Strange diagonal structures caused by iRay
AndyS
Posts: 1,438
Hi,
since I do some experiments with the new nVidia iRay render engine I obseerve strange diagonal structures on almost all the older textures. (see attached clipping of the Aikanaro environment).
With 3Delight this never appeared.
Can the DAZ development team please explain, why they delivered a that buggy component at that early development step?
There are that many bugs in that render engine - See the enormous number of posts in this forum.
- Displacement problem,
- Refraction not correct,
- commercial HDRIs not really working,
- Environment- and map intensity not working seperately,
etc., etc., etc, ...
It is time for a general homecall and a fundamental rework.
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
The default cube, with a UV test grid and the Ueno Shrine HDR from sIBL.
No sign of artifacts, only some residual noise, since I let Iray work only for 14 minutes. The zooming on the corner shows uniform colors where they should be uniform and razor sharp transitions.
Since the Ueno Shrine HDR is very reddish, the red parts are (correctly) overexposed.
It's far likelier that 3Delight's refraction was "wrong". I was musing tonight on building an imperfect camera lens to stick in front of the camera to see if it produces the natural abberations you get with analogue lenses.
First thing I do with a surface in iray is change the surface angle to 30/40 degrees, and any textured surface has its diffuse value changed to full white or down to eighty percent (textures multiply the base diffuse surface).
It's a different renderer. You just have to bite the bullet and accept it's pretty much nothing like 3Delight.
Displacement is NOT broken...it is just not the same as 3Delight (not much else is, that is one of 3Delight's strongest points...in fact, some studios just use 3DL to render displacements! And use other renderers for other parts of the image and composite all together). You cannot really compare the two, because they are so different. About the only thing you can say about them is that they are radically different.
Refraction is correct...the geometry used for water is key. Liquid in almost any PBR is volumetric in nature...and the geometry used must reflect that. Or you need to cheat...not only on 'faking' the volume, but on the IOR/refraction weight (strength) to get the 'look' you want.
Don't see where/what the problems with HDRs are when they are correctly set up/used. If they are of sufficient dynamic range, are of sufficient size they can do lighting, with nice crisp shadows AND be used as a backdrop. Not adequate range or large enough resolution, then one or both will not turn out 'right'. Basically anything under 8K (pixels in width) is going to give mediocre to poor results as a backdrop. Anything in the low to mid EV range is going to give poor shadows and even poor color rendering.
I think that is one of the problems, people expect the old 3Delight scenes to render the same with IRAY, IRAY is a very different renderer so you can't expect that.
IRAY is pretty well tested, it's been around for a long time, the only thing new is the interface between DS and IRAY and there can of course be some bugs, but the actual renderer should work just fine, just don't expect it do behave the same way as 3Delight, it's an nVIDIA procuct that is not designed to be compatible with 3Delight from the beginning, but DAZ has done a pretty good job making it work nice with DS.
If you really think, the iRay-refraction is correct, so please see the domumenting post I created some time ago.
--> http://www.daz3d.com/forums/discussion/57352/
And for the iRay test each surface was changed by applying the necessary iRay shaders delivered with DAZ4.8
Even the usage of cubic "water containers" didn't change the situation.
I still don't think there is anything wrong with IRAY refractions, it has been used for many complex scenes in the past and as far as I know there are few complaints on refrcations.
I know there was some problems with glossiness in combination with refracrations 2 or 3 years ago, but that is pretty much it.
But as I said, this is a very different renderer compared to 3Delight and there are big chances that something goes wrong in the interface between DS and IRAY.
Yes for sure, it is a very different render engine. And it's got a lot of good aspects, no doubt.
But that hasn't to mean, that some natural sets are not possible to use !! :bug: :grrr:
And who is responsible to clrify the interface?
DAZ development team refused. And I'm really fed up to discuss in that kindergarten way "you failed <--> no, you failed".</p>
As clearly can be seen in my test results, the rays coming back from out of the water got the wrong "direction-" information.
And this is only one point that very early popped up as I tried to use iRay.
Other users detected a lot of different issues.
As long as this big amount of bugs is still open, the release has to be taken back for rework.
The problem with Displacement is two fold, poor textures and weak or wrong settings.
Settings of strength 100% min -0.1 max 0.1 will barely register in Iray, to get anything to show with those settings you need to be using an Iray specific shader like the Iray Uber shader, that way you have access to the SubD Displacement Level, as you need to move it up a few levels, just don't over do it as DS will CTD.
Far to many of the HDRI I have tried have crap dynamic ranges, this means the light and shadows you get never match the "image".
Yes for sure, it is a very different render engine. And it's got a lot of good aspects, no doubt.
But that hasn't to mean, that some natural sets are not possible to use !! :bug: :grrr:
And who is responsible to clrify the interface?
DAZ development team refused. And I'm really fed up to discuss in that kindergarten way "you failed <--> no, you failed".</p>
As clearly can be seen in my test results, the rays coming back from out of the water got the wrong "direction-" information.
And this is only one point that very early popped up as I tried to use iRay.
Other users detected a lot of different issues.
As long as this big amount of bugs is still open, the release has to be taken back for rework.
Is your water just a flat surface plane? Iray's PBR so probably needs a volume like you find in the real world. Try a cube instead of a flat poly. If your camera's inside the water, it would probably need a "bubble" for the camera to sit inside of as found with fog volumes.
Would you mind reducing the scene to a very simple one and post the .DUF?