Daz 3D is part of 
Connect
DAZ Productions, Inc.
7533 S Center View Ct #4664
West Jordan, UT 84084
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2025 Daz Productions Inc. All Rights Reserved.
Comments
Since Studio's base unit is cm, one would assume that it is in cm, too...
""What unit is used in "distancWhat unit is used in "distance" tabs? Scatter measurement distance etc...mm, cm, m , inches?"
... SSS amount / SSS distance = coefficient..... very confusing i known 
.... Arnold can explain that very well 
Nothing at all
i use mostly 0.1 distance and 1 sss amount = 10
jcade as far as i remember 0.07 and 0.75 amount = 10.7
but you could also set 0.05 and 0.5 = 10....
that's why many different settings have a good (similar) result
You're welcome.
Experimented with the "Fabric - Silk - Blue" example a bit, applied on a simple cylinder primitive (20x2 cm) using an AndySpot (15000 Lumen, Temperature 5000 (K), Geometry Disc 10cm (ringflash)). That light created a blueish tinted highlight. Turning the Temperature to 2900 (K) yellow/orange color (similar sunrise/sunset) the material appeared greenish, but the highlight, too. That seemed wrong, observing the reflections from different colored lights on a silk bedware sheet I own. There the reflections repeat the color of the light shined on.
So I set "Share Glossy Inputs" to Off and put the color from "Glossy Color" into "Backscattering Color". Then I turned the "Glossy Color" and "Top Coat Color" back to white (if it's too strong, you can lower it to 240, still white).
Using the 2900 (K), the material still appeared in a greenish tint, but the reflection yellowish/orange. Switching to 5000 (K) showed the deafult blue material color, the reflection appears white. That looks more than it should be.
You may try to put the "Glossy Color" you found in your hair shader product into "Backscattering Color" instead. If you're on a PBR Specular/Color Mix, as I guess, you also may put the same color you use in your "Specular Color" slot into "Top Coat Color" (Might also be the more correct solution to repeat it there, since Top Coat hasn't a seperate Glossy Color on it's own).
yes... the problem is that we can not control WHERE the "cheated" secondary specular is... backscattering drives light in shadow areas...
anisotropy could in theory beeing controlled by a UV map.. and create a shine following the fiber direction.. ( a lot to test )
Or you could start to implement all your theory and build us a nice hairshader
That's correct. Measurement Distance scale is in Studio's cm world scene scale. That's why you've to adjust the distances in an imported MDL. NVIDIA uses a meter scale there as default.
If you could afford it... I'd charge an amount similar to a "Commercial Game Developer License".
Reading though the Siggraph 2003 doc, there it's mentioned that hair would have some kind of backscattering build in. Must be from the same madman that created human skin...
ok: results.... backscattering dosent work on fiber... it follows edges in geometry,,, and a fiberhair with 3 profiles has edges everywhere..
But using colored specular AND glossines color.. then anisotropy with full amount (1).... and top coat with correct IOR and white glossiness color... base/top coat weigthed with 0.3/0.7,,, looks promising...
Anisotrpy smears the colored highlights/specular in the fiber direction.. (i am using a UV now)...
the colors are overdone for testing... i think that looks promising - zero texture...
Fibremesh and similar are still way too slow and thick for me for anything other than the most basic or short hair. Personally I'm a big fan of using geometry shells + transparency maps. You can get a good ammount of depth that way with a lot less overhead
(one thing that I do think is true for both strand and traditional hair shaders they really do need some sort of transparency, especially for lighter hair, blonde hair is lighter in part because it is finer than darker hair tends to be)
The render below has 2 geometry shells on it. All three layers have a cutout opacity of about .75 or so (excluding the scalp)
my tests convinced me otherwise... they render relatively fast in Iray.. the fiberhairs which i bought last year in the store.. use to thick hairs to keep the polycount low.... i was surpised to see that i can render 2 mio polygons up in Iray now... what slows down is the viewport opengl.. handling in DAZ is the problem...
The hairs above use it all... translucency, sss, refraction... and they render on my weak card reasonable.. that's what made me look again into fibermesh for DAZ...
Your render looks good... nice hairs and method...
