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
Vibrance, from what I can understand, essentially equalizes saturation (+vibrance, less saturated boosted, -vibrance, more saturated reduced).
Given the great decrease in vibrance, how different is it from just desaturating?
If you just desaturate you take all of the colour out, the vibrance retains some, but not all, of the colour. I've been doing it with my textures since I first found out about it. Finding a decent program that had it was a bit of a trial though, PT Photo Editor is free.
CHEERS!
Most desaturation things are on sliders, so you don't have to take it all out... but I might poke at PT Photo Editor, thanks.
Of course, what I REALLY want is 'unbake the shadow information from this skin color texture,' which is complicated by things like darker skins tend to darken preferentially in wrinkles.
I wonder if one could do something with a good bump map. Hrm. (Like take bump map, invert, use it to slightly value correct the diffuse map...)
Almost... every skin type, be it from someone of asian, caucasian or african descent; filters MOST light waves under 600nm, the lower the wavelength, the stronger the effect. What differs at an african skin type, is that the curve towards the lower wavelegths is way much steeper. (see the graph below. Source: Sheng-Hao Tseng, Alexander Grant, and Anthony J. Durkin "In vivo determination of skin near-infrared optical properties using diffuse optical spectroscopy", University of California, Irvine Beckman Laser Institute Laser Microbeam and Medical Program).
The wavelengths for yellow lie in the narrow range between 575 and 585 nm, 600 nm is more orange (585 - 650 nm), red is from 650 - 750. Unfortunately most studies cover more the near infrared ranges, which isn't very helpful for our cause.
Unfortunately I hadn't the opportunity to test them yet. The cooler of my PC's power supply decided to continue to fulfill it's duty any longer (deserter). And I invested in a long overdue enlargement of RAM first. But from the information given by one of their creators in the IRay and a bunch of new map types in Xtreme Merchant Resources Set 01 they seemed to took care of Iray usability.
Yeah, but that depends on the quality of the normal and bump maps. The normals for Victoria 6 f.e. I don't find that great.
Umm, that's Alessandro_AM's building site... Look At My Peach Fuzz.
Or just consider the characters are all freshly shaved, no fuzz left... 
Oh, right, I know what you mean. You mean the Hue/Saturation controls, they tend to be a bit fiddly. The vibrance one is just input value and click. Gimp does interesting bump and displacement maps and they retain the colour of the diffuse map, I don't know if you're supposed to make them monochrome afterwards or what, but I'll have a look at them and see.
CHEERS!
This is interesting:
https://www.lightworkdesign.com/images/support/3dsMaxDocumentation/features/materials/conversion-materials.html
If you look at the section relating to Iray + Skin it shows you what SHOULD be in an Iray skin shader and just how much Iray Uber has missed out. If we had all the correct bits with the correct maps then we really would get some incredible results.
CHEERS!
Thanks, Roger, most interesting indeed. From the renderer description I get, that you should take into account that Iray+ would be Lightworks' advanced version of NVIDIA's Iray. You can tell that from the instance that in their Iray+'s SubsurfaceScattering you're able to use full RGB colors (which enables to detemine different scattering coefficients for each the red, green and blue wavelength ranges), where the common Iray renderer still use greyscale (uniform scattering over the whole wavelength ranges, though I've been told that NVIDIA planned to work on a change). Most of the parameters matching IrayUber's layout, though the naming is a bit different.
Unless DAZ will get an updated version from NVIDIA with similar capabilities, or their hands on the source of Iray+ to be compiled for DS (through buy, beg, borrow or steal) , it's pretty useless to currently adjust the IrayUber layout to Iray+ capabilities, for the common Iray renderer would just plain simply not know what to do with these settings.
I see,
Right now, no matter how much I adjust the Transmitted Color or the other settings, it seems to make no difference at all. It seems to be the translucency settings that are doing what we perceive as sub surface scattering. You said yourself that the common Iray renderer uses greyscale, so putting any colour in there isn't going to make the slightest bit of difference, is it!?
I certainly haven't seen anything noticeable, have you!? If you look at this gallery image of Cath's, this is what we need, but I don't think we'll get it with Iray Uber:
http://www.daz3d.com/gallery/#images/66270
That is her shader at work, we need this shader, stat!
