itsjustme

I've been looking into fake SSS as well and have been thinking that Toonnation's "Villa" is something that might help. "Villa" affects transparency and reflectivity and can make an object more or less transparent at its' edges. I wish it could affect translucency instead of transparency. If we had decal control of translucency, it might also be helpful...but that is currently not available. I saw one tutorial that used translucency along with internal geometry acting as the skull...I can't find it at the moment. It had its' drawbacks, like you wouldn't be able to open the character's mouth, but looked pretty good otherwise.

Those are very high "Half extinction distances", which is providing it's own smoothing. "skin1" is (1.3, .85, .56) and "skin2" is (.95, .6, .5) in this post from Yves: http://www.hash.com/forums/index.php?showtopic=25867&p=212206 If you were to set the "Half extinction distances" to either "skin1" or "skin2", my guess is that you would find that the "Averaged normals" isn't providing any smoothing with SSS turned on (the smoothing works fine without SSS)...at least that has been my experience over the past couple of weeks.

What settings are you using on SSS, Matt? On higher settings, there is some smoothing inherent in SSS. "skin1" and "skin2" are supposed to simulate human skin. Most of the settings I've seen being used are much higher.

That may be the case...I'll try a few more things. If nothing clears it up, I'll see about faking SSS.

Yes, "Average Normals" is a surface property. When I look at Porcelain in v18, it has the same settings in its' surface properties, so I assume they are the same...although it used to be a different "Attribute" type called "Geometry" (I'm not sure when that went away). In v17g+, Porcelain and "Average Normals" gave the same results...in v18g, they were slightly different. I tried a 50% setting on the "Average Normals" and it didn't fix the problem. With "Average Normals" set to 100, the character is smooth until I apply the "skin2" SSS settings (.95, .6, .5), then it is no longer smooth (I tried v18g, v17g+ and v16b). When Jeff Bolle made his model, he used a higher SSS setting (6, 4, 2). The higher setting gives a more waxy appearance (which is fine for an the Alien character) and makes it appear smoother. The "skin1" (1.3, .85, .56) and "skin2" (.95, .6, .5) settings are supposed to be closer to human skin. I used Jeff's model so that there would be a common model that anyone could test on...it's located here (I deleted the maps for my tests to better see the changes between different settings): http://www.hash.com/forums/index.php?showtopic=46636&p=399715

I'll try it in a few other versions...haven't done that yet. I set the "Average Normals" to the same setting as was in the Porcelain material with Jeff's character..."100".

I'm not sure if anyone else has experienced this, but it seems as though SSS negates smoothing using "averaged normals" or Porcelain (same thing done as a material). Also, "average normals" and Porcelain appear to give slightly different results even though they should be the same. Anyone know how to get around this? Here are some examples (the names of the files denote the settings):

I just posted an SSS test for Bertram here: http://www.hash.com/forums/index.php?showtopic=32649&p=400464

I made some attempts to make Bertram more "toon", but I didn't like what I came up with...so, I just changed the ears, nose and eyebrows and had to redo the hair. There are a lot of things that still need smoothing and tweaking, but, here is where the SSS is currently ("skin2" with a different color). I also need to turn up the "Relative density of SSS samples" to get rid of a few artifacts. Here is a repost of the SSS settings posted by Yves: For the hair, I used the MuhHair shader...which caused problems initially with the eyebrows. I dug up Richard Harrowell's explanation of the MuhHair settings and used that to correct the problem: http://www.hash.com/forums/index.php?showtopic=26834&p=222085

Very cool, Steve!

Happy Birthday, Will!

I haven't tried that yet, Steve...but it's worth another experiment. I'll do some quick messing with it. After a few more experiments, I had something similar show up on Bertram's mouth when lighting the head from behind. Setting the shadow casting light to have ray-traced shadows with the darkness set to 100 percent got rid of it. Hope that helps, Steve.

Nosir, I used the setup that I posted with the test in this post: http://www.hash.com/forums/index.php?showtopic=33048&p=400217 It's just one klieg light.

