Jump to content

Skin SSS Point Cloud


Overload

Recommended Posts

Could someone point me in the right direction to get the Skin SSS shader working? I found older posts on SSS, but none on how to bake out the point cloud needed for this new shader. Does the PC need to be calculated based on the camera? Or maybe its something in the ROP that I'm missing. Thanks!

Anthony

Link to comment
Share on other sites

The shader on which the SkinSSS material is based was written before pcwrite() became available, so yes, the material itself doesn't generate the pointclouds (the way the pyro shader does, for example), instead, they need to be generated at the SOP level using the SSS Scater SOP (wow, 5 'S'es). A historical limitation with all sorts of fun implications.

Here's an example file:

testSSS.hip

Go into the sphere object and click the Render button in the "GeneratePointcloud" SOP to create the point cloud. Then, every time you change the "Scattering Distance" parameter in the example file's shader, you should regenerate the point cloud.

In building this little demo file, I noticed there are a number of "funky" things going on with this material's treatment of the sss component... I think it may be in need of some lovin'. Anyway... yeah, as they say, "YMMV" with this thing.

Link to comment
Share on other sites

Ahh sweet. Thanks! Now it works. I didn't realize that SOP existed. Just curious, not that the PC takes long to write, but how come you need to re-render that when you update the Scatter Distance? Also, in the sampling of the shader, should that be set to the exact number of points in your PC file? Thanks again!

Anthony

Link to comment
Share on other sites

Just curious, not that the PC takes long to write, but how come you need to re-render that when you update the Scatter Distance? Also, in the sampling of the shader, should that be set to the exact number of points in your PC file?

Only because in the example file I attached in my previous post, I tied the shader's scattering distance to the number of points in the cloud. This is just for efficiency (long scattering distances don't need as many points as short scattering distances), but it is not a requirement.

It's perfectly OK to have a constant number of points, independent of the shader's scattering distance, as long as there are sufficient points to properly evaluate what you anticipate will be the smallest setting of the shader's scattering distance parameter. However, if you overkill the number of points too much, the renders won't be as efficient as they could be (they will take longer than necessary for comparable results).

The sampling in the shader doesn't have to match the number of points in the cloud -- it's somewhat proportional, but not directly related. The default of 50 (and up to 120 to 200) should work in most cases, unless you have a huge number of points and a very short scattering distance, in which case it can/should go higher.

But again... while playing with this material a little now, I find, among other things, that the sss component is unstable (render the same frame with the same settings twice and you get two slightly different results; i.e: it "flickers" with constant parameterization). Something's not feeling quite right in there, so "Caveat Emptor".

Link to comment
Share on other sites

Awesome, thank you Mario. This makes much more sense now. Yeah, the flickering is kind of odd. Figured I was doing something wrong, but without changing anything you do notice a fairly big difference even on the same frame. Is there anyway to blur or smooth that so its not so drastic? I would think these results would stay consistent if your not re-baking the PC file.

Link to comment
Share on other sites

Is there anyway to blur or smooth that so its not so drastic?

Yes, by debugging the material, :)

...which means an RFE or BUG report to SESI.

I have no personal involvement with that material whatsoever, other than very indirectly through some R&D work I did five years ago (5!... sheesh! time flies), so I haven't a clue about what might be going on in there.

Mmmm... debugging VOP spaghetti... yum!

Link to comment
Share on other sites

haha yeah, it's one big spaghetti network thats for sure. Oh well, I might render the sequence like this, just to get some abstract renders for fun. Sucks that the SSS is for the most part useless if you need more than a still frame. Looks nice when it cooperates.

Link to comment
Share on other sites

Hmm, I'll give it a shot. We already rendered the spot in Vray, but for my own personal enjoyment I wanted to see what Mantra could do with the same shot.

Hey guys,

you should try this SSS VOP from Houdini's exchange:

http://www.sidefx.com/index.php?option=com_wrapper&Itemid=146

I believe you will still have to generate the pc at SOPs but, this shader gives nice results.

Best,

Ronen.

Edited by RonensArt
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...