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
Perhaps the links to the tutorials of SickleYield could be helpfull. --> http://www.daz3d.com/forums/discussion/56788/iray-start-here
But in general there isn't much.
According to this picture
there's nothing wrong with the shadow line moving towards the normal.
I really doubt that someone can "use an IOR in the wrong direction". It's just a factor. What someone can do is messing up the formula to calculate something, though. But I really can't see any evidence that this would be the case on this example.
Anyways, this forums would be a very inaccurate place to criticize the way Iray works, or complain about bugs. Have you been at the NVIDIA forums to explain your concerns?
The official DAZ documentation for the Iray Uber shader can be found here.
Useful introductions to the realm of PBR rendering/shading can be found here, and here, and here. Although they're for different renderers than NVIDIA's Iray, the good thing about PBR is that the underlying principles are always the same, regardless of the render engine used.
Still very stuck ... right now I'm just annoyed with the Sclera settings. I've tried them all from all the notes I've taken. I thought it was the lighting but when I just turned it all all off and used just a Eye Accent light and the HDR to light the scene and her eyes, same issue. Just this very flat look to the Sclera with a strong shadowy look around the outer edges. This setting is just using the Water setting with the maps put back in and the Refraction Weight turned off. Not going to pull my hair out over this but I'm just very technical and I really need some advice as to what to try. All the notes I got from Cath's thread looked very good but nada with the results. Here they are:
Rendered using MEC4D's lighting rig
I've been working on eye settings and i think I've hit on some that look nice. Here they are for whoever is interested:
Eye Reflection: default Glass-Thin-Clear shader
Cornea: Glass - Solid - Clear - Dispersive (rainbow) shader, cut-out opacity set to .25
Sclera:
Base color - White + Diffuse Map
Translucency weight: 0.25
Base color effect - scatter and transmit
Translucency Color: 212-61-215 + Diffuse Map
SSS Reflectance Tint - white
BaseBump : 0.35 + Diffuse Map
Thin Walled: off
Transmitted measurement distance: 2.5
Transmitted Color: 204-24-255
scattered measurement distance: 0.1
SSS Amount: 1.0
SSS Direction: -0.5
Iris & Pupil
Base color - White + Diffuse Map
Diffuse Roughness: .5
Translucency weight: 0.25
Base color effect - scatter and transmit
Translucency Color: 212-61-215 + Diffuse Map
SSS Reflectance Tint - white
Glossy Layer weight: 0.5
Share glossy inputs - off
Glossy Color: White + Diffuse Map
Glossy color effect: scatter only
BaseBump : 1.0 + Diffuse Map
Top Coat Weight: 0.5
Top Coat Color: White + Diffuse Map
Top Coat Color Effect: Scatter and Transmit
Thin Walled: off
Transmitted measurement distance: 2.5
Transmitted Color: 204-24-255
scattered measurement distance: 1.0
SSS Amount: 1.0
SSS Direction: -.5
Another bit of Eye Info:
Several people have asked about working with eyes and I've seen various advice so I'd thought I do a quick example and see how it turns out.
To try this out I decide to go find a reference eye to check against. A simple image search yielded several. I decided to compare against this photo of an eye. I knew I wouldn't be able to match it exactly but I wanted to see real reflections on a real eye. I also referred to these diagrams to make sure I had the right anatomical concepts.
The attached image is what I came up with. I did not do anything with skin, lashes, lacrimals, etc, they are just the standard conversion. I was focused on the direct eye materials only. I used the Genesis 2 Female with it default loads and converts except as noted below.
Here was my approach:
Tear/Eye Surface: Apply "Glass - Thin.duf" and change the index of refraction to 1.336 to match tears.
Sclera: Changed the base color texture from the default to number 5 from Awesome Eyes cause I liked the way they looked. I then turned the base color down to .74,.74,.74 because the texture was a bit too bright. I then reverted all other settings to their defaults except Glossy Roughness which I set at 0.25. This was the roughness I arrived at through empirical observation so it may depend on eye set up. I increased the bump to 1.0.
Irises: Setup just like the Scelera except the base color was standard white (1.0,1.0,1.0) and glossy roughness was set to 0.5.
Pupils: Pupils where setup like Sclera/Irises except the glossy layer was shut off by setting "Glossy Layered Weight" to 0. I did this because the Genesis 2 pupils are not physically modeled. So I need them to act like a hole even though it is not. The map sets them black for me, but I don't want them to reflect so I removed the glossy reflection as noted above.
