Sub surface scattering again...

Started by Hannes, November 26, 2014, 06:42:58 AM

Previous topic - Next topic

j meyer

Just saw the turntables and I like what I saw.
The glowing ears are of course due to the depth map and were
the reason I chose that map for the static pic.
For an animation I'd try it without a map first.
Don't know how the industry guys handle SubSurfScattering mapping
for animation,maybe they can be animated,too,like displ maps.
A map is needed even for stills to deal with the differences of tissues
and bones as far as I know.
The translucency feature in TG seems to be connected to the shadows
somehow,at least it's mostly noticeable in those areas.
Maybe we have to find a balance between mapped and unmapped
translucency.
Tried it with a lambert shader on some Vdisp thing with and without
a map (a mask rather),but ... not quite there yet.


And again,if someone wants tests with Daz/poser models provide a
model or a link to a free one or do it youself,please.I don't have such
a model.

j meyer

Ulco is that "water" separate geometry like a plane or so
or just the planet surface with an other disp setting than
the river bed?

TheBadger

QuoteHere is a (literally) small movie of the scan head in motion.

Well that the most detailed looking human I have ever seen in TG. I think that it could really look great with fine tuning.
One thing I was looking for as the head turned, was to see light coming through the cartilage of the ear, from the sun. Don't think I saw that? Not sure. Ill have to look again when I feel better (sick as a dog right now)

Quoteif someone wants tests with Daz/poser models provide a
model or a link to a free one or do it youself,please.I don't have such
a model.

I can't think right now. Yossam knows where to find free stuff like that. If you are reading this yossam, please post a link if you have one that would be good here.

Going to go die now. See you all on the other side.
It has been eaten.

Dune

@Jochen: it's a plane. Otherwise I couldn't get the front sides of the stream to be transparent. It's supposed to be a sediment/airbubble laden fast moving stream in this experiment, so only a bit of transparency was needed.

Hannes

Jochen, the thing is, we don't have real sub surface scattering here, only translucency which works for thin one sided objects, like paper or fabric. Real sub surface scattering is kind of volumetric which takes the thickness of an object into account. So to answer Michael's question: the external ears are backlit all the time, because TG thinks the head is kind of a thin shell and not a volumetric object. The depth map simulates the volumetric nature of SSS by increasing the strength at certain parts like the nose or the ears, but unfortunately all the time, no matter if these parts are actually backlit or even covered by other objects (or itself).
Other programs that include SSS shaders seem at least to be able to simulate this volumetric behaviour somehow. I did some tests with my wax ball (a few posts ago) and put a long black object right through the ball, just to see if there is some lighting reaction of the translucency shader. Nothing.  :(

So the search continues...

(I hope my english is halfways understandable!)
Michael, I hope you feel better soon!!!!!

Hannes

Something else. Ulco, what do you think about using this method for water? As far as I can see in your (fantastic!) efforts, the fake SSS thing would only be useful for water areas like high waves seen from the side like those that Ivan Konstantinovich Aivazovsky had painted so incredibly?! I think, if rendertimes are so much higher than those of the water shader, there wouldn't be a benefit, right?

Dune

You're right, but I'm still testing the river and it seems the slowness comes from something else, now that I've replaced the default setup by a simple color PF. I think it may be a waste of time pursuing this, as no doubt Matt can think of a way to make the glass shader see both sides of a plane, and thus the shadows or (wave) backside foam. So I'm out, for the time being. Other things to do.

Best wishes from me too, Michael. Rest, rest, rest, that always helps (well, almost always).

bobbystahr

Didn't know this cold was catch-able online but I've as well been struck by sniffles in the sinus and and sandpaper throat...in addition to a screwed up left wrist which has kept me from guitar playing for a week and a half now...goin snakey with no music fix and listening to it just don't cut it. I figger it'll be worth payin' some lil kid to shovel the snow this year if this is how it treats my wrist.
something borrowed,
something Blue.
Ring out the Old.
Bring in the New
Bobby Stahr, Paracosmologist

Hannes

Best wishes to you as well, Bobby!!

I think I have to correct myself. I just made a test of a stylized cartoon head made of a simple sphere with extruded (and smoothed) faces representing the nose and the ears and with the usual fake SSS shader attached. I then placed a slightly smaller black simple sphere with no faces extruded inside that first cartoon head. So it's kind of like a head with a skull inside that lets no light through of course.
Look at the image, there is no depth map included. It seems objects placed inside other objects do affect the SSS effect!

bobbystahr

something borrowed,
something Blue.
Ring out the Old.
Bring in the New
Bobby Stahr, Paracosmologist

bobbystahr

Quote from: j meyer on December 02, 2014, 11:58:03 AM
Just saw the turntables and I like what I saw.
The glowing ears are of course due to the depth map and were
the reason I chose that map for the static pic.
For an animation I'd try it without a map first.
Don't know how the industry guys handle SubSurfScattering mapping
for animation,maybe they can be animated,too,like displ maps.
A map is needed even for stills to deal with the differences of tissues
and bones as far as I know.
The translucency feature in TG seems to be connected to the shadows
somehow,at least it's mostly noticeable in those areas.
Maybe we have to find a balance between mapped and unmapped
translucency.
Tried it with a lambert shader on some Vdisp thing with and without
a map (a mask rather),but ... not quite there yet.


And again,if someone wants tests with Daz/poser models provide a
model or a link to a free one or do it youself,please.I don't have such
a model.

I believe the Free DAZ 4+ comes with some models in it's bundle.
something borrowed,
something Blue.
Ring out the Old.
Bring in the New
Bobby Stahr, Paracosmologist

Hannes

Quote from: j meyer on December 02, 2014, 11:58:03 AM
Tried it with a lambert shader on some Vdisp thing with and without
a map (a mask rather),but ... not quite there yet.

Jochen, Matt once said, that in TG the Lambert shader is nothing but a simplified shader with only diffuse and translucent settings. So it's nothing special. There are no further translucency settings than in a regular shader.

j meyer

Bobby - thanks,but I'm not interested in having that.And just for testing to help out
           others...nay,too much hassle.So either a model someone really uses or
           someone else has to test that,sorry.
           As for your wrist keyboards or sequencers can help in times like that,at
           least a bit.In my experience that is.
           Hope you get well soon!


Yes Hannes,I know,but there can be circumstances where a default shader is
not an option,like in my tests with Vdisp "terrains".
In this example I had to use a lambert shader.
And my experiments seem to suggest that the lambert shaders "use existing colour"
feature doesn't take reflectivity into account,which can be problematic at times,btw.

[attach=1]

Ulco,the reason I asked was another strange effect I saw when rendering
translucency on a different Vdisp thing.It looked like TG rendered an addi-
tional layer for translucency first and on top of that the final surface.
Didn;t notice that behaviour with the other file(s).Kinda strange.
Your pic reminded me of that,but your answer shows that it's most likely
something different.
More thinking and testing required apparently.

Get well everybody!


Dune

Nice shapes!

The dark is probably the shadow, like I saw in the waves with the fish.

Hannes

Here is a test animation of the cartoon head. See how the ears or the nose are almost not glowing until they get into the light.