Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.Daz 3D is part of
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
Just following directions Andy. My lighting is the same as it was.
I'm assuming you mean Translucency Weight to zero?
this is close to impossible ...
post a quick render like before where i can see the volume settings on your screen ( hidden from the progress windows before)...
OK, with SSS to zero and Translucency weight to zero I'm getting nicer results but the skin needs to be a bit darker as it's tanned. So at least I'm on the right track finally!
Yes but we dont want zero.....JUST for a testrender of the base - let me see a render....
Your results before are not normaly - nothing changed, that cant be...so we start from looking on the base...
Stopped the render at 94%.... close enough.... See the blackhead/zit thing on the tip of her nose? I think it was from that SSS map I made. I guess I'll have to go in and smooth out some of those areas with the smudge or blur brush...
with sss map you mean bump or normal map?.... The render shows now only base and bump/normal.. (step by step.....
looks OK to me.... now slowly turn in ONLY sss amount - set it to just 0.5 first.. AND translucency weight to 0.4 - do a render.... just some iterations! and upload...
OK... here is a render. Very very red. Wait till you see the settings.... very low...
Wolfie, try this. Go to Image Editor on your diffuse map, change the gamma to something like 1.3. I had an interesting problem along these lines and that cleared it up though it's really a stopgap solution.
EDIT: Good Lord, why do I have a random name now?
this is definitly odd!
1. close and restart daz!
maybe even reboot your system..
2. still the same problem....
remove one map - do a test render..- add the map agin - remove the next... and so on.. maybe you can identify the reason that way....
3. install daz 4.9 -> this reseted some dlls which are shared in both versions.(aka 4.8 suddenly worked also again).. solved a strange render behavior with ligths in my case some weeks ago....
Yea, I think this will probably be resolved when 4.9 is released. It's pretty sucky. If I have those two thing turned down to zero I get a pretty nice render so that's telling me something is up for sure.
Hello all, this is what I have been working with so far. My first attempt was to try to get something that approximated skin without using any maps at all. The settings are shown below.
I'm doing test renders with Redspec's free HDR pack because I like the lighting setups. They're nice and clean with a lot of neutrals to choose from, so no worrying about color. The first is 01, rotated 270 degrees. The second is 08 with no rotation.
Not perfect, but I liked how it looked pretty well so I set base color and translucency color to pure white, then plugged in the G3F default diffuse map. This caused a horrible looking sunburned color I haven't saved an example of. Funny, because the diffuse map average color is almost identical to the color I had plugged into the base originally. Reverting to the previous shader setup, I plugged in a flat image file of that same color into the diffuse channel, and it also came out horribly sunburned though it should have been identical to the previous images. DS is handling gamma (at least I think it's gamma) differently between image maps and plain numerical color input. Why? It's a little irritating.
Changing the gamma of the diffuse map to 1.3 in Image Editor removed the sunburn. This was a quick and dirty fix; these would be better off adjusted individually in an editor. The eyebrows in particular suffer from being lightened more than they should. I think this is a halfway decent start, though. Loving the response to rim light. Still no bump or specularity maps.
This setup also looks fairly nice with the translucency color left as red with no maps, except that places like eyebrows of course aren't masked at all. Plugging a simple black/white eyebrow mask into translucency and transmittance caused the skin to glow much brighter red in addition to masking the eyebrows, so DS is handling images wonky there too. For some reason straight white multiplies the effect of the color in the channel. Very very weird. I haven't tried to work around it yet.
Been meaning to post this screen grab. Here are the 5 maps. Hopefully you can read what they are. Maybe I can tune up one or two of them to help improve things but I really think it's a bug in DS that's partially to blame for this craziness...
did you install /deinstall other graphic software - i did ..deinstalled 3d coat and other trials.. and this messed up daz 4.8...
cant help more right now....
you can install the beta... dont worry... you must not use it - 4.8 will be still there.
I like what your doing here. Looks awesome so far. I hear ya about the diffuse map but I'd rather look into what's causing this issue rather than desaturate all my hard work, ya know what I mean?
RAMWolff
Glossiness map is clearly wrong - looks like a roughness map... nose should be brightest - INVERT this map in photoshop. but that has nothing to do with your problem..
Specularmap you dont need! but looks ok expect the eyebrows... you dont want them metal set them to a grey (way darker)...
SSS you use in translucence and set a base color dark brown.. that should work....
base should work...
Thanks! I like how you're coming along with the texture too! If you do try the fix, it won't do anything to change your work, don't worry. Going through Image Editor in DS will just tell DS to use a different gamma. All your textures will remain the same. This might be an issue with the iray shader expecting a lighter image in the diffuse map.
EDIT: I hear you on waiting for a fix though. There is definitely something very strange going on.
@recursivelement_922de179e2
yes i noted some strange map behavior too in my test on 4.8 some weeks ago.....
i use now only the beta... and did not check out everything again yet.... problems with to much red are gone in 4.9.... so always post which version you used for a testrender with strange results..
Good point. I'm using the latest beta, 4.9.
Yeah, in 4.8 if you used scatter+transmit for translucency you couldn't plug a map into translucency weight. Its been fixed in 4.9 I think
Surely doesn't look like it to me. Will post broken pics tomorrow.
Here's Victoria forme in 4.9 with skin settings using scatter+ transmit and a simple black and white cross texture plugged into translucency weight center was white and is matching the transluceny of the rest of the skin sides were black and have no translucency
For comparison here's the same setupin 4.8 with victoria looking like someone ripped her face off.
