien Perron <[EMAIL PROTECTED]> wrote:> This is very interesting,> So I guess it solve all the problems.>> Next time I will double check what I write.>> It's a "must be very in needs" solution though.>> Next time I see a bright 
question and solution, I will compile it and edit> it to put on my website. Only if I have time for it : )>> Gary is right, the lists and post on a forum is not as good as a small> webpage accessible to anyone in one click in a tutorial or faq section.>> IF, everytime someone 
is stuck with a problem and ask this list, when he> find the solution, he should write a small resumé of the problem and the> solution and give it to me, I will put them on my website. I have 2 website.>> That should be the punishment of the person daring asking a question on 
this> list and bothering everyone (--exageration--). I mean the people that take> time to answer the questions (most of time it's the same 4-5 person) deserve> that little service. And they should write the name of the people giving> them the solution as a (thank you).>> We 
should not keep the knowledge to ourselves.>> I'll show the example, I will update my website with the trick about the> displacement rendering speed (someone gave me a while ago on this list).>> Jean-Sebastien Perron> www.neuroworld.ca>> -Message d'origine-> 
De: [EMAIL PROTECTED]> [mailto:[EMAIL PROTECTED] De la part de Boris Jahn> Envoyé: Wednesday, September 20, 2006 2:54 PM> Ŕ: user-list@light.realsoft3d.com> Objet: RE: Ok, OK I will do it, only because I can.>> Hi Jean-Sebastian,>> > I like the feel that this list is 
private. Even if it's just an illusion.> >>> > The point is that if we knew these reply were compiled on a database (and> > accessible), we would not discuss as we did in the past.>> But this list isn't private since long time. Simply check the both links> to the 
mail-archives I've posted.>>> -->> Bye> Boris - http://www.3ddart.com -> Realsoft Image Contest - http://www.realsoft.org -

Re: no wiki problem

[Timo Mikkolainen]

Wiki moved to http://rs3dwiki.the-final.info quite a long while ago.

On 06/09/06, Arjo Rozendaal <[EMAIL PROTECTED]> wrote:

what do you mean? as far as I can see the wiki is still up.

Arjo.

>
> I was impressed enough at the time to sit down and create one of
> the first RS Wiki pages ; bringing RG'z to the attention of wiki
> surfers .
>
>  That effort (still one of the very few efforts) appears to have
> been rendered null , for some reason . Yes , it's been deleted .
>
> Considered "defunct" by the Wiki powers that be , somehow ...
>
>   I tried to track the deletion but I'm only allowed to see
> the changes that have occured over the last 30 days . (guess
> how many changes ? yep . zero)
>
> Makes you stop and think , doesn't it (I hope) ?
>
> Garry
>
>
>

Re: Next Tag Problem

[Timo Mikkolainen]

Or just Self.SetTotalScale(Self.GetParent().Get("SCL"));
Assuming that the tag is a vector.

On 08/08/06, Timo Mikkolainen <[EMAIL PROTECTED]> wrote:

Try

Self.SetTotalScale(new
r3Vect(Self.GetParent().Get("SCL").x,Self.GetParent().Get("SCL").y,
Self.GetParent().Get("SCL").z));

Copy&Paste might fail as I didn't test it, but I'm sure you get the idea.

On 08/08/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:
> Hi all,
>
> my next "little" problem:
>
> http://the-dimension.com/pulsing-via-tag.zip
>
> the render time (preprocessing time) is
> growing up from frame to frame, while rendering this
> testscene. Is there a command to flush the memory or something
> else like this?
>
> This is the poor script for the "Use parent level Tag"
> The parent level has a Tag "SCL"
>
> daddy=Self.GetParent();
> scaler=daddy.Get("SCL");
> xa=scaler.x;
> xb=scaler.y;
> xc=scaler.z;
> Self.SetTotalScale(new r3Vect(xa,xb,xc));
>
>
> Matthias
>

Re: Next Tag Problem

[Timo Mikkolainen]

Try

Self.SetTotalScale(new
r3Vect(Self.GetParent().Get("SCL").x,Self.GetParent().Get("SCL").y,
Self.GetParent().Get("SCL").z));

Copy&Paste might fail as I didn't test it, but I'm sure you get the idea.

On 08/08/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:

Hi all,

my next "little" problem:

http://the-dimension.com/pulsing-via-tag.zip

the render time (preprocessing time) is
growing up from frame to frame, while rendering this
testscene. Is there a command to flush the memory or something
else like this?

This is the poor script for the "Use parent level Tag"
The parent level has a Tag "SCL"

daddy=Self.GetParent();
scaler=daddy.Get("SCL");
xa=scaler.x;
xb=scaler.y;
xc=scaler.z;
Self.SetTotalScale(new r3Vect(xa,xb,xc));


Matthias

Re: How to access a Tag with Javascript

[Timo Mikkolainen]

http://realsoft.fi/reference/javascript/real/code/r3ttag.html

Each tag type has its own get. No idea if it works though.

On 07/08/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:

Hi all,

 is it possible to acces a user defined Tag
(Properties Window-->Tags) via Javascript?
And if yes, how?


 Matthias

Re: Ray Path Categorization

[Timo Mikkolainen]

The paper's edge detection is purely based on identifiers, so all
internal lines would require a difference in identifiers. If you want
contour lines in RS, use the edge post filter in Geometric mode (this
might not be in the current Win version yet, but is in the Lin
version). Reflections have to be handled by storing normal and
distance from secondary rays to custom channels and adding them to the
Distance/Normal channels in post before doing the edge detection.