CHEERS!
Ok, you'll love this,
I put a monochrome colour into Transmitted Color, and it made absolutely NO difference, neither did changing any other settings!! All those supposed settings that the 3Delight lot thought were SSS settings do precisely jack in Iray Uber. To increase or decrease what appears to be SSS in Iray Uber you adjust Translucency Weight. This was what I did to the ears on Darius here:
Nobody on the Nvidia forums can help with those scatter settings as they don't work in Iray Uber, period!
CHEERS!
EDIT:
Ok, after a bit more twiddling, they do something, but not SSS in the way it should be that I can tell. Translucency weight looks to do that job to me. You do what you like, but, I like the results I got more than doing anything with those settings.
You can see it a bit more clearly in this render:
I also vary the top coat settings according to the body part, so the head is different to the rest of the body, as are the ears. You do have to be very careful with this as you will get lines where the edges of the material zones are (You can see this on the back of his neck.). As far as I know this is unavoidable, but, when you add clothes and hair this won't be as noticeable.
CHEERS!
PS (A month ago I was still using 3Delight and would never have dreamed I'd be doing renders like this.)
Ok folks, major newsflash....
why should a map kill sss?..... all my tests showing that i can put any map in this slot and sss is perfect - that's why TINT is used to correct colors right there! and where you use a "shadow catcher map" a transmission map is actually the real thing..but i am still trying to figure out how to create one using multiple UV tiles... any suggestions here? Knald seems not to support more then one UV..
thickness map or transmission map is what you replaced with a "shadow catcher" map... the datas are for sure different.
No, I'm talking about these settings:
None of them actually does anything that really looks like subsurface scattering to me. As far as I can tell they just add an odd hue to the skin, nothing like the effect that adjusting translucency weight gives. I put a little edit in my post, but, to me those controls hark back to the AoA Subsurface shader and they worked for that, but, I think things are different here.
CHEERS!
Yes well, if you have that reflectance tint then you are using different settings. I am only reporting what I have found using the settings I have been using. Anyhow we've moved on from there and you will see that I've found things that work for the way I have things set up.
Other people can set things up how they like, I'm getting good results from what I'm doing and don't want to complicate things by adding other things that have nothing to do with that.
CHEERS!
It uses greyscale for SSS Amount, that's why it is just a slider by now, and that would need to be changed on the IrayUber when/if NVIDIA updates the Iray renderer, Transmitted Color still is supported in full color. That you're still confusing those tells me that you may first get yourself a little bit more familiar with the IrayUber.
If you look at the date of render, that was made at a time when Iradium was still in a beta status. I somehow doubt that she made her own special shader setup in Shader Mixer, since it wasn't that complete then... and still buggy. The render you're talking about would also largely depend on what textures and settings she used, and I also recall that she was using ZBrush a lot for improvements. Well, then you could go and ask her if/when she plans to release a skin shader.
See above. If SSS wouldn't work on the IrayUber, you couldn't do something like the below with it.
Simple SSS experiment: cylinder primitive, 10 cm long, 2 cm in diameter, Translucency Weight at 0.5, Transmitted and Scattering Measurement Distance both at 2.00, SSS Amount at 0.5, SSS Direction at 0.5. Single SpotLight at 6500 K, 20,000 Lumen, camera put directly opposite the SpotLight to shine through the volume right into the lens of the camera. Aux Viewport with NVIDIA Iray:
"SSS1" has Transmitted Color set to 1.00, 1.00, 1.00 as you can see. Result: no absorption happening, as intended, the light leaves the volume unchanged, the cylinder appears to be of a white color, right? 1.00 simply means, that 100 % of a special wavelenth will come out the other side of the volume, nothing will be filtered out.
"SSS2" has Transmitted Color set to 0.10, 0.92, 0.20 as you can also see (if one guesses that this the exact color used in the colored_wax.mdl example... he's guessing right!). Result: various absorption happening over the wavelength ranges, the light leaves the volume changed, the cylinder appears to be of a green color, right? Thats the result of absorption happening on lights' way though the volume.
Conclusion: The people on the NVIDIA forums, which are employees of that corporate, can most obviously help with those scatter settings, as SSS works on the IrayUber as intended. Period! (Now you write that a hundred times, alright?)