The reason why i test out the limit of fiber is.. backlight on hairs.. and flying away hairs (the glow of single hairs around the main strands)... plus relatively easy to create morphs.. wet hairs (clumping)... action poses..(flying hairs).. and so on.
here two images rendered exactly for 5 minutes on my poor 96 cudas 1GB...
a closeup to give you a idea about the thickness i am using (stil a little bit to thick.. but using thinner hairs for the hairline is always an option)...
the polycount in the geometry editor is 1.7 Mio.... plus a fully textured V7 is also there...
No argument on the render speeds, because of the lack of maps they can be relatively fast. And I do think they can be viable for short hairstyles. But those hair strands are still way to thick and too few. While I don't think its always completely necessary to go fully accurate 100,000+ I think you'd still probably need at least twice as many strands and have them be at least half as thick. (although you could probably get away with less strands by making covered areas emit less strands). As it stands currently it doesn't really look any more realistic than modeled hair to me, just differently unrealistic
For reference when using blender's strand renderer I usually use around 60,000-80,000 strands, (thankfully blender is is really good about child strands so the memory hit is actually really good)
well... i am testing mainly shaders on fibers right now... (and thickness) ... that's why i use a half finish hairstyle for testing. for sure she must go first in the hairsaloon to look great
Clear to me is that it wont render fully on 1GB when i am finish.... But it have not do do that because from the 100k hairs usally just 20-50k are to see ... i see more problems with handling a long hairstyle in DAZ then with Iray.. and that realism is better with fibers, for that are many great reference renders in the net.
@j cade -> Did you render one of your blender fiber hairstyles in DAZ Iray? would be great to see one.. Also i still struggle with a shader setting for blond on fiber...
There is a lot of good strand hair renders out there. I'd argue with you on them having only less than 50k strands although if you show some examples I'd be happy to be proven wrong.
I've done some blonde strand renders in blender cycles, but nothing in Iray yet. I've been doing a lot of blender iray compositing actually. Its easier to setup the skin and such in Iray, but blender's hair system and shaders are much easier
there is no argument about the number.. i think about 80k up are needed for a good looking hairstyle.. what i meant above is.. that there is never a view which shows all of them...from the 1.7 mio polygons i use for testing right now. are most on the back as example.... no hair on top at all...
So - i will split my hairstyle in different props and layers... switching on and off to be able to handle it... we will see.. it is a challenge for me.. learning experience.
I am a little confused.
If it is in cm then I would assume I should use a real life measurement.
If it is a coefficient then I assume that there is some correlation to a physical thing...is that where "10" comes in?
In a physically based workflow I would think that "measurement" refers to a real world unit.
@j_cade I was going back to your gallery (stunning) and I am curious about the "gianni´s back" image. Did you use bump and other maps that came with or have you manipulated them or even authored your own. I have never really been able to create such lovely highlights yet(Strong, but broken up) and bump and gloss is next on my agenda. My specular highlights are either too diffused or smooth and plastic. Do I need to go into top coat, maybe...I have so far not gone there for some reason.
Seems some of the wet body products out there use Geo shell...
Either a real-world measurement or something close to it. When using a real-world measurement, most are given in a /cm^-1 scale, but there are also some in a mm^-1, the latter need to be converted in cm^-1. Just divide by 10.
I wrote here about my investigations about the volume properties group of the shader, including the formulas of how to calculate the appropriate color values from a set of known coefficients. That solution needs some valid measured data and some math juggling.
There is a more easier solution: just pick a RGB color value directly from a Diffuse texture of your character in dependence of luminosity, like explained in MEC4D's posting in the "Show Us Your Iray Renders. Part II" thread. Put the picked RGB color values into "Transmitted Color", set "Transmitted Measurement Distance" to 1.0... done. "SSS Amount" and "Scattering Measurement Distance" are a bit tricky, "SSS Amount" uses a monochrome color, where in fact it would needed to be in full color like the "Transmitted Color". So it can't be set to be anything near a more physically correct set of parameters.
The "10" in Andy's example is the scattering coefficient he uses, divided by a "Scattering Measurement Distance" of 0.1 you'll get a "SSS Amount" of 1.0.