On 09/07/06, Neil Cooke <[EMAIL PROTECTED]> wrote:

I think the paper Matthew referred to might have had an effective edge
detect. Their Figs 8, 9 and 10 showed an internal line within the sphere
that was not part of a reflection ... although they then say, in the Fig 10
caption, that the black lines were painted in after. However it was painted
on using the AA data. The paper's examples use primitives or analytics so
the edge detect thing is not an issue since using RS analytic objects no
identifier is used.

The main issue, for me, is not reflections. It is the working lines of an
SDS object which are not wanted and the internal lines of edges within that
object which are wanted. The identifier gets rid of the working lines but
maybe AA data could restore internal edge lines. The problem would remain as
to what to do with the internal lines that would start showing in any
analytic spheres, for example, in the same scene.

The RS tutorproject shows the diminishing line strength with distance which
was not part of the paper Matthew refered to.

Neil Cooke


- Original Message -
From: "Matthias Kappenberg" <[EMAIL PROTECTED]>
To: 
Sent: Sunday, July 09, 2006 11:21 PM
Subject: Re: Ray Path Categorization


> Is this not something like that:
>
> Realsoft3D\tutorprojects\rendering\outlines.r3d
>
> Matthias
> > - Original Message -
> > From: "Matthew Hagerty" <[EMAIL PROTECTED]>
> > To: 
> > Sent: Saturday, July 08, 2006 6:14 PM
> > Subject: Ray Path Categorization
> >
> >
> > > Does anyone here understand this paper well enough to possibly
implement
> > > a material or script for Realsoft?
> > >
> > > http://www.cs.princeton.edu/~diego/academic/publications/NehGat00.pdf
> > >
> > > It seems lately that I'm finding more and more a need to have high
> > > quality outlines or rendered wireframe objects, which are currently a
> > > real pain to do.  Maybe Vesa might find it interesting enough to add
the
> > > functionality to Realsoft 3D directly? :-)
> > >
> > > Matthew
> > >
> >
> >

Re: Ray Path Categorization

[Timo Mikkolainen]

The project that Matthias pointed does pretty much everything in the
paper; reflections can be added by e.g. adding the ray identifier to
the surface identifier in a secondary ray shader. The only novel part
about the paper is the collision avoidance, which is a bit hard to do
in VSL. You'll just have to make sure that there's enough difference
between different objects' identifier values. You might notice that in
the project you don't get a line between the textured text and the
background: this is because they share the same identifier value.

On 09/07/06, Matthias Kappenberg <[EMAIL PROTECTED]> wrote:

Is this not something like that:

Realsoft3D\tutorprojects\rendering\outlines.r3d

Matthias
> - Original Message -
> From: "Matthew Hagerty" <[EMAIL PROTECTED]>
> To: 
> Sent: Saturday, July 08, 2006 6:14 PM
> Subject: Ray Path Categorization
>
>
> > Does anyone here understand this paper well enough to possibly implement
> > a material or script for Realsoft?
> >
> > http://www.cs.princeton.edu/~diego/academic/publications/NehGat00.pdf
> >
> > It seems lately that I'm finding more and more a need to have high
> > quality outlines or rendered wireframe objects, which are currently a
> > real pain to do.  Maybe Vesa might find it interesting enough to add the
> > functionality to Realsoft 3D directly? :-)
> >
> > Matthew
> >
>
>

Re: Tips for animating a long sequence?

[Timo Mikkolainen]

File instances load the object from a file. See
realsoft3d/reference/realsoft3d/models/geometrics/fileinstance.html

On 30/06/06, Matthew Hagerty <[EMAIL PROTECTED]> wrote:

File Instances?  Can you explain in more detail?  Almost all my objects
will be appearing in every cut.  Sorry for all the lowly user questions...

Matthew


Timo Mikkolainen wrote:
> I'd do every cut as a separate file, to keep away clutter. You can use
> file instances for objects that appear in more than one cut to keep
> the objects consistent (as in you only have to tweak the geometry in
> one file).
> Realsoft.info has a model repository.
>
> On 30/06/06, Matthew Hagerty <[EMAIL PROTECTED]> wrote:
>> Now that my model (lightcycle) is done (thanks to everyone for the color
>> changing tips), I want to animate it of course.  What good is a
>> lightcycle that does not zip around a game grid and smashing into walls?
>> :-)  This will be the longest animation I've ever attempted and any
>> insight would be greatly appreciated.
>>
>> The first question that keeps popping into my head is whether the whole
>> 3 to 5 minute sequence should be set up in a single file, or should
>> various shots be set up in smaller scene files dealing with only 5 to 30
>> second clips?
>>
>> Here is a link to a few renders I made last night.  Is there still a
>> repository of models being archived somewhere?  If so, I'll upload the
>> lightcycle.
>>
>> http://digitalstratum.com/programming/tron
>>
>> Matthew
>>
>>

Re: Tips for animating a long sequence?

[Timo Mikkolainen]

I'd do every cut as a separate file, to keep away clutter. You can use
file instances for objects that appear in more than one cut to keep
the objects consistent (as in you only have to tweak the geometry in
one file).
Realsoft.info has a model repository.

On 30/06/06, Matthew Hagerty <[EMAIL PROTECTED]> wrote:

Now that my model (lightcycle) is done (thanks to everyone for the color
changing tips), I want to animate it of course.  What good is a
lightcycle that does not zip around a game grid and smashing into walls?
:-)  This will be the longest animation I've ever attempted and any
insight would be greatly appreciated.

The first question that keeps popping into my head is whether the whole
3 to 5 minute sequence should be set up in a single file, or should
various shots be set up in smaller scene files dealing with only 5 to 30
second clips?

Here is a link to a few renders I made last night.  Is there still a
repository of models being archived somewhere?  If so, I'll upload the
lightcycle.

http://digitalstratum.com/programming/tron

Matthew