Speaking of, I just got answer from JanJordan from NVIDIA today on our question, if a texture would work in the Transmitted Color texture slot.
Quote: For a transparent material there are 2 places where you can absorb light /color:
-color the transmission bsdf: this is happening at the surface and is a fixed amount. this can easily be textured
-color the volume itself through absorption. this is then depending on how far light travels through the surface and can not be easily textured (right now not at all in iray). End Quote. So it's a NO on that.
Final word of advice: when working on skin settings, don't use primarily the included HDRi, better use meshlights; when you think you`re satisfied, you can try it under different SpotLight/HDRi-setups to check if your skin will look as good as in any other lighting solution. AND... don't use caucasian skin parameters on a african descent skin and expect a good result. They won't work that well, since the latter's skin has very, very different absorption and scattering properties, see the graph in my posting above and compare. Jus' saying.
Ok, based on what you were saying, yes, maybe those settings do make a difference, but only if you are using photometric lights, which I'm not. I'm using Iray Sun Sky and nothing else. I'm finding it gives me the best results for the kind of renders I want to do. The only other lights I'd use would be the emissive type like DZFire's Real Lights.
Nothing else I've looked at or seen looks right to me. Anyhow, I'm done with this thread.
well thanks arnold....
the only way to improve what we have right now are better maps...... so i am asking again: does somebody know how to create proper transmission maps (thickness) for a Gen2 or 3?
This maps are created from the mesh and not in photoshop. until yet i only found Knald which is a great solution and i will buy it- free 30 day trial.... but it does not handle daz multi tile UV's or i could not figure out how to load them :-)... is substance painter able to work with multi tile uv's correct and can it create transmission maps?
Sorry, I meant thin walled off, Atleast we are talking about the same things at least.
Here are some fun examples of what the Transmitted Measurement Distance, Transmitted Color, and Scattering Measurement Distance settings can do. As far as I can tell Transmitted Measurement Distance answers the question at what depth does the color of the sss switch from the translucency color to transmitted color. Transmitted Color, is the color the sss shifts to as we get deeper in the object; wheter it is a bug or limitation of the hardware, sticking a texture map here the shader will act as if the trasmitted color is white (and therefore the sss acts pretty much like standard thin walled translucency). Transmitted Measurement Distance is the overall sss scale.
My 2 examples. Both are rendered with the same lighting setup 2 large(so soft) strong lights behind the figure one less strong light in front, and a very ambient environment. There are no maps in any of the color channels, I also turned off the specular so there would be nothing to distract
Here I set the translucency color to purple and the Transmitted Color to green. Thinner areas like the nose and ears pick up the translucency color, thicker areas like the neck and head pick up the Transmitted Color. I recommed sticking these colors in and playing with the settings to get a feel for wnhat they do with more noticable visual feedback. (also some nice strong backlighting)
For this one I used the principles I figured out to solve a potential real world problem; Specifically how to simulate stronger sss on the ears without the whole figure lighting up like a Christmas tree. To do this I raised the Scattering Measurement Distance, so that the overall sss was brighter, but then made the Transmitted Color darker so that the thick areas saw less of the effect. With the Transmitted Color set to the color it was originally (or a larger Transmitted Measurement Distance) the neck would have been glowing and look completely unnatural. *edit, another useful thing is to either desaturate or shift the hue of the transmitted color to something in the bluer or greener range (at a low value of course) this can help remove some of the overall redness, without removing it fro the thin areas*
For me at least transmission depth maps are less of a priority, they're more of a holdover from using translucency without sss. given that Iray's SSS is raytraced it might be easier to just add some primatives underneath the skin to simulate the bones.
Just for you Rogerbee, here are those settings using just the sunsky.
also it goes without saying that I'm not saying the values i'm using are correct or what I'm going to use. these are very testy test renders purely to test what the values do.
"For me at least transmission depth maps are less of a priority, they're more of a holdover from using translucency without sss. given that Iray's SSS is raytraced it might be easier to just add some primatives underneath the skin to simulate the bones."
But i would say that also in your example above, this is exactly what is missing.... your settings are good and look perfectly on the ears and also partial in the face.. but the skull and forehead skin looking now way to soft.... mapping would solve that.