Cornea: The Cornea turned out to be the most complicated to set up as it ends up doing multiple jobs at once. I first applied "Glass - Thin.duf" and changed the index of refraction to 1.3360 to simulate the "aqueous humour". I then turned the "Top Coat Weight" to 1.0 and turned the "Top Coat Layering Mode" to Frensel with a "Top Coar IOR" set to 1.376 to model the surface of the cornea. This was giving me good results but I found the line between Irises and Sclera too pronounced. This turns out to be an old issue with the Genesis 2 models and luckily many of the major DAZ Characters come with maps to address it. This map is found in "[Content Directory]/Runtime/Textures/DAZ/Characters/Genesis2/Victoria6V6AnnaEyeTr.jpg" for Victoria6. For 3Delight these maps go in the opacity channel but working from a tip I received at the DAZ office I instead put it in the "Refraction Weight" channel, then opened the "Image Editor..." and set "Invert" to On. If you don't want to use the "Image Editor.." method you can also invert the map in your favorite image editing program. I also added this map to the "Top Coat Weight" and again used the "Image Editor..." option to invert it. To finish the blend I then added the same "Base Color" map that I used on the other surfaces to the cornea. Finally I changed the "Glossy Reflectivity" to 0.4.
To finish the image I added the an hdri from HDRI Archives called "Old Industrial Hall" because it had dominate windows in it. Then adjusted tone mapping to match the hdri.
When none of this worked out for me on the Sclera (the Iris and pupil and Cornea look OK for my needs so that's OK I guess) I tried the iRAY Water shader and just put back the maps in and as mentioned turned the Refraction Weight back to zero.
HELP?
@RAMWolff, does your sclera UV match the Bree texture you are using?
Bree? This is Dawn from Hivewire. This is my skin that I'm working on for her, hence the endless questions about finding a way to get nice effects. I got the nails figured out, the skin looks good and the inner mouth. The teeth are almost there but the eyes... those darned eyes! lol
The list in your post says V5BreeEyesM and V5BreeEyesB textures on the sclera. Did I misunderstand?
I just switched those out for my own maps.... sorry forgot about those side notes there.
So been having more issues with this set up. The Glass - Thin - Clear shader works pretty well but the eyes look like they are under a dark glass. I've double checked all my settings to see if I had any of the diffuse channels set to anything other than pure white, and that's a no. Only thing set darker is the Cornea but it's really not dark. I was trying to take Andy's advice and keep the settings for the Cornea and the Sclera the same (except for the maps on the Sclera)
I've also noticed a very much unwanted effect on the Sclera ... I'm getting a very grainy firefly effect that doesn't clear up after the render is done. I remember running into this a long long ways back and can't remember the fix for that.
My lights in the scene are supplied by one distant light and one point light for extra flashes of light on the eyes and then the HDR supplies the over all rest of the lighting in the scene. The fireflies were much worse before adding in the Distant light and the Point light.
Thanks for the help! :-)
Hi Arnold,
I didn't say anything different.
Yes, but in my pictures you see, that the rays move away from the normal in the renders.
Those discussions/experiments were very helpful.
Looking at the results and simple tests I did in my bathtub, I'm sure that DAZ simply submits the inverted IOR for shadow/light rays instead of the IOR itself (or the other way round) into the interface for the render engines. So they (DAZ Developers) use the IOR in the wrong direction for this part of the refraction effect.
As I stated, my bug report is over one year old and meanwhile accpted by nVidia via the DAZ dev team.
So as long it is not solved, it is not really useful to use refraction due to unexpected "dark" effects in the render.
I let a light ray enter the water in a very gentle angle and even this was broken down to the ground with a big intensity. Opposit to it in iRay it already disappears below a steap angle of round 42°.
Sorry, but of course you did:
When lightrays enter a medium, their photon's velocity is changed by a factor which is given by the Refractive index (aka Index Of Refraction, short IOR) for that specific material. When, f.e., lightrays enter a volume of water with an IOR of 1.3333, the photons travel at a velocity 1.3333 times slower than in a vacuum, which's IOR is always 1.0).
"When light enters a material with higher refractive index, the angle of refraction will be smaller than the angle of incidence and the light will be refracted towards the normal of the surface. The higher the refractive index, the closer to the normal direction the light will travel. When passing into a medium with lower refractive index, the light will instead be refracted away from the normal, towards the surface."