I use the topcoat set at a weight around .15 with a tiling bump (the top coat lets you use a secondary bump map and I love it) I use a custom skin detail map now based on one of the downloads here (I didnt do that much really, just made them tiling), but in the past I've gotten away with just using a 1k map just using the generate noise tool in photoshop. Iray is great because you can tile individual textures seperately so I then tile my top coat bump to something obscene like 22 and boom, nice broken up glossyness
Just for the record....
It's possible to render 80k hairs with profile3 and 30 segments on a 1gb card... (plus a V7) (7.2 Mio polycount for the hairs)
and as expected the problem is the viewport in DAZ and not Iray... the viewport is still "workable" but lags extrem....
Render speed is reasonable... 500 iterations on 96 cudas about 3 hours -> on a GTX 780/970 up it would take maybe 1 hour for 1000 iterations...(or faster)
preparing for the 1 iteration on the GPU card was about 6 minutes - i was sure the system will crash .. but it did not
Is there a way to speed up the viewport in DAZ ? How much does a good graphic card improve the normal viewport?
Thanks for explaining Arnold.
this ^ makes it so much more understandable what is happening. Now I just have to read the rest of that thread
J cade, ah I see. I have been using a tiled micromap (i think it was andy that posted it a while back). It has been great, but I have not tested it in top coat yet.
@Kaboom
The reason why after many tests i settled with a SC of 10 for skin .. is a compromise between theory and results...
the problem with using measured values (real world SC of 16 (or higher) for a red wavelenght as example) is that small details getting washed out to much... Our used textures have inbaked SSS allready!
RamWolff.
I'm the author of the "Spanish" tutorial. You can try my image tutorials in my gallery: http://www.daz3d.com/gallery/#users/37283/
Also you can find interesting information in http://www.scratchapixel.com/
And of course, in this post, that is very interesting!
I am going to take away all the diffuse and specular and bump maps from a character preset I made and try only the PBR shaders. That's what they are intended for right? Your sclera has the same specular properties and the shine depends on the light, same with the cornea over the iris and the skin throughout although I will have to maybe make layers for melanin concentrations and types of melanin and so forth.
What is the best turtorial for this? I must go back and read this entire thread. It seems the easier is to walk before running so a simplified albino skin should be the first type I try to model.
I am so slow, maybe MECD will have her product out first and save me much work...although now I am curious because I will always have need to know how to customize with PBR.
However, it seems to me if PBR means anything then once you have the layers and masks set up to create standard human PBR shading with the ability to vary the type and amount of melanin then the model is mostly homefree. You'd need to only need to intoduce masks and gradient masks for things like tan lines and freckles.
The wrinkles you'd want the deep ones as morphs and prably want to ignore the shallow ones, at least with the state of the models and technology as it is now. The pore and hair follicles are mostly the same but vary between sexes and pre-pubuescent ages most notably, but you could have settings for those to vary their denisity and size and have them on different layers to take such factors into account. And then there is the matter of the super fine wrinkles we don't really notice that people have once they grow beyond infant size on the arms and legs, most noticeably under magnification.
++++++
OK, an update I'll post this David Bowie-ish looking diffuse textureless & mapless PBR shader render.
So I created a material preset where I've loaded the Genesis 2 Male iRay optimized textures and then removed all the diffuse textures and various maps, except the eyelashes (as a reminder to do things step by step) and then changed the red of the skin PBRs to a peachy pink that I think matches the pictures of the Van Winter brothers or the Russian model & her brother and then rendered in iRay.
Just that alone surprised me at how much better the lighting looked compared to the same scene using the Logan textures. True the model looks cartoonish or plastic but one can't dispute the lighting looks much more natural on the model.
That's just to get to get me started. I am going to look up a PBR reference site and try to build the skin back to a more natural look but one that does not use diffuse textures or specular maps that introduce highlights that are appropriate for only one lighting situation. I am not sure if it is possible from me to buy generic PBR mappings of traits of human skins for different regions of the body that have been created with the nice neutral diffuse light needed to avoid PBR texture exaggerations or not. If not I'd have to hand create such textures and LOL, will likely decide to proceed no further than a cartoon PBR preset then.
I will probably post again in the next few days to this same thread after a re-render when I add in all the color values and settings as the PBR reference I will be using are added to the material preset dictates, but then it will be a long gap before I post further results as I either make or obtain various PBR maps to use with the material preset that give the render a correctly lit render. I'm a complete texuring newby outside of simple cartoon textures I've done for some simple models I created in Blender.