Good to see also on the nose, the tip is red - in nature it is not that way.... the nostrils transmit light, the tip is darker. your examples showing me extactly all the small missing things which happenend to me on my tests too :-)
Great examples above using colors and explaining the different values - thx.
Confusing is maybe also, scatter and transmitt or scatter only..... i always use the first one for skin. and those modes have really different effects.
Yeah, I have no idea how or why some of the settings do what they do. But I'm starting to figure out what works well, and if something isn't working how to fix it. I have no idea for scatter and transmit vs. scatter only is perhaps something to do with simulating backscattering? I have used both depending on how my textures are set up. (if you're using scatter and transmit and a map for translucency strength it will override the slider value rather than be multiplied with it which is a PITA)
another fun test, one of the reasons I dont like transmission depth maps (unless you have a way to stick them somewhere like Transmitted Measurement Distance) is that what they do is too much of an approximation, what they should ideally do is say "okay the light travels through all the skin at roughly the same rate, but in these areas at a certain point there's bone and that is not translucent." instead it often gets stuck somewhere like translucency strength where it says "okay the translucency is not as strong here" but the thing is things like the front scattering should be just as strong, just there shouldnt be any light reaching from the back.
Here's another fun and ugly expiriment. I put some spheres inside g3, (I forget where but I saw this suggested in a cucles totorial somewhere once) I turned the sss way up to show the effect. If you modeled the (very) basic shape of the sceleton, you couln't get much more phisically accurate (I wonder if you could transfer the old skeleton to g3?)
Also for gen 2 dimension theory does have some translucence strength maps at the store here that could potentially be rehashed, the question is where to put them. I may have to go shader mixer diving and see if it's possible to add a map to Scattering Measurement Distance or Transmitted Measurement Distance.
Scattering Measurement Distance - This is the distance light must pass through an object before getting the SSS Amount.
Transmitted Measurement Distance - This is the distance light must pass through an object before getting the transmitted color.
i think those two values are independent from each other ..... they can also overlay....
Scatter & Transmit - This works like Scatter Only, except with reflection, and based according to the Base Color's RGB; however, the remainder is sent to the Transmission layer instead of absorbed.
so -> we have overlaying parts PLUS more light (the reminder is sent to transmission)
if they overlay it looks too waxxy instead like skin while on thicker volume it looks perfect (the values are correct!).... this happens even on MECs renders.
That's why i think a correct transmission map is needed which blocks TRANSLUCENCY complete on forehead and other places.. but scattering still happens if you dont use a to high distance value...
ears... they look waxxy because the thicker parts are not proper masked... to much effect ... most just reduce then sss and translucency amount.. and the skin looks liveless again.
so the question is can we map also distances? scatter and transluceny? here my test stopped ... because i dont have proper masks yet .
I also had the idea with a skeleton but modeling one as the base of a ge2 g3 then exctract displacment (subtract differences to a normal base map) and mix it with a thickness map (nose ears) if transclucency weigthing works (zero blocks 100%).. well this should be the best possible way.
Nice examples, J. But... simulating "bones" in this situation is a bit senseless, since visible light on most parts of the human body doesn't even reach that deep. Explanation given here, a few good and easy understandable sources named here. The thing you shall aim for to put in a texture map is the hypodermis aka subcutis.
The labeling for the IrayUber are sometimes a bit strange/misleading. The "specular BSDF" (which is responsible to simulate specular reflection) has in MDL three different "scatter modes":
scatter_reflect, (reflection only)
scatter_transmit, (transmission only)
scatter_reflect_transmit (reflection & transmission).
An explanation of reflection and transmission can be best viewed here. Since all of them are part of the "scatter_mode" field, the "Scatter Only" parameter would actually have to be labeled Reflection Only and "Scatter & Transmit" Reflect & Transmit to accurately describe by name what they are there for.
In combination with the weird and nonsense Transmitted Color settings on Generation 7's volume material properties it makes me think that someone urgently needs to RTFM (as you english language people so formidably use to say. *20 page essay 'til next monday, in combination with an official apology in a traditional japanese style (on video for sure), may get you off the hook.*).