The possible Bertram skin? It might need a slightly darker tint...I'm still tweaking it. It has nearly the same settings as "skin1" (1.3, .85, .5 instead of 1.3, .85, .56), what I think you're seeing is the blue tint on the more shallow areas along with a very stark white light. Gelatinous would be more like the higher settings.

I haven't tried that yet, Steve...but it's worth another experiment. I'll do some quick messing with it. Okay, so here is a set of test images. The first is the settings for "chicken1", "skin1" and "skin2" from Yves' numbers. The second image is using the half extinction numbers from Jeff Bolle's "Batgirl" image here: http://www.hash.com/forums/index.php?showtopic=46491 The third image is using the half extinction numbers from Steve's "Buying A Coffee Shop" thread here: http://www.hash.com/forums/index.php?showtopic=44044 The fourth image is the possible Bertram skin...which is pretty much the "skin1" setting with a couple of minor changes. I didn't get any hot spots in these tests, so I'm thinking it might be something to do with the lighting where they appear...just a guess though. Hope that helps, Steve.

I haven't tried that yet, Steve...but it's worth another experiment. I'll do some quick messing with it.

I worked on some lighting on the smoking area a couple of weeks ago starting here: http://www.hash.com/forums/index.php?showtopic=38232&p=399990 I put a version of Bertram in the set and decided I needed to do some tweaking to him before I could adequately evaluate the lighting. I tried several things in order to "toon" Bertram, but ended up just updating his nose and ears (which made me have to re-do his hair)...I didn't like the other changes very much. After that, I started working on his skin...which led me to an exploration of SSS. In order to get a better handle on SSS, i did some tests using the extinction distances with color values posted by Yves here: http://www.hash.com/forums/index.php?showtopic=25867&p=212206 I have attached the test files and renders in the "SSS_TESTS.zip" in this post. I made my test box extremely small so that I could see the color changes, which also meant that I had to turn up the "Relative density" to get rid of artifacts. The numbers in the name of the renders is the "Relative density" setting. From these tests, I think I got a skin setting for Bertram that might work (image is attached). I'll post an image of Bertram with the settings applied once I get his hair finished. The SSS settings are "skin1" with some minor tweaks. SSS_TESTS.zip

Have you tried updating your graphics card drivers? At the very least, it couldn't hurt and might allow you to use OpenGL3.

Great stuff, Mark!

Looks great, Chris!

Thanks!

I got a great grin out of this one, John. Very cool.

I don't remember the render times for each AO setting, but (using 16 passes), AO at 70 percent quality with the three sun lights was about 1 hour and 45 minutes (of course, the time gets faster as you reduce the quality), with the 17 bulb light rig and the three suns was close to 1 hour and just the three sun lights without any ambient light was about 18 minutes. I'm not too concerned about the render times since I'll be compositing Bertram into the scene. Also, my render times are pretty high due to the underpowered laptop I'm currently using.

I think I didn't write that very clearly...the test images start at 30 percent and go to 70 percent in increments of 5 (I'll tweak the original wording).

I rendered the smoking area with AO only with a white Ambiance Color to better see the AO grain. The quality settings (for these test images) go from 30 percent to 70 percent in increments of 5. If I were to stay with AO, the lowest setting I like is 65, but I still prefer 70. I've added the renders in the attached ZIP file as individual images and as layers in a GIMP XCF file for comparison (you can just hide layers in GIMP to A/B each setting). I then set up a 17 light hemisphere of blue bulbs to replace the AO and the render time was reduced considerably. I also reduced the saturation of the yellow sun light and put the present version of Bertram in the shade to see what it would look like...attached is that test as well. There is still a lot to work on...it needs bounce lights, possibly a couple of fill lights, etc. I'll also need to finish Bertram so that I can see how he will really look in a closeup in the shade (texturing will help). So, I'll have to start making some decisions about how the final Bertram will look. AO_quality_percentage_tests_10_16_2014.zip