+++++
OK, second update. I've added color that I think is a good match for PBR settings for albinos to the iris, pupils, lips, and nipples and except for the diffuse color I added to the skin everything else is still the same defaults except I stripped the textures and maps away. You can see that the lips are way to glossy and specular. That shouldn't be a surprise as I'm still using the DAZ PBR defaults for the Genesis 2 Male iRay Optimized settings and they do the lips up fashion model style, male or female, so I'll need to adjust those.
One advantage without the textures and maps with this PBR shading is I can see right away that the mouth is trying to surpress smile and that has to be edited out. Also the lips are fashion model female style lips and that geometry has to be edited to match the ID. Can I make those little geometry corrections directly in DAZ Studio Pro 4.9? I don't know but that's my next lesson.
If you look at the lighting differences in the two pictures it is because the 1st is done using the DOF Closeup Camera and the 2nd the Perspective View. What's similar between the two is the Portrait Golden Ratio and the perceived distance from the lens. For some reason when I save a scene with a camera the camera gets stuck with a viewport size equal to the 'Portrait Golden Ratio' maximized for the Viewport I'm using rather than how I originally set it up - which is the Portrait Golden Ratio maximized to fit in the UHD dimension settings.
If DAZ Studio Pro 4.9 won't let me edit the mistaken lips geometry directly in DAZ Studio I'll save that for last otherwise I'll do that next. Then will come the base skin bumps. I think, but ain't sure yet that I will not be creating glossy or specular channel maps and stick with the mapless defaults but it's a matter of one step at a time and add what is needed that the PBR shaders aren't already sufficient for. Actually the 1st maps I'll do are the irises and then I will try making eye presets using opaque texture layers in LIE in the Metallic Flakes Layer to represent melanin in the eyes (I'm guessing how it's done here).
+++++
@Kaboom You're welcome.
Edit/Preferences (or shortcut F2). In the "Interface" tab set "Texture Resources" all the way left to "Performance". That can speed up DS's viewport a bit. Check if "Display Optimization" is set to "Best". To speed up a bit more, you can set "Hardware Anti-Aliasing" to "Off".
I can't wait to get HW that can have an iRay viewport...
Just saw this post. Thanks so much. I found you on Deviant through another thread here and downloaded your image tutorials to study off line when I have more time. Thanks so much!
Yeah, those will be useful to check my work when I am done.
I'm using the Really Short Haircut Freebie but I changed the diffuse color to instead be a texture from the Real Short Hair cut product I bought from the DAZ Store. I also added the glossy (specular) map from the same product.
I'm not as pleased with the eyebrows. I did the same thing as the hair but you see the eyebrows are splotchy for some reason. I quickly eliminated the black eyebrows on the diffuse texture in Gimp and you can tell. However, I will later go back and to a better job.
The hair & the eyebrows are both freebies from Laticis Imagery.
The eyes are Nevio 6 eyes. The skin textures are from Ryder. I feel the skin textures are too red. They've had the Genesis 2 Male iRay Optimized Shaders Setting applied. The hair & eyebrows have the IRay Ueberbase setting applied. The lighting is from the iRay default environment.
So my question is how can I fix the splotchy eyebrows to be as good as the hair? Since the diffuse textures and the maps come from Ryder is it possible to make them less red?
Hmmm...I rendered again & upload as I though 95% converged left too many unconverged pixels in the shadows. I also rotated the HDR Dome to 120 degrees.
I been trying to follow this thread and i am still scratching my head. Can someone in laymens terms tell me what happens when you convert skin to iray shaders. To my eye it really does a lot of things but especially to skin tone and fine skin details. I rendered some tests of a comparison of Vicky 6 rendered in Iray with her default mats, then rendered with everything the same except converting skin to iray. I also did Danika for a comparison as well doing the same. The 01 tests are the standard mats...the 02s the only difference is applying the Genesis 2 female iray optimized setting. So what exactly the iray optimized mats doing to the skin. I find the base mats prettier in comparison but maybe i see the world differently than everyone else lol. What i have noticed in Vicky 6 for example she is more grey and less saturated but in Danika the opposite is true she is more red even her tattoo is no longer black...but a reddish color.
Daniel