Also just want to say... Holy crap you got the g3f defalt texture to look so nice! You should probably be sainted, because I think that counts as a miracle
Although it looks pleasing, you have some glitches build in: From an artist's point of view, you can still say "I give a ... (whatever you want to put here)", if you want to get physically based shading right, you can't say that. How you want it to be is always up to you.
Investigation: Why having a monochrome Transmitted Color and a dark Translucency Color is a sub-optimal™ idea.
A few pages ago we discovered that the "Translucency Color" is a Multiplicator (Modifier) for our Transmission. The idea behind having a modifier, which is of optimal use when designed as a variable, is to modify a constant, a thing we either don't want to or simply are unable to alter; like the speed of light, f.e.
When using a light (or monochrome) "Translucency Color" in combination with a dark "Transmitted Color" (avoiding values of or going against "0.00" which we found out is something really Bad™), we will have another instrument of control over our Transmission at our hands with the "Transmitted Measurement Distance" (TMD) parameter. Raising or lowering it's value will have a clearly noticeable impact on the final color of the Transmission. (See screenshots SSS2 to SSS4 below, for presentational purposes the "Translucency Weight" is set to 1.0 to show the Transmission at it's full strength).
Now, having a dark "Translucency Color" in combination with a light (or monochrome) "Transmitted Color", we will almost loose that instrument, raising or lowering has a very smaller impact, and at a certain point, changing the TMD won't have any further noticeable impact. (Screenshots SSS5 to SSS7).
"Transmitted Color" in the shader represents the color of our skin, caused by melanin and haemoglobin absorption effects within the skin tissues, a higher melanin concentration will make a darker skin color, and a lower one a lighter skin color. "Translucency Color" in combination with the "Translucency Weight"parameter just determines how translucent a material is.
Under that point of view, using a variable as a constant and the actual constant as a variable doesn't make any plausible sense. As mentioned above, from an artistical point of view you can do anything that suits your taste or needs. But Physically Based Shading it then isn't anymore.
Thank you for your interest, and have a nice day.
Thanks! And thanks for the comparison. My problem in both cases is not the same as what was showing up in 4.8, although it looks to be along pretty similar "DS handles image maps wrong" lines.
First set of renders is comparing the "too red" issue in 4.8 and 4.9. Rendering the non-image-based setup looks exactly (exactly!) the same in both, so changes and issues are solely in the handling of texture maps.
4.9 has a new and exciting but still map-related issue. This is what happens when you take a shader setup, blank the diffuse color, and plug in an image map which is exactly, or in the case of the full texture, very close to the same color. Red, red everywhere. In preview, they match without gamma correction.
Finally, plugging a strength map into translucency strength works okay (yay). But plugging them into transmission color or translucency color does this. The difference here is literally just plugging a flat white map into transmission color and translucency color while leaving the colors already in the channel as they were. This should be changing absolutely nothing about the final render. White should change nothing. Instead, it causes the color to explode.
Radioactive, radioactive...
Maybe there's a great reason for image maps working in a completely different way from colors, but I don't see it. Looks amazingly broken to me.
Excellent analisis.
There's a potential problem with calculating specularity like this. Human skin has a specular which should correspond with the refraction of the volume. But, it also has layers of oils on top, which do not correspond with that refraction index. This might perhaps be best accounted for by using "skin" specular values in the base, and then adjusting for oiliness with a topcoat. However, this gives up the ability to use that topcoat for additional effects like wetness. Often, specular of a shiny coat on top of a relatively diffuse coat completely overpowers the base coat - example, paint a car flat black, then topcoat with clear gloss coat, it is now glossy with minimal effect of diffusion from the base coat. This is because the top coat smoothes the roughness and dominates the specularity. I will have to look into whether accounting for skin + oil separately is worthwhile.
As you've noted, I haven't actually turned on refraction yet.
Will get to the other post later; you have some excellent points.
Here are my latest tests using a greyish diffuse and a fairly saturated diffuse map for the translucency color on a more pale skin. The bump map is, of course, heavily modified. I created a new pores overlay that introduced a bit of variation to the pores in size and depth. I smoothed the tip of the nose, and the slightly smoothed the chin area. I also dodged out the typical dark areas on most bump maps around the bridge of the nose and eyes. I then manually drew the thin wrinkles directly under the lower eyelid back in. For the specularity map, I did the regular T-zone adjustments, though I need to brighten them up a little more, I think.
I'm still toying around with the negative vs. positive setting for the SSS direction. There are some aspects of using a negative direction over a positive I like. For example, I'm not thrilled with the overall look when not using a geometry shell inside and a positive direction. In the tests below, I didn't modify anything other than the SSS direction, and whether the geoshell was visible. The bluish cast seen on the test with the positive SSSD without the shell is pretty difficult to get rid of without mucking up the overall tone.
When using the negative SSSD, I think the translucency around the ears is much more appropriate to the lighting I was using (basic 3-point with a somewhat strong back light).
Thoughts and critiques welcomed
Attached images are the same except for image size and zoom.
(Please excuse the aliasing around the edges of each head, the latest updates to PS CC2015 are not agreeing with my system)
(Can someone tell me how to insert an 800px image directly into the body of a post? I tried drag and drop, and copy/paste, but couldn't get it to work)
Kurzon,
Did you desaturate the map in photoshop or did you use the channel controls in DS? Also, are these general skin settings available somewhere so that I can try them out for myself? I like the way this looks so far.
I actually like the middle one the best. While indeed there is probably just too much overal translucency causing the nose to glow a little brighter than expected, I do not see much benefit gained from the geoshell version. Not that there is no advantage at all, just I lm not sure the memory overhead is worth it. As I stated before, the green/blue aspects of skin are important and the geoshell seems to reduce that effect. Still, wonderful skin effect overall!!