"Scatter & Transmit" for translucent materials is the correct choice, incident light which doesn't get reflected will be passed through the current layer and passed to the underlying layer(s).
Mapping distances is a really good idea Andy, but unfortunately "thickness" isn't be determinable in any way in MDL (I asked). Determining this by "Weight" is the only (and most abstract) way to do.
*Mapping distances is a really good idea Andy, but unfortunately "thickness" isn't be determinable in any way in MDL (I asked). Determining this by "Weight" is the only (and most abstract) way to do.*
Yes - that's where i am heading... because the only slot where we can control a part of the transmitted light is "translucent weight"...
https://s3.amazonaws.com/docs.knaldtech.com/docuwiki/transmission_capped_new.jpg
The above example is such a transmission map... such a map in addition to a "skeleton difference map" is the goal... and well... i wont use a skeleton... actually i will model more a mummy - i need the bones which are directly under the skin, skull, ellbows and so on... plus epidermal Map (veins and so on)..
as far as i understand the shader mixer.... we can not change the layers (iray works with 3 layers).. but we can add more map slots!
I also see in your other thread that you figured out the scales... what is actually distance measured nm mm cm ? ... i dont get it... just trial and error yet
and because you seems to be the one which put most facts about skin and Iray togther .> question...... i always tought skin looks silky instead waxxy because we have also backscattering effects... but trying this in iray even with the slowest amount - results in strange places where the effect is best to see (under the chin in shadow as example).
Thanks Andy, too much the honor, I know that at least V3Digitimes knows much much, more about the skin topic and optics, and you yourself more about lighting and tonemapping settings than I might ever could learn.
Silky instead waxxy is right, and backscattering also. I'd have yet to find any study that investigates where the silkiness comes from.
There's a slight difference in real-world and IrayUber where this backscattering is resulting of. In real-world, backscattering results from incident light encountering the hypodermis in around 4 mm depth, from which it gets reflected back to the skin surface due to it's high refraction index (around 1.7). That would be a subsurface backscattering effect, right? A thing we unfortunately don't have.
IrayUber's backscattering seems to be tied to specular reflection and transmission distribution functions (part of the glossy layers), which would make it a surface backscattering effect. So we sure have backscattering available, but it happens at a wrong place on our skin, up high instead down below. I'd say that this silkiness is a combined result of backscattering and anisotropy, and so a combination of subsurface and surface effects. In a meso scale, our smooth skin doesn't look that smooth anymore, wrinkles and pores make it look more like a World War I battlefield, with craters and trenches all over, worse than our moon's surface.
Some time ago I read a good example explaining the anisotropic effect (unfortunately can't find that anymore). Imagine a polished metal plate, an almost perfect mirror, then scratch thousands of more or less parallel very fine lines in it. The lines will scatter the light wildly all around and stretch the reflection to both sides along the axis of the scratches. That what happens up high. Down below anisotropy is mostly the result of a change in light's direction by collagen fibers, very tiny cylinders of about 0.5 - 3 micrometer in diameter and around 300 nanometers long. 70 % of dry weight of the human dermis is made out of those.
Human skin is a very complex thing, and it's inventor was either really, really clever or completely crazy... maybe even both. Iray´s abilities are limited, and we surely can't mimic all aspects properly. Specular reflection, refraction, absorption, anisotropy are the things we can work with, and hopefully we'll soon get some of the capabilities from 3ds Max's Iray+ four our Iray, too. Full color SSS Amount and subsurface backscattering would be nice to have.
About Shader Mixer: well, you could build a completely different and new one out of the bricks available, even write your own skin shader MDL and put that together. But before you even think about that, just try and do an "Import From Scene" from something using the IrayUber... me, I did a Homer Simpson.
as always you put some facts in your posting which i did not know yet :-) thx...
...
...
ok we know we have subsurface scattering and subsurface - backscattering (without parameters, automatic more or less)...
