In these renders I'm comparing 3 different SSS techniques in 4 different lighting situations. Each row is a different light setup. Each column is a different shader. All four shaders have lots of stuff besides the SSS, but in each only the SSS distribution technique is changed. The SSS color is the same. The amount was adjusted so that in the 50/50 light (first row), they produce roughly the same amount of SSS. They differ in how the SSS is distributed. Of course all of these can be further tweaked up the wazoo, but I'm trying to do some kind of basic comparison here.

The left-most figure has no SSS. The other three do. Each technique is based on just a few key shader nodes. I'm not going to describe just yet the whole shader, but the key nodes that produce the different SSS distributions (IN NO PARTICULAR ORDER) are:

Cosine technique (by me): 
Cos(Diffuse - a constant)

Incidence technique (by face_off, see http://www.physicalc-software.com/tutorials/realskinshader):
Vector Dot Product of surface normal (N node) and Vector to light

Toon technique (by Stahlberg, see http://www.deepwaterstudios.com/maya/modules/wfsection/article.php?articleid=5):
a Toon node

The clear winner on ease of setup is the Toon. It's just one fricking node.

What's your opinion about which one seems most realistic. If you don't think there's a clear winner, than say what you like for each lighting scenario.

And can you guess which is which?

Nobody is allowed to say they like the leftmost one best :-)

I'm gonna go with the far-right column....

even though to my old eyes it's hard to really see a difference other than the obvious color changes in the last couple of rows....

(and yes...I did open the full-sized image) ;)

Yeah, it's a toss-up between the far right and the 2nd from the left, but I think
the far right wins out.  (not trying to make any kind of political statement  )
As for lighting I like the 2nd row.

1. No SSS
2. Toon ?  Looks fine in the top 2, but the ambient/luminance seems to have too much affect in low-lighting conditions.
3. face_off ?
4. baggins ?

...not sure on 3 & 4, which is which.  I guess I'd only be surprised if one of those was the toon one.   Obviously #3 shows less of the SSS influence in lower-lighting conditions (I generally use a 20-30% IBL, btw), but I find both 3 & 4 about equally desirable, I think.

I would have to say the furthest right wins out overall, but the second from left is also good in the first two rows. Second from right is better in the low-lighting conditions.

Same here, column 4, followed closely by 2. Not sure about the lighting, what time of day or setting is it representing?

Right and then the second from left.

my first vote, from a mac power book: 4th column (far right). 

edited to change vote on zooming.

The 4th column looks best for me...

Column 4

Thank you all for your feedback.

Spanki was right: the images are

#1 None, #2 Toon, #3 Incidence, #4 cosine

Any of the techniques can be made to look very close to any of the others by changing parameters. My goal was to find a technique that doesn't require so much parameter changing.

Of course the incidence technique (#3) looks really good, but it has always troubled me that it needs to be updated when you move lights, needs adjustment in different lighting conditions, and doesn't work so great with IBL.

The cosine technique isn't perfect either, but it seemed to me that, overall, it is closer to what you want with less tweaking.

I've stared at these so long, and am so invested in the cosine technique, I was concerned about my own bias influencing my belief that it was better than the others.

my second vote from a pc with a viewsonic lcd: column 3

Interesting.  Overall, it looks like you have a workable solution  - it probably has too much influence on the skin color/tone in low lighting, but I could live with that if the node setup was easier to deal with.  I'd be interested in seeing your node setup so I could try it under some other lighting situations.

Spanki -

Do you want just the Cosine based front-side SSS part of the shader tree or the whole thing with incidence SSS, toon SSS, freckles, petechiae, moles, two-part specular sheen, backlist SSS, soft reflections, moveable head bruise, moveable poison ivy, and HSV and color bias dials? All of these effects are in one shader and can be adjusted with parameter dials on the figure - no need to go into the material room and edit the shader.

This is all going to be in one of the demos for my free software, Matmatic, I'll be releasing "any day now". I've pre-release a number of demos since then, but this Cosine SSS is not among them.

So do you want the Matmatic script, a Poser 6 material collection, a Poser 5 mat pose file, or individual materials? And do you want any of the other goodies or just the SSS?

If you want to get involved in the Matmatic or Parmatic beta check out these threads:

Matmatic Announcement and discussion: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=228105

Matmatic Download: http://loftydesigns.net/Beta/Beta/MatmaticBeta1.zip

Parmatic: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=226110

Parmatic SR3 Issues: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=234750

Threads where I've used these for special effects:

Deep Ocean Water: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=226110

Incidence SSS on Jessi: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=229578 (although I think I did that one wrong)

Algroithmic Eye Makeup: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=229726

Reptile Scales: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=228528

Goldfish Scales: http://www.runtimedna.com/mod/forum/messages.php?ShowMessage=233907

Heh.. actually, I was curious about just the Cosine based front-side SSS part of the shader tree .

