Human Progress

Chris Jones

New member
Seeing as this is a perpetual work in progress and there are a few threads scattered about the place with people wanting to know more, I thought it might be an idea to bunch everything together here to make it easier to keep track of things.

The story thus far:

Displacement & texture workflow test: http://youtu.be/Nee46ySt2DQ and related thread: http://www.zbrushcentral.com/showthread.php?167603-Human-Experiments

Hand Test: http://youtu.be/DYJRVUDuZvo

Eye and Wrinkle Test: http://youtu.be/PNiuRnisK98 and related thread: http://forums.newtek.com/showthread.php?132336-Eyes-and-Wrinkles&highlight=

Some other threads documenting progress: Eye Bulge Solution ; Finger Flipping Good ; Quest for the Ultimate IK Hand Rig ; Tension Based Displacement ; Using Morphs to Drive Textures ; Metalink Morph Symmetry Issue ; Expressions - Smooth Clamp? ; Expressions Crash ; Tension Trouble

The latest test up to this point:


Now, wrinkles!

These are the two main wrinkle sculpts in Sculptris, although I also made some revisions and added platysma and neck wrinkles in separate sculpts.

View attachment 117127

They were baked to displacement maps in Blender and combined into two sets of wrinkle maps in Krita, which I've enhanced here for clarity. The maps are applied to the base mesh (which can be seen in the zbrushcentral thread above) as both bump and displacement maps.

View attachment 117128

The bump and displacement nodes have evolved somewhat since the Eye & Wrinkle test; here's what they look like now. I've added some colour coding and annotations* to the bump nodes to try and clarify what's going on. The displacement nodes do the same job using a different configuration.

View attachment 117129

View attachment 117130

And here are the actual nodes themselves... feel free to adapt or improve them. Two of the image nodes are failing to export for some reason, so they'll have to be added manually.

View attachment 117131

That's about it for now.

(c:

*Feature request: the ability to add annotations within the node editor. That way I can remember what's going on in the network, which nodes were plugged in where if I temporarily unplug them etc. In fact regarding the latter, the ability to enable/disable nodes without having to unplug them (ie allow the data to pass through unaffected) would be even better.
 

Attachments

  • wrinkle_sculpts.jpg
    wrinkle_sculpts.jpg
    95.3 KB · Views: 9,238
  • wrinkles.jpg
    wrinkles.jpg
    67.3 KB · Views: 7,426
  • bump_nodes.jpg
    bump_nodes.jpg
    625.7 KB · Views: 6,847
  • displacement_nodes.jpg
    displacement_nodes.jpg
    513.5 KB · Views: 4,801
  • wrinkle_nodes.zip
    4.8 KB · Views: 1,806
  • wrinkle_nodes.zip
    5 KB · Views: 2,047
*Feature request: the ability to add annotations within the node editor. That way I can remember what's going on in the network, which nodes were plugged in where if I temporarily unplug them etc. In fact regarding the latter, the ability to enable/disable nodes without having to unplug them (ie allow the data to pass through unaffected) would be even better.

You can add notes... http://www.youtube.com/watch?v=7BV73Y4-NDE&list=PL1C4072533A16B807&index=4&t=01m18s
 

Thank You...!!
king.gif


Awesome share!
smile.gif
need to dig more into the setup later on... ...neat! \:]
 
ugh, you gonna make learn Nodes.
now i gotta listen to RH for the next
1 million hours :sleeping:

...............what was I saying, oyea
Thanks for sharing. :beerchug:
 
A milestone in LW CA
has been mined by Mr.Jones
at the coal-face (interface).
A big up-- well done.
I hope LW3DG are jotting down
notes as we are?
 
Great information, wonderful examples and thank you so much for allowing us to take a peek under the covers, esp., with the sample files!
 
Just an eyeball. YouTube's transcoder has not taken kindly to this one - checking HD makes it marginally better though.