Subsurface Backscattering = the light which leaves the skin.... coming from front lights on our object ->
Backscattering = light which shines trough a object lit from the backside -> here we have actually translucency in our shader... but it does not scatter! I know it doesnt scatter because the effect in nature is that skin edges get a white shine (silky).... using translucency, ears are just plain hard contoured but showing our expected red (light from the backside)....well demonstrated from j cade above
Glossy has a backscatter weight... .. well... now i am guessing,maybe because RGB in iray scattering is limited yet - Nvidia came to this strange terminology -> translucency should actually be backscattering - but cant shine as white (close to the light source color) wavelentghs back... so they put it seperatly under glossy ( named it here correct "backscattering)..... starting to make sense to me now
However the effect i am looking for is one of the missing details yet - and all skin threads and examples overlooked that yet.... or maybe i just overlooked the one important posting how to use glossy backscatttering the right way for skin
Your hint about using anisotropy brings me on the idea that top coat is also a part of specular and glossy.... and maybe i never switched on anisotropy there at the right test moment
As always, you're welcome, Andy. Do not despair, brother, that can easily be fixed: 'The Appearance of Human Skin', 'Optical properties of the human skin'.
That's correct, the technical method to simulate backscattering in Iray is by SSS Direction, a single parameter. So we can determine where to our light is reflected to, but not where at this occurs.
I fear I put my explanations a bit unclear and you confuse diffuse transmission (translucence) ("light which shines trough a object lit from the backside") with backscattering, which is in physics "radiation deflected by scattering processes at angles greater than 90 degrees to the original direction of the beam of radiation." Miller-Keane Encyclopedia and Dictionary of Medicine, Nursing, and Allied Health, Seventh Edition, 2003
Well, the effects are somewhat tied, since you can't have backscattering on a non-translucent material. But a translucent material on the other hand doesn't need to neccessarily have backscattering capabilities. On human skin, the main scatterers and absorbers (melanin & haemoglobin) are forward scatterers, away from the lightsource. If it wouldn't be for the adipose cells (fat cells) of the hypodermis, light would travel all way down and be able to reach deeper into the human body. Although melanin and haemoglobin absorb most of the low wavelength/high frequency radiation, like UV, imagine what a high intensity UV radiation could cause to the tissues of our inner organs if it wouldn't get deflected. Ever got a serious sunburn?
So the translucency parameters are in some way correctly modeled, and the color/mapping of that would be a mix of what you'd see if you remove the layers on top, one after another, until looking at the surface of the hypodermis, the great backscatterer.
That really depends on what skin layer your Top Coat is thought to represent. It could be used to mimic the sebum (oily film) alone, or the stratum corneum in combination with a Top Coat Thin Film for the sebum. Recently I'm used to set the Glossy Layer of the Base to off (playing epidermis and dermis, so Glossy for that needs to be deactivated, since they both take a major part in the subsurface diffuse transmission and diffuse reflection story, but only a minor, if any, in the surface specular reflection story), and let the Top Coat simulate the two uppermost layers.
double posting
No i understand absolutly what transmission and translucency and backscattering means
... you explained that well... and i am using other render engines since 25 years (pixar renderman).... the problem is irays terminology...
translucency does not scatter in iray (sss does). But it should have a back (light) scatter - which brakes the wavelenghts in other ranges then subsurface scattering does...
The effect is that light travels TROUGH a volume and shines stronger on the other side (greater then 90 degrees - 180 degrees) specialy to see on EDGES (bundled light after scattering)absorbation. it shines again close to white/yellowish on edges even in shadows....
this is GLOSSY BACKscattering in Iray -> missed in 99.9% of all skinshaders examples here....
I am close to this missed effect right now with my test shader setup-> important is to switch of shared glossy inputs and TINT glossy backscattering correct...... THIS results in a paler face (the redish problem is solved) and edges where the light shines-> this effect goes directly to TOP COAT... and is best to see on a wet skin....
Problems i have with the direction right now (no parameters - one for all scatter effects )....
See my example 2 -> Glossy Backscattering on - chin line is right, nose, fingers too.... EARS, same setup, but lit from behind (strong translucency) = strange glossy backscattering efffect ....(oily skin is intendent to see the effect better).
example 1 - no Glossy Backscattering -> looks good on the first view - but contoures on fingers chin nose are to STRONG..... I use reference light sources 5000k for a color neutral setup.
Texture is standard Gen3.... made for 6500k - there it looks best /which is wrong - to warm (tan) in neutral light 5k)... i will take care about that later