Emily

Started by Hannes, November 03, 2019, 04:28:28 AM

Previous topic - Next topic

j meyer

Definitely improved. :)

WAS

I was wondering why her skin always looked off to me, but i think I see way now. I don't know if it's just me, but it appears all her pours are inverted. Lots of "light" circles rather than dark circles for divots.

Hannes

First, here is an an updated version with higher volume density and a bit of higher contrast (inside the TG render settings - no postwork). The nostrils look better now, I'd say.

Hannes

Jordan, you made me think about the bump.
I did a test with negative bump, and I think the first version (positve) is the correct one. Sometimes there's a certain convex bumpiness (sort of small pimples) in some areas, especially on the forehead. The negative bump looks a bit strange, especially on the lips and nose. The difference is subtle, but I'll stick to the initial version.

Dune

Difference is very hard to see. What about the curved banding in the lower neck? Are they jpg artifacts or is it mesh?

Hannes

Unfortunately it's mesh. Before I used the Emily head model I loaded it into Poseray, recalculated normals and also subdivided the mesh. It's still there, but a bit finer than before this procedure.
I'm not sure, but this problem has been mentioned as far as I remember, but I couldn't find this thread, and I don't know, if there was a solution for it.

Hannes

Quote from: WAS on November 05, 2019, 01:44:21 PMbut it appears all her pours are inverted
At the moment I'm rendering another model, and I converted the included normal map to a displacement map (using Crazy bump and for testing Materialize as well). Unfortunately in some of the areas it seems both apps read the normal map's information wrong, so there are some parts that seem to be inverted. Maybe this happened for Emily's maps as well.

For my actual model I edited the created map in PS, and it hopefully will work.
I'd love to see TG supporting normal maps.

WAS

Quote from: Hannes on November 06, 2019, 08:17:28 AMJordan, you made me think about the bump.
I did a test with negative bump, and I think the first version (positve) is the correct one. Sometimes there's a certain convex bumpiness (sort of small pimples) in some areas, especially on the forehead. The negative bump looks a bit strange, especially on the lips and nose. The difference is subtle, but I'll stick to the initial version.

Though chicken skin is a dermatological condition, and not normal. You shouldn't see raised pores on the face. o.o When you invert it the skin highlights look realistic, even on the lips there is more lineage than splatter, which gives actual stretched skin appearance.

I think the inverted does look better, you should just think about adding actual skin imperfects than using inverted skin pores as imperfections. Small bumps for blackheads and whiteheads here and there, not all over.

RogueNZ

Have to agree, inverted looks better in my opinion!

WAS

I also noticed, leaving the image open on my monitor and taking a step back cooking at about 15 feet, the overall highlights seem more photorealistic in the inverted bump, than the left non-inverted, which seems more soft and 3D ish with the highlights being not as defined/bright?.

DocCharly65

I needed a long time watching and doing some speps back from the monitor to see it too, but in the end I think the negative bump looks more realistic.

Oshyan

The thing that I see is negative bump looks better everywhere except the nose. :D So maybe indeed some parts got inverted from the original conversion and some did not?

- Oshyan