iRay caustics issues

iRay caustics issues iRay caustics issues kcpr raffaEl
14 Sep
2011

***UPDATE***

This article is related to iRay 2.x rendering engine and eariler.

Look here:

Houston, we have a caustics! iRay 3.x Sneak peak

************

After almost a year of using iRay, many hours spent on rendering, lighting, and materials tweaking I decided to write some posts, to point iRay issues, which I couldn't handle.

In this article I would like to express my opinion about so-called "physically-correctness" of iRay, which, in my opinion, in some situations is only marketing term...


Don't understand me wrong - in most cases I love iRay, and I've switched to it with most my works, and I'm aware that it is in it's developement stage, but for me the TRUTH and FACTS are most important, and I have to figure out what is the future of my renderig tools - iRay or vray.

I'm open for your comments - and your solutions, because I'm not saying that I'm completely right. I just want to make these things work, as Mental Images and Autodesk claims.

So - let's start, first part - THE REFRACTIVE CAUSTICS and TRANSPARENCY.


During my work, which is mostly architectural visalisations, I didnt't noticed the lack of caustics. But the time has come when I discovered, that something is wrong...

I've found the great tutorial of caustics for Mental Ray by DM Multimedia, which has attached scene to download.

I was curious about the quality of renderings produced by iRay compared to Mental Ray.

First, I prepared the scene for iRay, changed lighting to mrPhysical Sky with sun scale 1, corrected the exposure settings, and lowered caustics multiply to 1 (physically correct value) and did test renders with Mental Ray.


This is what I get:



Then, I've switched to iRay.

The effect I've got was dissapointing:



Why does this water looks like old, smelly sewage? ;)

Why there's no transparency, no reflective and refractive caustics?


OK, maybe I'm doing something wrong?

Lets try...

First idea:

This must be Advanced Rendering Options -> Refraction -> Maximum Distance / Colour issue.

Ok, let's disable it:


Nope.

Here's Mental Ray version:

Second idea:

The water shouldn't intersect the pool edges.

So, I've created circle shape then converted it to poly, changed the Camera to get better view and here's the render:

Still no caustics, not even near any "reality".

Mental Ray:

One more thing: Do you see how the shadow casts both on water and on the pool. It's physsically, correct, iRay version doesn't produce that effect (for now).

Ok, maybe the water shouldn't intersect ANY object.

Let's see:



Still the same.

And Mental Ray....:

Third idea:

- the water should be 3D object, not just a plane.

I've created 3D water, which for sure doesn't intersects the well.

To make sure, that displacement map doesn't cause water to intersect pond,
I've created non-displacement version of water material, which I've attached to all
geometry of 3d water except top surface.


Here is effect:


Still nothing.


Mental Ray:

To be absolutelly sure I've tried the version with 3d water intersecting the pool to eliminate reflections on the bootom of pond:

Mental Ray:

Next idea:

Let's change the material of 3d water to gray plaster, except the surface. And let's eliminate all objects from scene that may produce any unexpected shadows:

The render:



No use. What are the strange shadows?? There's no object here!

Mental Ray:

Then I've started to experiment with transparency value and IOR values,

and I've noticed, that at IOR = 1.15 (non physically correct for water) value something started to act better:

But still NO CAUSTICS!

Mental Ray:

So, finally, I've decided to render ONLY water, and some samples of high-reflective and refractive objects:

Here is Mental Ray version:


And iRay:



Hey! There are some caustics near spheres and teapot ... there are some but the level of these is absolutelly far too low, not physically corect.

The strange thing is, that at the begining of render caustics seems to appear stronger, but as the time pass they're cleared out:


I've attached .max scene (Max 2012 Version), so you can experiment by yourselves. All kinds of materials I've used are in the Matrial Editor, and objects are distributed to layers.

Any ideas to get refractive caustics work?

NOTE: Copyrights of source .max files used in this article beleongs to: DM Multimedia.
The .max file was used and modified only for educational purposes.

Last modified on Tuesday, 03 November 2015 11:23

6 comments

  • kcpr raffaEl We need "iray next"

    Yes, I absolutelly agree with you. I'm looking forward to iRay Next in 3ds MAX...
    Thursday, 28 February 2013 23:48 posted by kcpr raffaEl Report Comment Link
  • Nobuyuki Nakata I have the same problem on 3ds max 2013. We need "iray next" to have caustics. http://blog.irayrender.com/post/31979770119/iray-next-dispersive-caustics Wednesday, 30 January 2013 15:39 posted by Nobuyuki Nakata Report Comment Link
  • David iray 3.0 supports caustics, previous versions did not, this is why you could not produce caustics. They are generally presented in a renderer as a special case of global illumination. Monday, 27 August 2012 10:33 posted by David Report Comment Link
  • kcpr raffaEl Check the ZIPped file that I've attached: The sun size is 4 - look at the soft edges of shadows - it's because light source has a shape and is not a small point. I've tried even with the sun size 10 and circular ligt source very near objects. So the shape visibility isn't the source of issue. Saturday, 17 December 2011 00:38 posted by kcpr raffaEl Report Comment Link
  • zaparine In my opinion you should make your light source a visible shape and then it works Friday, 16 December 2011 15:14 posted by zaparine Report Comment Link
  • Erich Ocean iRay does see them, but they are found *very* slowly, and once found, are not pursued further.

    This will happen any time you are viewing a caustic through a reflective/refractive surface, and is a problem with unbiased renders. This is simply an effect they cannot **efficiently** produce.

    The reason the bottom of the fountain looked so dull is that it is entirely lit by caustics, which aren't passing through the surface!

    Maxwell Render has the same problem, BTW.
    Friday, 16 September 2011 10:06 posted by Erich Ocean Report Comment Link

Leave a comment

FaLang translation system by Faboba

Latest in Portfolio