I've been following some of your Matmatic/Parmatic work (very impressive), but I hadn't had a chance to play with that yet.  I love the ocean-water thing, btw.

Overall I'd say the rightmost coloumn, but it differs in the various lightning situations :o)

They all look good.

Attached you will find a Poser 6 material collection with head and body shaders for James.

Becauuse of forum posting rules I renamed it.

*** THIS IS NOT A WORD DOCUMENT FILE - REMOVE THE .DOC - IT IS an .MC6 FILE ***

There are parameters in here, usable with Parmatic. If you want, you don't have to have Parmatic. Just go into the shader tree, find the nodes labelled PM:something. These are just simple numbers that feed into other nodes.

There are only four parameters:

PM:cos SSS - the amount of SSS to apply - my default is 1

PM:cos SSSQ - controls how narrow the SSS band is - higher is more wide - my default is 1

(Edited - i said higher is more narrow - that was backward)

PM:cos SSS Bias  - controls a bias - higher is more, lower is less - my default is .4

PM:cos SSS MV - controls the "middle value" - lower moves SSS into shadows, higher moves into bright areas - my default is 1

There are also a three other easy to find parameters:

PM:Specular:Specular_Value, PM:Blinn:Reflectivity, and PM:Diffuse

These should be obvious what they do.

If you set PM:Diffuse to 0, and set PM:cos SSS to 10, you can see just SSS and specular, which is helpful when trying to figure out what's going on.

I'll post some renders next.

I can no longer edit the earlier post. I reversed the pictures first time.

So Q Up is more narrow. Q Down is more wide.

A couple other notes about this shader you may find different than normal.

I subtract half the diffuse value from the specular color (WHITE). This softens the speculars and avoids color blooming artifacts that produce pink-yellow bands around shiny areas in bright lights.

I also subtract half the specular from the SSS, so SSS does not show up where the specular is strong. Does this have any correspondence with reality? I have no idea, but I like the effect.

bookmark

I'm not sure I'm much of a fan "skin shader comparison's" that include my shader, since I've got no control over whether the user is using my skin shader correctly.  Also, you comparison image is way too small to compare anything!  I doubt anyone in this forum is doing realism renders features James that small.  Give me a 600x600 render of James, and I can make a good guess if it is using my skin shader, but not on images these small.

Also, you are comparing your setup with Stahlberg's (which must be at least 3-4 years old now - and he no longer uses that technqiue!), and my original technique, which is nearly 2 years old, and doesn't represent the techniques used in the more recent Skin Realism Kits (which have significant improvements). 

My current skin shader produces results like that attached.  To me that looks realistic.  It was done with 1 spot light, and 1 IBL fill light.  I'd be happy to render that scene with your node setup to give a better comparison between the shaders.

Hey Paul!

I didn't say that was your shader. I was just giving you credit for inventing the technique in Poser. Same with Stahlberg, just giving credit.

Lest anybody is misled, I made that shader using information I got from you. It was a simplified incidence angle shader without any of the other features you put in. You're complete shader is way more than just an SSS function.

I'm just exploring different ways to do it, because it's an interesting puzzle. The renders by you and others with your shaders are, without question, the best skin I've seen in Poser renders that don't involve HOURS and HOURS of custom texture work.

I'm not trying to compete with you, by any means. I don't think anything I've posted is going to cut into your sales :-)

umm, can i just say i highly appreciate the contributions that both of you have made to realism in poser?  wow, face_off, that picture is incredible. i won't even ask who the figure is, because those shoulders are so great i can't imagine it's anything but a figure in beta.  bagginsbill, it would be nice to have an option that doesn't need recalculating with light changes.  if i were better at lighting i wouldn't need to experiment so much, but i'm still working on that.

again, thanks to you both.

Attached Link: Luxology forum Ben Kingsley thread by Jacques Defontaine of Belgium

Follow the link. Scroll halfway down.

That looks real.

Keep on trying and never give up. My favorite is the one farthest to the right. :D
:b_smile:

No problem Baggins.  I just get a little sensitive when ppl compare my shaders with theirs.  Most of the time they use favourable light and renders settings for their shader, which is not fair.

Anyway...Any SSS technique that does not take into account the positioning of the main light source is going to compromise realism.  I think your issue with the Incidence technique is that you need to run the shader again if you move the main light.  In practice I don't think this is an issue for more users, since they finalise their light positioning prior to running the shader.  Regardless, the next version of the product has a callback which automatically reruns the shader if the user changes the main light position.

