On Friday 14 May 2010, Neil Cooke wrote:
> Hi LeeE,
>
> Dont read this one!!!
>
> Some quick crits ... I have not yet solved the red sun issue. As
> in your pic it can often end up as almost the darkest thing in
> the sky ... how can that be when it is an energy source? An
> anti-matter sun ... neg-light?

Umm... It's supposed to be a small red dwarf in a close binary 
system with the larger yellow sun, which is the primary light 
source (this limited where I could place the major light source as 
being a close binary system the two stars need to be fairly close 
to each other).  The Red Dwarf does emit some light, but not very 
much, and given how little light it does contribute, and the fact 
that being so low on the horizon means it effectively casts no 
shadows, I guess I should just make it a low ql point source 
instead of an 11 ql spherical special light source.  I actually 
want to make it a little darker and perhaps mottle it a bit.  Part 
of the reason for doing this high ql render was to identify such 
issues though.

> Same for me, the only time I tried 
> it I was in a hurry, so just threw it out and made the thing
> orange but not saturated orange.
>
> My usual recipe for a sun is simply to select a colour between
> orange and white (through yellow) and give it the "unshaded"
> material. But have yet to spend time on the red star thing and
> suspect that the atmosphere around it has to do all the work with
> the star itself just 255 white.

The problem is that with a dynamic range of just 255 steps we're 
always going to be limited with how bright things can appear before 
they just end up at 24bit white.  There's not a lot we can do about 
this, apart from moving to higher bit depths.

>
> Also volumetric anything ... time is against me and I just use a
> procedural on a back plane if it is the sky and dont really
> bother for anything closer. If I need clouds to move in front of
> a moon or something then I just use a transparent setting on a
> surface in front of the further one. Stereo is not an issue since
> the distance is way beyond stereo separation range for human
> sight.
>
> The landscape foreground is just too boring for me. Landscapes
> for still images for me are displacement mesh surfaces with a
> colour map that is also copied as a bump map for ground cover.
> However mostly I use SDS surfaces for landscapes.

I agree that the foreground is a bit boring and I really need to add 
some features there.  Using SDS for landscapes is a bit expensive 
though, to my mind, and I'll generally start with an analytic plane 
as a starting point, and then perhaps add some nurbs or SDS 
features on top of it.

>
> The pylons that hold the tubes up seem clunky. I think some
> variation of detail on at least the closer ones might be useful.
> Even rows of rivet sets in "I" beams, maybe some rust patches.

I need to adjust the AO settings to better illuminate the tube 
supports, and perhaps add some sort of access ladder and a few 
panels, but this is supposed to be advanced technology, so no 
rivets and definitely no rust ;-)

>
> The actual capsule has guides that touch the inner surface of the
> glass it seems to me. There could be some kind of rail at those
> contact points since friction could be an issue .... although
> mag-lev at those points or future tech-magic etc. Lol.

Well, advanced tech again here, and if you had rails then you 
wouldn't need the entire tube.  As it happens, I've already got an 
idea for a soap-bubble type material for the tubes, which are 
supposed to be some sort of force-field constructs anyway, and I 
want to make the 'shoes' less clunky than the simple polyhedrons 
they are now as well.  I might add a bit of glow to the bottoms of 
the shoes too.  I might also add a few smaller (notionally robotic) 
barnacle/limpet type vehicles travelling along the bottom (and 
perhaps even the sides and tops) of the tubes, between the legs of 
the large pods as well.

>
> Which all just goes to show that your are ambitious and not
> scared of the tricky stuff and I am a coward!! Lol.
>
> However, the style of your pic is so close to a sci-fi
> illustration genre from mid last century. (But the dark red sun
> remains an issue for me there) So, on that basis you are bang on
> I reckon. I personally never really liked that genre, thought it
> was poor workmanship, but that's just me..
>
> Thanks
>
> Neico

Heh - the 1950's style is a little earlier than I was aiming for but 
I'll happily admit to being greatly influenced by the work of Chris 
Foss in the latter _quarter_ of the last century.

LeeE

Reply via email to