> Therefore you probably have to deploy an ICETree to each instance master setting the materialID
Correct, and as Steven mentioned, this is how materials are propagated in CrowdFX, by applying an ICETree to set materials on the source geometry. The 'Materials ICETrees' are created using a python script, so you can take a look at the code (I'm not in front of Soft so I can't tell in which file it is exactly, but I guess it is in the helper module of the CrowdFX addon). That could help you in writing it in C++ (if you really need to, as calling directly native c++ commands from python is not really slow). Cheers, Guillaume On Thu, Jul 12, 2012 at 8:28 AM, Leo Quensel <[email protected]> wrote: > It does work in creating the actual topology yes. Thank you for that. > For the material transfer I think there is a limitation in that you cannot > access material data in ICE unless it has been set by ICE (At least I > couldn't do it). > Therefore you probably have to deploy an ICETree to each instance master > setting the materialID and use that back in the created mesh to > reassign the materials. > > I think this is a good exercise for writing a first C++ plugin for me and > see how far I can get ;) > > > > > -------- Original-Nachricht -------- > > Datum: Thu, 12 Jul 2012 10:53:03 +0100 > > Von: Ciaran Moloney <[email protected]> > > An: [email protected] > > Betreff: Re: Freeze ICE tree to geometry? > > > Did Grahame's or my compound work for you? I'd say his would be your best > > bet since the point of mine was to allow attribute transfer onto the > final > > mesh, which adds some overhead for sure. > > > > Ciaran > > > > On Thu, Jul 12, 2012 at 9:29 AM, Leo Quensel <[email protected]> wrote: > > > > > Well it kinda works if you plug in each instance master manually - it > > does > > > not support groups however, which I totally need. I have over a hundred > > > instance masters in my case which would make the problem of plugging > > them > > > up one by one rather cumbersome. Even though it can possibly be > scripted > > it > > > would make the ICEtree interaction ultra slow. > > > Furthermore, the demo scene does not copy the > > > instances exactly but randomly distributes them it seems - the point > > cloud > > > and the mesh do not match, while I need an exact match from the point > > cloud > > > to the generated mesh, preserving UVs aswell as Material information. > > > > > > Another possibility is scripting this using the point cloud information > > to > > > distribute duplicated meshes, which is pretty easy indeed. This is > > however > > > ULTRA slow - it took almost 90 seconds to distribute 50 copies of my > > > instance masters (simple grids). I need to distribute 12.0000, so this > > is no > > > solution either (the Duplicate function is really, really slow in > Soft). > > > > > > The best solution so far? Bring the pointcloud and instance masters > over > > > to Maya and do the duplication there :( > > > > > > > > > > > > -------- Original-Nachricht -------- > > > > Datum: Wed, 11 Jul 2012 13:55:37 -0700 > > > > Von: Steven Caron <[email protected]> > > > > An: [email protected] > > > > Betreff: Re: Freeze ICE tree to geometry? > > > > > > > im confused... when i used it i had issue with particle clouds with > > many > > > > different instance shapes and if i deleted particles. maybe you > > updated > > > > it? > > > > > > > > regardless ciaran's looks like it works and if that doesn't i know a > > user > > > > can employ the same technique used by crowdfx to get a polymesh per > > > > instance shape. > > > > > > > > On Wed, Jul 11, 2012 at 1:51 PM, Guillaume Laforge < > > > > [email protected]> wrote: > > > > > > > > > No, it supports multi-instance masters, unless my memory is really > > > > bugged > > > > > ;) > > > > > > > > > > But I doubt as I remember supporting more than one shape was the > > > > > interesting thing in building this compound. > > > > > > > > > > Cheers, > > > > > > > > > > Guillaume Laforge > > > > > > > > > > > > > > > On Wed, Jul 11, 2012 at 8:23 PM, Steven Caron <[email protected]> > > > wrote: > > > > > > > > > >> yes, but it was already linked :) > > > > >> > > > > >> leo's issue with this compound was that it worked with only one > > > > instance > > > > >> master. > > > > >> > > > > >> > > > > >> On Wed, Jul 11, 2012 at 11:01 AM, Marc-Andre Carbonneau < > > > > >> [email protected]> wrote: > > > > >> > > > > >>> Is that what y’all looking for?**** > > > > >>> > > > > >>> > > http://www.si-community.com/community/viewtopic.php?f=15&t=1579**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> *From:* [email protected] [mailto: > > > > >>> [email protected]] *On Behalf Of *Steven > > Caron > > > > >>> *Sent:* 11 juillet 2012 13:45 > > > > >>> > > > > >>> *To:* [email protected] > > > > >>> *Subject:* Re: Freeze ICE tree to geometry?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> its there, its just not a single click.