Skin realism is all about having the skin look realistic relative to the figure's surroundings.  Shaders need to respond to the lighting and objects in the scene, otherwise the figure is going to look out of place.  "which skin looks better" comparison's that have the figure on a black background are not testing the shader in an environment that users will be using it.  The incidence effect starts to build a skin model that is reacting to the scene lighting.  If you want more realism, I think /simplifying/ the SSS model is going in the wrong direction.  You need to be building a shader than takes into account MORE of the scene information, not LESS.

actually, i like face_off's picture better than the kingsley one.  while the kingsley one has a lot of detail, he looks dead.  and really cg.  i would have known it was 3d in an instant without being told.  face_off's i would have just assumed was a photo.

sometimes, more detail is just gilding the lily.

LOL, I had preferred Paul's settings as well :)

No problem Baggins.  I just get a little sensitive when ppl compare my shaders with theirs.  Most of the time they use favourable light and renders settings for their shader, which is not fair.

Cool. Everybody should understand I was testing 3 shaders of mine, each built using different techniques, two of which I did not invent. But they are my implementations, and if they suck it's my shaders that suck, not incidence shaders in general. Maybe I didn't build mine same as yours. Maybe yours doesn't blow up under different light conditions. I don't know - I don't have any of your shaders and have never seen the shader trees.

However, I was comparing shaders fairly. Each row is a single render with four figures - therefore they have identical render settings and lighting conditions. And the four rows was deliberately trying to examine how the appearance evolves as the lighting is changed. Again, a single render yielded whatever results it yielded for all four. There was no tampering.

*Anyway...Any SSS technique that does not take into account the positioning of the main light source is going to compromise realism. *

I agree, but why did you feel the need to point that out? I ask, because all three shaders do take the lights into account. Additionally, the toon and cosine technique take IBL into account. They also take into account other lights, not just the main light. Every time you introduce another light, change any of the intensities, change light types (spot, inf, point), apply AO, or move any light, the toon and cosine techniques produce a different distribution of red. The incidence one doesn't do as much tracking of the lighting as the other two techniques. I've also built a multi-light incidence shader that tracks any number of lights. I've also built one that uses a point-by-point recalculation of the vector-to-the-light that works much better than the single vector technique for point lights and spot lights. But in all my side-by-side tests, I felt that the cosine technique pretty well matched the results.

I think your issue with the Incidence technique is that you need to run the shader again if you move the main light. 

Actually that is a concern, but not my main issue. Since I wrote Parmatic (dynamic shader parameters), the callback thing is easy. I haven't released any of this yet, but I'm adding stuff to Matmatic to allow it to create shaders in memory, not just write a file, and Parmatic can call it. So I have the ability to create new nodes, rewire nodes, change parameters, etc when you add a light, or move the camera or move a prop and so on, in real time. 

No, my main issue is that the incidence technique, in order to take all my lights into account, including point lights close to the figure, is way more complicated than the cosine technique, while at the same time the cosine generally produces equal or even better results and runs faster. The cosine is also more predictable and requires less adjustment. Once it's set right for a given texture, it seems to do a good job in any lighting. And with only 4 easy-to-understand dials, I've been able to make it very easily adjustable.

*If you want more realism, I think /simplifying/ the SSS model is going in the wrong direction. *

Simplifying at the expense of results is wrong. However, given two systems that do the same function equaly well, I choose the simpler one.

In this case, my question wasn't about simplicity, it was "does the cosine work better than the incidence, all other things being equal?" That's why I did the comparison the way I did it. Given one lighting condition (50/50), I adjusted the techniques to produce good and nearly identical results. The first row is that. Then I varied the lighting amount only and they became different. 

My question to the forum was, which of the three has evolved in a way you expect/find pleasing. The fact is at any of those light levels, I can make all three look identicial by adjusting parameters. But if the concensus was that the cosine one was the desired result, then why bother making adjustments to the incidence or toon, or worse, try to find even more algorithmic structures to force them to behave like the cosine one.

It's funny you mention "unfair comparison." I actually started off aligning the three results with the 90/0 light combo, and then leaving all the dials alone, rendered with the 50/50. The incidence shader, adjusted to look like the cosine at 90/0, went completely berzerk at 50/50, going wildly red. I figured people (especially you) would cry foul, so I started the demo at 50/50, giving the incidence shader the best adjustment in its worst scenario. I bet with some more work I could get it to stop behaving like that, but what's the point, if cosine is good enough?

