Is this possible? I’ve been trying to use them together, but the mia_material will not go in the diffuse color of the SSS skin compound no matter what… Actually it will go if you disconnect the SSS first, then connect them, then reconnect the SSS… weird. But even then the result is completely blown out with no shades…
This is without any indirect lighting.
A better way might be to hook them up in the opposite way. ie put the SSS output into the diffuse slot of the archi shader. This way you can use the SSS node for the diffuse (and the SSS diffuse/scattering slots for your textures etc) and use all the nice archi features like blurry reflections, real fresnel, energy conservation etc on top of the SSS.
Of course you may want to disable the SSS shader’s specular so that you dont get reflections twice.
thats completely wrong, plugging something into the diffuse port will not replace that shaders diffuse calculations, it simply means the shading will be applied twice, causing everything to go too dark, doing it this way will also cancel out most of the SSS effect…
the correct way to do it would be to simply add the two shaders together using a mix2colors set to “add” with the weight set to 1.0.
note that you need to use the sss compound in v7 to do this, the actual phenomena shaders output cant be used together with any of the other shaders in the rendertree
Ah i see what you mean, thanks for the correction.
So i guess you would set the SSS diffuse to black and just use the archi diffuse node in it’s place ?
you could do that yes but I don’t see any reason to not use the SSS shaders builtin diffuse component, I’d keep that on and just use the reflective part of the architectural shader since the SSS compounds doesn’t do reflections at all.
Thanks for the help guys 
- I wondered if there was a more correct way to do this - I will try it out!
Some notes and concerns I’d like to share:
On the subject of SSS, anyone figured out what’s wrong with SSS + FG? The colors are all blown out (and setting the weights to 0 does not give black)… Someone on xsibase found that you could connect the result of the deepscatter part of the shader directly to the output, and that gets rid of the problem. It works to an extent (looks black when weights are turned to 0), but I don’t understand why it works, and why it was wired the other way in the first place…?
The xsi base thread can be found here:
http://www.xsibase.com/forum/index.php?board=6;action=display;threadid=34836
Saturn created a compound which fixes some of the issues with the default xsi7 one (lack of backscatter depth for example) One thing I found with his compound though is that setting everything to 0 weight does not give 100% black either, even with the fix mentioned earlier - it’s due to the way he rewired the diffuse weight, which wasn’t correctly wired in the default one (it overwrites the epidermal weight)… which makes me wonder… what is the correct way to wire it?
Here’s Saturn’s node:
http://community.softimage.com/downloads.php?do=file&id=122
Also, is there any way to speed up SSS with FG? the FG map generation stage takes much longer than without the sss. I tried rendering sss in a seperate pass with only direct light, and then comping it with the fg pass, but I didn’t get very good results…
I’m hoping people post their findings on sss, as it’s a complex topic, and there doesn’t seem to be a proper solution for FG yet, even though it works fine in maya and max.
A
If you are using XSI7, you are lucky. I figured out the problem. It is, that XSI (and only XSI) has a node for FG calculation, totally unnecessary, that adds to the SSS. Actually FG is considered in the lightmap, so using it absolutely unnecessary.
You have to use the SSSSkin compound I have attached. If you have any questions, just post a reply.
If you are using 6.x, try to get EMAN’s alternative skin sahder, it works properly, and fast
PS. Few other things, that for some goddamn reason, XSI has set the falloff to .2. It’s preferable to set it 2 (as in MAX or Maya), the 300 sampling is way too high, for tests even 64 is enough, and for final render you can screw it up. Other reasons of extreme render times can be, if the model has clusters. If the clusters have the same skin shader, delete the clusters, they are calculated multiple times.
Awsome. I’ve tested your compound, and am getting great results so far. The epidermal back scattering controls are really useful too. Also, The render time is much better with the sampling settings you suggested.
I have one question though, the diffuse color has an effect on the final render even with it’s weight set to 0 - how does that work? It seems like it’s because it’s connected to the lightmap generation node, but I’m not sure.
Thanks for your help!
Alex
I think that’s because FG. Since there should be something that captures the FG effect. But I’m not sure. I’m gonna check it. Good news, that I had reported it as a bug toward Softimage and after few mail, they confirmed it.
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.