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> also implementing it through ICE means you have to implement the > > > logic > > > > >>> too... thats actually the hard part in my opinion. at least if > you > > > try > > > > to > > > > >>> do it with pure ICE workflow. if you want a run once script that > > can > > > > be > > > > >>> done quite easy and i think has been done many times over. which > > > leads > > > > me > > > > >>> to another question/point... do you want animation on the objects > > > too? > > > > >>> cause that changes some things...**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> s**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> On Wed, Jul 11, 2012 at 8:43 AM, Tim Crowson < > > > > >>> [email protected]> wrote:**** > > > > >>> > > > > >>> Exactly. That's the problem I'd like to find a solution for. > > > > Obviously, > > > > >>> I don't want to go converting 40 bazillion ICE polygons to real > > > > polygons. > > > > >>> But a few thousand, sure. > > > > >>> > > > > >>> Modo's replicators, while not in the same league as ICE in terms > > of > > > > >>> power, make it very easy to convert those instances to geo by > > merely > > > > >>> 'freezing.' Of course, Softimage lacks any native function for > > > > converting > > > > >>> instances to independant geo, so I guess it's a related obstacle. > > > > Still, > > > > >>> with ICE modeling, seems like the tech is there to allow this > kind > > of > > > > >>> conversion somewhere inside Soft. > > > > >>> > > > > >>> -Tim**** > > > > >>> > > > > >>> > > > > >>> > > > > >>> On 7/11/2012 10:23 AM, Leo Quensel wrote: **** > > > > >>> > > > > >>> Still it remains a pointcloud and not geometry which leaves**** > > > > >>> > > > > >>> you with the same limits as before.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> -------- Original-Nachricht --------**** > > > > >>> > > > > >>> Datum: Wed, 11 Jul 2012 15:13:08 +0000**** > > > > >>> > > > > >>> Von: Sandy Sutherland <[email protected]> > > > > <[email protected]>**** > > > > >>> > > > > >>> An: "[email protected]" > > > > <[email protected]> <[email protected]> > < > > > [email protected]>**** > > > > >>> > > > > >>> Betreff: RE: Freeze ICE tree to geometry?**** > > > > >>> > > > > >>> If you freeze the pointcloud - you can easily add an ice tree to > > it > > > > and**** > > > > >>> > > > > >>> for e.g. just change the instance shape to another using that new > > ice > > > > tree -**** > > > > >>> > > > > >>> the system is very powerful that way, the new ice tree just > builds > > > > upon the**** > > > > >>> > > > > >>> data frozen into the point cloud.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> S.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> _____________________________**** > > > > >>> > > > > >>> Sandy Sutherland**** > > > > >>> > > > > >>> Technical Supervisor**** > > > > >>> > > > > >>> [email protected]**** > > > > >>> > > > > >>> _____________________________**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ________________________________**** > > > > >>> > > > > >>> From: [email protected]**** > > > > >>> > > > > >>> [[email protected]] on behalf of Tim > Crowson > > > > [[email protected]]**** > > > > >>> > > > > >>> Sent: 11 July 2012 16:50**** > > > > >>> > > > > >>> To: [email protected]**** > > > > >>> > > > > >>> Subject: Re: Freeze ICE tree to geometry?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> I can, and I do eventually, but I don't have any control over > > the**** > > > > >>> > > > > >>> subcomponents in that frozen system. I can't model on top of it, > > > > extract anything,**** > > > > >>> > > > > >>> or tweak anything. It's literally a locked pointcloud with > > > > inaccessible geo**** > > > > >>> > > > > >>> in it.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Or is it? Is there some way of accessing the raw geo inside it > > that I > > > > have**** > > > > >>> > > > > >>> overlooked?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> -Tim**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> On 7/11/2012 9:39 AM, Ciaran Moloney wrote:**** > > > > >>> > > > > >>> Can't you just freeze the first scatter system? It should keep > all > > > > the**** > > > > >>> > > > > >>> particle attributes as they were. Or, drop an empty simulation > > region > > > > into it,**** > > > > >>> > > > > >>> so the scatter system will only ever be evaluated once instead of > > > > every**** > > > > >>> > > > > >>> frame.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Ciaran**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> On Wed, Jul 11, 2012 at 3:29 PM, Tim Crowson**** > > > > >>> > > > > >>> <[email protected]<mailto: > > > [email protected]> > > > > <[email protected]>> wrote:**** > > > > >>> > > > > >>> It's actually nothing very interesting. In fact it's bad for > > > > performance.