My next step was going to be to keep the light intensities the same, but move them around to produce various extreme lighting angles, and again query the forum to see which technique produces a response that is most expected/pleasing. Again, if the answer is cosine, then I'm done. If it's not, then I'm not and will keep studying the matter.

You need to be building a shader than takes into account MORE of the scene information, not LESS.

Yes, well, that was already my point. The part of the shader that produces front-side SSS needs to work for all lights, of any kind. My question to the forum was, which of these is doing that best?

Finally, of course I know that skin imperfections, soft reflections, global illumination, back-lit SSS, yada yada are necessary for realism. All that stuff is in the mix too, and when I "fire for effect" it will be there, just not for this test of front-side SSS. I've got those pretty well producing predictable, reliable, flexible results. It was this SSS that still bugged me.

all three shaders do take the lights into account.

Certainly the toon shader does NOT take into account the light direction.  I really don't know anything about your method, but from looking at the node set-up, there doesn't appear to be doing any light direction calculations.  The effect you get from a toon shader is quite similar what you get from the incidemce effect when the main light is at the camera.  The incidence effect comes into it's own when the main light moves away from the camera.  At main light angles of 90 degrees, the incidence effect is too strong, and I recommend either choosing a dimmer light closer to the camera, or turning the effect down - since as you rightly point out, it reddens the skin too much.

I've also built a multi-light incidence shader that tracks any number of lights.

Well this would certainly be an improvement on the single light incidence effect - and is the way the Blender SSS python script works.  But wouldn't you be in the same situation of having to re-run the shader when you moved a light?

Paul

Thanks for taking the time to do that render with the shader I posted. I can see a few things in it that I can improve.

However, I'm not overly concerned that it isn't as good as yours. First of all, it's not a complete shader, it was only a demo of the cosine technique for the front-side SSS effect, coupled with very straightforward diffuse and specular elements. To really compare, we'd have to include my implementation of fresnel, soft reflections, skin imperfections, etc. Also, I don't think you bothered to use any of the parameters to adjust it. But you had to have used a texture other than James, so it needs adjustment to go with the texture. I've seen you make the same point when someone asks if they can use your RealSkinShader with other textures and figures. You always point out that it was tuned for those textures and the results will degrade.

"Certainly the toon shader does NOT take into account the light direction.  I really don't know anything about your method, but from looking at the node set-up, there doesn't appear to be doing any light direction calculations. "

"Any SSS technique that does not take into account the positioning of the main light source is going to compromise realism."

You keep saying things like that to me and I don't know why because I am taking the position and intensity of the main light source (in fact all lights) into account. The diffuse node calculates the dot-product of the normal with the light vector times the intensity. It also takes into account spot cones and shadows. When I use the diffuse node as input to my algorithm, by transitive closure, all of those things will alter the outcome.

As evidence, I present the above renders. I stripped my 3 shader types down to just the SSS part and ramped it up high so you can see it. I then rendered 3 times, with different light positions. Clearly all 3 algorithms are doing pretty much the same thing.

You can argue that the contour, distribution, fadeout, whatever of the SSS effect is not right, but you can't keep arguing that I'm missing the key information of light direction and intensity for front-side SSS. The three are almost identical.

Where they totally go different is when there are multiple lights, or a spot light, a point light, or IBL in the scene. Stick a single point light next to the girls left arm and show me your shader.

This is with two infinite lights, with 45/65 intensity.

The diffuse co-mingles the info from two infinite lights and gets all muddy. In the middle, I've used a double incidence shader, and it keeps all the details straight.

Too bad poser doesn't have a parameter on the diffuse node to pick which light it is responding to.

OK, now in the render above, reduce the scene lighting by 50%, do not touch the shaders, and rerender.  Both the cosine and toon column will change (since they are relative to the total light in the scene), moving the red band into the lighter face polys, whereas the Incidence column will still accurately track poly's parallel to the camera.   The render I did with your shader is evidence of this - since the total lighting in my scene was different from yours - your shader overexposured the red - creating the glow effect.  To me, it is far easier to rerun the incidence python script (if you happen to move a light), than to have to manually adjust the cosine and toon shader light amounts.  Also, for people who do not have such a great eye for realism, they will struggle to get the red content right with your shader.

I don't think that simply increasing the red content of mid-intensity polys is a step towards further realism.

In recent experiments, I've had outstanding results from a shader that combines the incidence effect with ray-tracing, and a tiny tiny amount of toon.  The ray-tracing is providing a substancial amount of what we used to call SSS (but which is infact reflections of other parts of the body) - and the incidence/toon componenets have been turned right down.  I think methods like this are a step in the right direction.

Oooops - you got in before me :-).  My post refers to the image /above/ the post above.  And you are arguing my point.

Yay we agree.