View attachment 118255

I say "just" an eyeball, but it was actually riddled with complications, many of them stemming from this. Ultimately I managed to get transparency to work with Sigma 2, but it necessitated the use of the Dielectric shader which caused erratic caustics behaviour and prohibited the use of SSS on the iris. I even sculpted an iris with this in mind (see below), but without SSS it looks worse than a simple image map.

View attachment 118249

Caustics also had an "explosive" effect on the illumination of the sclera the instant light touched the cornea, and for some reason it doesn't show up on the sclera at all now (though it should). The geometry itself plays a part in this, but determining the "correct" way to model the eye assembly has proven elusive.

Additionally, weight maps don't work as gradients on SSS shaded objects, so I had to use Z coordinate gradients instead; artifacts kept materialising around the sclera/cornea transition; radiosity causes blurring of the iris; refraction blur messes up caustics, and so on and so on. Suffice to say, there's more to this than meets the eye... :cool:
 

Attachments

  • iris_sculpt.jpg
    iris_sculpt.jpg
    261.7 KB · Views: 5,316
  • eyeball278.jpg
    eyeball278.jpg
    150 KB · Views: 1,966
Stunning result though despite the pain
:)
Brilliant how you're pushing the software to produce these hyper real animations. Have you reported these issues to the dev team? Looking forward to updates.
 
Last edited:
Apologies if I am completely out to lunch here, and this bears no relevance on what you are doing, but for transparency in the simpleSkin shader, if you supply a subdermis depth where the depth is greater than the depth of the geo, you get translucency. I think if you keep going with it you will get transparency. I was going to try this on a jellyfish, but never got around to it.
Basically run a (black=no depth, white=really deep) texture through DP's remap for your subdermis depth. Keep increasing the +ve scale and you will get translucent. Keep going for transparent. If I have time I will give it a go tonight & post the results.

Cheers
T
 
Progress for sure :D Thanks for keep sharing these milestones of yours Chris.

One question about the eyeball. I presume the Sclera (white part) has SSS. There were moments where I felt the surface was faceted, not smoothed. Now this could be because of two things. Skin shader is faceting, as I have seen happening (maybe mostly on double sided), or the youtube video compression is the culprit. Anyway the sensation is quite mild. I was just curious what was causing it.
Also some "striping" in the back of the eye, on top, at the start and end of the video. Surely these are parts of the model one won't see in your project. Just curious.

Cheers.
 
Last edited:
Thanks everyone!

Have you reported these issues to the dev team?

Not as such - I can't be certain it's not user error so I was hoping for confirmation first. Plus, I find the process of reporting a bit draining... so hopefully someone from dev will pop in here and save me the trouble ;)

Apologies if I am completely out to lunch here, and this bears no relevance on what you are doing, but for transparency in the simpleSkin shader, if you supply a subdermis depth where the depth is greater than the depth of the geo, you get translucency. I think if you keep going with it you will get transparency. I was going to try this on a jellyfish, but never got around to it.
Basically run a (black=no depth, white=really deep) texture through DP's remap for your subdermis depth. Keep increasing the +ve scale and you will get translucent. Keep going for transparent. If I have time I will give it a go tonight & post the results.

Cheers
T

Not out to lunch at all, although I have some doubts about whether it will work in this case, considering the object is basically a sphere with an inverted refractive shape inside it, with the geometry overlapping at the cornea. Could be worth looking into though.

I presume the Sclera (white part) has SSS. There were moments where I felt the surface was faceted, not smoothed. Now this could be because of two things. Skin shader is faceting, as I have seen happening (maybe mostly on double sided), or the youtube video compression is the culprit. Anyway the sensation is quite mild. I was just curious what was causing it.
Also some "striping" in the back of the eye, on top, at the start and end of the video. Surely these are parts of the model one won't see in your project. Just curious.

I think the faceting you're seeing is YouTube's interpretation of the the finer veins in the bump map. The "striping" is due to the UVs being squashed up towards the back to give the front as much res as possible, along with me being careless with the detailing back there. As you say, it won't be visible (unless I need to pop an eye out of a socket, in which case I'll have to go back and do it properly...).

(c:
 
Back
Top