**** > > > > >>> > > > > >>> I'm using Milan Vasek's (awesome) scatter tools to scatter small > > > > spikes on**** > > > > >>> > > > > >>> a stick (somewhat like the small spikes on an evergreen frond). > > Only > > > > about**** > > > > >>> > > > > >>> 100-150 of them. Then I parent the resulting pointcloud to the > > > emitter > > > > geo**** > > > > >>> > > > > >>> (the stick) and use that as the instance source for a second > > > > scatter**** > > > > >>> > > > > >>> system. This gives me the control I want for both scatter > systems. > > > > However, if I**** > > > > >>> > > > > >>> manually model the equivalent of the first scatter system, and > use > > > > that as**** > > > > >>> > > > > >>> the instance source for the second system instead of another > > > > Scatter,**** > > > > >>> > > > > >>> performance is far better. Which stands to reason.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Would someone smart kindly make an L-system in ICE?**** > > > > >>> > > > > >>> :)**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> -Tim C.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> On 7/11/2012 8:11 AM, Mihail Djurev wrote:**** > > > > >>> > > > > >>> Hello SI list!**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Tim, could you share with us how you achieved that?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Mihail**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> On 10.7.2012 г. 18:06 ч., Tim Crowson wrote:**** > > > > >>> > > > > >>> I'm rendering this with Mental Ray. I found a solution that lets > > me > > > > render**** > > > > >>> > > > > >>> what I want without having to freeze it to geo. But it would > still > > be > > > > nice**** > > > > >>> > > > > >>> to know how to do it.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> -Tim**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> On 7/10/2012 12:45 AM, Sandy Sutherland wrote:**** > > > > >>> > > > > >>> What are you rendering with?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> S.**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> _____________________________**** > > > > >>> > > > > >>> Sandy Sutherland**** > > > > >>> > > > > >>> Technical Supervisor**** > > > > >>> > > > > >>> > > > > [email protected]<mailto: > > > [email protected]> < > [email protected] > > > >**** > > > > >>> > > > > >>> _____________________________**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ________________________________**** > > > > >>> > > > > >>> > > > > >>> From:**** > > > > >>> > > > > >>> > > > > [email protected]<mailto: > > > [email protected]> < > > > [email protected]>**** > > > > >>> > > > > >>> > > > > [[email protected]<mailto: > > > [email protected]> < > > > [email protected]>] on behalf of > > > > phil harbath**** > > > > >>> > > > > >>> [[email protected]<mailto:[email protected]> > > > > <[email protected]>]**** > > > > >>> > > > > >>> Sent: 10 July 2012 07:05**** > > > > >>> > > > > >>> To:**** > > > > >>> > > > > >>> > > > > > > [email protected]<mailto:[email protected]> > > > <[email protected]>**** > > > > >>> > > > > >>> Subject: Re: Freeze ICE tree to geometry?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> momentum instancer perhaps.**** > > > > >>> > > > > >>> ----- Original Message -----**** > > > > >>> > > > > >>> From: Tim Crowson<mailto:[email protected]> > > > > <[email protected]>**** > > > > >>> > > > > >>> To:**** > > > > >>> > > > > >>> > > > > > > [email protected]<mailto:[email protected]> > > > <[email protected]>**** > > > > >>> > > > > >>> Sent: Monday, July 09, 2012 6:24 PM**** > > > > >>> > > > > >>> Subject: Freeze ICE tree to geometry?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> I'm needing to freeze the results of my non-simulated ICE tree to > > > > geometry**** > > > > >>> > > > > >>> so I can optimize its use in another ICE tree. (wanting to > > > scatter**** > > > > >>> > > > > >>> something that was built by a scatter to beginwith).**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Concerns about geo density not-withstanding, is there a way to > > freeze > > > > a**** > > > > >>> > > > > >>> non-simulated ICE tree to editable geometry?**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> --**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Tim Crowson**** > > > > >>> > > > > >>> Lead CG Artist**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> Magnetic Dreams Animation Studio, Inc.**** > > > > >>> > > > > >>> 2525 Lebanon Pike, Building C. Nashville, TN 37214**** > > > > >>> > > > > >>> Ph 615.885.6801 | Fax 615.889.4768 |**** > > > > >>> > > > > >>> www.magneticdreams.com<http://www.magneticdreams.com> > > > > <http://www.magneticdreams.com>**** > > > > >>> > > > > >>> > > [email protected]<mailto:[email protected] > > > > > > > > <[email protected]>**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> --**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> --**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> --**** > > > > >>> > > > > >>> - Ciaran**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> --**** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >>> -- **** > > > > >>> > > > > >>> **** > > > > >>> > > > > >>> ** ** > > > > >>> > > > > >> > > > > >> > > > > > > > > > > > > > > > > -- > > - Ciaran >