So, the shadow line within the medium with the higher Refracive Index doesn't "go down", which would mean it it would be bend away from the normal. I took myself the freedom to add some lines to your render. The orange line would be the direction the lightrays would travel if they wouldn't get refracted. The blue line is the direction the lightrays travel after entering the medium with that said higher Refracive Index. As you can see, the shadow line "moves upwards" (is bend towards the normal), which is exactly the behaviour you'd have to expect. So, there's absolutely nothing wrong here.
The problem with the "critical angle" is another story. DAZ's Iray Uber uses Schlick's approximation, a simplified method which doesn't take polarisations into account. Nevertheless, when something's wrong there it would rather be the one at DAZ to blame who wrote that thing, than to state that NVIDIA's Iray is broken. Which would only be the case if you write your own shader in MDL, test it, and the result would be the same.
Hi Arnold,
sorry, but it is only partially correct what you wrote.
In the picture you see two big issues
1. the direction of the shadow:
Above the waterline you see the shadow following the blue line of the drawing. On that drawing the line continues under the waterline correctly calculated for the IOR of 1.33.
As you see, due to the refraction of everthing from under water, it seems to be a straight line. Somehow the entered refraction value in the parameter tab is only near the real IOR.
So far - everything correct.
For a minute please forget, that the shadow in the iRay render doesnt't continue, where it normally should. This is the 2nd part of this bug.
But what you see is, that the shadowline is folded upwards (the green line). Considering the 2nd part of the bug, at least it should follow your orange line.
So you have to agree, that somehow the render engine calculates with a somehow wrong value.
2. The gap of the shadow:
It is very obvious that the render engine calulates the refraction from the location where the shadow is created, instead of the real refractive surface.
And it uses the wrong value which is already the reason for issue part 1.
I don't want to judge, who did the mistake. As I reported this bug over one year ago, after some month the DAZ dev stated, that it must be the vault of nVidia.
So you must have something read which I didn't write. It was the DAZ team blaming nVidia. This is a further fact you didn't read correctly in my posts, as well as my statements about the bending down.
But to write an interface needs a lot of communication of both parties. And perhaps there are two parameters in the interface: One for the picture our camera records (upwards) and one for the light entering through the surface (downwards). It is only my private speculation, but if they simply inverse the downward parameter in the interface, everything would be fine.
The fact is: The part of what our camera records is correct, but the calculation for the light entering downwards somehow strange. This already could be observed for 3Delight renders. But not in that dramatic consequence (gab of the shadow and turning black).
Attached a real live experiment and the original picture explaining refraction and used in the render.
btw: using this angle in the iRay render the water already appears black.
Quick and dirty refraction experiment, using several cylinder primitives: the first is thought to represent a common drinking glass, the second to represent a liquid with a distinct IOR (and a "Water" / "Water - Dispersive" Iray Uber Shader preset applied to), and the third a pencil or drinking straw. This experiment's purpose is to check if you'll get a result similar to that in this picture:
Render A is made with the liquid having an IOR of 1.3333, regular water, and render B I gave it an IOR of 1.541 (representing C6H5CH2OH or Benzyl Alcohol), to see if the lower part of the straw, beneath the liquid's surface, will get to be more bend towards the normal, as the laws of optics tell us has to happen: "The higher the refractive index, the closer to the normal direction the light will travel."
Render C is the visual confirmation of a "being bent towards the normal" (green line) in comparision to "no refraction happening", the blue line. I can't see anything that would suggest a glitch or bug on Iray's (or the Iray Uber's) refraction calculations. Happy rendering.
Yes, sorry Andy, you're right, I confused the normal with the interface.
But still, if there would be some glitch in Iray's refraction calcultions, we would see a complete reversed result of what we would have to expect, don't we?
I did an additional render using the "Water - Thin" material preset. As ist seems, on this one there are no refraction calculations in use.
You're welcome.
I think, it is somehow more complicated.
With the "straw Experiment" we see one part of the refraction effect. Information of objects under water coming out of the water into our camera. And this part is reproduced (almost) correctly.
But light entering into the water is a second aspect. All what we see in the situation I showed in my renders is of cause ever influenced by the first part of the effect, too.
I created a new setup with an under water light source. The first outcome was very interesting.
But could we please switch over to my refraction error thread, because this thread here originally was about skin?
Plenty of eye experts here. I'm messing around with eyes currently and getting some odd shadows on the iris when lit at an angle. Front looks alright?
Did you read my entry on the previous page (Jan 24th) related to the refraction problems of iray renders?
btw: Your eye is a really good prop with nice texture maps. Where is it from?
Ah yeah rings a bell. So it is simply a bug and unavoidable? The eyes use Eva 7 textures and my custom shader settings, also 100% cornea buldge, 50% Iris Realism Buldge and 10% Iris Realism Depth (j.cade freebies).
So what would be the correct behavior, no shadows on the iris at all?
If possible, have the light entering the eye in a steep angle.
Or better: Set "Thin Walled = ON". So the refraction isn't working. OK - it is a little loss of "real" eye, but better anyhow...
And please !!!
Please folks, open bug reports related to the refraction bug! I started mine over 1 year ago. I supported the DAZ-Team with a detailed analysis. Now thanks to the previous discussion above we have some furher ideas what's wrong. But DAZ still refuses to rework.
But your advice don't help bluejaunte.
A lot of you repeat to complain about "too dark" irises. And that's the reason why !
Let's fight to get the bug solved - better for us all.
Has there been some acknowlegement by Daz that this is even a thing?
I was having allot of issue with the Iris coming out dark in renders, didn't matter the lighting. I could even put a spot right in front of the face.. there was no difference. I figured out that the less maps you have in the channels the better.
Below are my settings. Anyone can tell me how wrong they are but doesn't matter as no one was able to really help me, feeling ignored I just set out on my own path and this works for me. Aside from the Bump and Normal map a little further down there are NO OTHER MAPS used... give it a try. Maybe it will work for you too...
Hi can you kindly post a screen shot of your eye settings and also skin settings? And what product of eyes and skin do you use? :)
first @ramwolff don't use any translucency on the iris, that will darken it right up.
I now have some documented evidence that the refraction "bug" is indeed probably not a refraction bug. If you render in blender's cycles you get a similar shadowing effect. Testing in blender is particularly helpful since it is much easier to break a material into separate parts (i even looked in studio's shader mixer, but could find no way to separate refraction from reflection)
image 1 The genesis eye in blender's cycles with a simple material mixing a diffuse node and glass node via an image texture to get the smooth transition. You will notice that it has a similar error to the problem we see in Iray
image 2 rather than mixing diffuse and glass node I mix diffuse and just refraction. There is no reflection component, but the weird shadow is gone, the problem isn't the refraction, but how the glossy is added
image 3 the goal, the glossy has returned but without creating the weird shadows. This is simple in blender, you can simply make the glossy not cast shadows this cannot be done in studio currently and is unlikely added let alone easily done any time soon
So what can you so in studio? Well, if at all possible change your lighting, provided you are using hdrs and/or photometric lights with relatively large geometries, you should never notice weird shadowing. If you absolutely love your lighting other wise but are getting the ugly shadowing on you eyes, rather than switching to thinwalled off, I would suggest setting your glossy roughness to something like .04 it won't look quite as nice as a roughness of 0 (and will take even longer to render), but it still looks imo much better than thinwalled off, which continues to look miserable soulless and dead.
Aaand a comparison of: Thinwalled on (0 roughness), Thinwalled on (.04 roughness), Thinwalled off. and all of these a quick renders that needed to cook more because I dont like spending 10 minutes on a tiny test render
Thinwalled on (0 roughness)
The best option IMO. Has depth, looks the most real, I spent a lot of time staring at my own eyes in a mirror while piffling with settings and this is the closest its the only one that gets the nifty inter cornea caustic effect (bottom right of the eye) that I can personally state my own eye definitely gets as well.
Thinwalled on (.04 roughness)
Takes forever to render and is flatter looking than with 0 roughness. But will get rid of weird shadowing without resorting to thinwalled on. Also a bit more even than 0 roughness if that's what your after. Probably the closest to the actual iris texture in terms of overall value.
Thinwalled off
The worst. Flat, soulless, blown out color, and messes up the sss in the sclera. Its the fastest to render though
also for all that talk of thinwalled off making the iris unnaturally dark, in these examples the thinwalled off versions are much truer to the texture maps of the iris. Believe it or not this is pretty neutral lighting (If I sample the skin from the image where it isn't in shadow its pretty and sample the texture in the same location the colors are nigh identical, likewise if I sample the eyes from my image and compare them to the texture I'm using they are also pretty close, significantly closer than the completely blown out thinwalled on version at any rate.)
Interesting j cade, really like your results too. When you say .4 glossy roughness, do you mean on the cornea? That looks very wrong here.
Ah .04, got it.
j cade, have you tried using top coat? Maybe that is handled differently than glossy.