Integrating 3D models with photography
Interested in integrating your 3D work with the real world? This might help
# 1 22-12-2002 , 10:12 AM
pitbull's Avatar
Registered User
Join Date: Dec 2002
Posts: 125

my ugly mental ray render...what went wrong

well.....all said :-)


got caustics,gi


:-)
# 2 22-12-2002 , 11:21 AM
Registered User
Join Date: Sep 2002
Posts: 10
you need final gather (or to pull the photon size down)

# 3 22-12-2002 , 06:31 PM
pitbull's Avatar
Registered User
Join Date: Dec 2002
Posts: 125
what does final gather do? i never tought mental ray would be more difficult to use than final render


:-)
# 4 05-01-2003 , 09:45 PM
undseth's Avatar
Subscriber
Join Date: Aug 2002
Posts: 1,135
Your GI solution can get much more better than this, but at this time I'm all buzy with wiping my HDD partitions in a preferred way.

I'll get back to you when my laptop is functioning properly user added image


I do not support Bush! Nor Saddam! Long live UN!
# 5 09-01-2003 , 04:24 AM
pitbull's Avatar
Registered User
Join Date: Dec 2002
Posts: 125
undseth :-p Thanks! btw....ur mr study was quite helpful :-)


:-)
# 6 09-01-2003 , 01:30 PM
undseth's Avatar
Subscriber
Join Date: Aug 2002
Posts: 1,135
From my memory, I'd say that your GI solution has too few photons + too low accuracy + and too low a exponent value.

A way to get a visual clue, in regard to photon count, is to set the exponent downt to "1" (in order to see the photons more clearly). And reduce the GI radius if needed. The point is to see the small "spots" that comes of this. Looking at the small spots you get an idea as to where the GI photons hit.

Not sure what an optimal solution would look like with GI, but go for a solution that has a decent distribution of photons onto your model.

Secondly; Fiddle with the radius setting. Not sure what is optimal, but I'd say, go for a radius settings that makes the photons overlap each other. If the radius is set too high, I guess the light will turn up wierd, but then again I'm not an expert.

With photon count and radius set. Increase the accuracy to get a smoother rendering.

Last (I guess), turn the exponent up from "1" to "2" or around that. I think the "normal" light exponent setting is "2" for natural sunlight.


I do not support Bush! Nor Saddam! Long live UN!

Last edited by undseth; 09-01-2003 at 01:33 PM.
# 7 11-01-2003 , 12:49 PM
Jim's Avatar
Registered User
Join Date: Jun 2002
Location: Adelaide Australia
Posts: 201

re ugly mental ray render....

Hi there,

I'm sorry if I missed something from not reading all of the post re mental ray (I've been busy and Lazy).

This may have already been covered in another thread re GI and mental ray.

Here is some images that I rendered with mental ray tonight.
My machine is a P3 450, not the sort of machine that you would hope to batch render GI stuff in.

The first two are using just a single spot light.

Time to render 1 minute

Spots are good for Photon use because unlike point lights that throw light and photons in all direction spot lights concentrate the light and photons in a single direction which means faster renders.

the middle image (same as top left) but with GI turned on.

Time to render 3 minutes
Settings are:

Spotlight:
Energy 180,000 (all channels)
Exponent 2
Physical Yes

Area lights turn "Off" (Unless you want excessive render times).

Mental Ray render globals:

Global Illumination tab:
GI accuracy 800
Gl radius 1.35

photon volume accuracy 35
photon volume radius 1.0

Max reflection photons 20, 20, 40


The lower images are the same but with 'final gathering' turned on to "smooth" the rough global illumination.

Time to render 4 minutes

final Gathering tab:

final gather rays 100
Min radius 800
Max radius 1600

Sampling tab:
Min 0
Max 2

Jitter Yes
Gauss filter

Hope this is helpful to someone. With renders this quick one could easily render animations not just pretty single images.

re Jim


Mayan elder......only in years.....
# 8 11-01-2003 , 02:48 PM
undseth's Avatar
Subscriber
Join Date: Aug 2002
Posts: 1,135
hmmm It looks that your GI radius is set too low, as you (apparantly, if I am not mistaken) get dark artifacts in the corners, (even with FG).

The photons are well distributed I guess, but imho the radius should be bigger, to smooth it out better.

I know nothing of the "photon volume accuracy and radius", so please tell us if you have something to teatch away regarding this setting. ( I never fiddle with this, as I have no clue at to how it works.)


I do not support Bush! Nor Saddam! Long live UN!
# 9 12-01-2003 , 12:25 PM
Jim's Avatar
Registered User
Join Date: Jun 2002
Location: Adelaide Australia
Posts: 201

re ugly mental ray render....

Hi there underseth,

I forgot to mention that I beveled the edges of the 'room'. This would account for the dark areas in the corners.

The reason I made the radius so low (forgive me if I'm wrong), was to 'capture' all of the detail in the head model. I found that I was loosing shading detail when using larger photon sizes.

Also, the post was mainly aimed at using final gathering to smooth out grainy GI solutions. I am keen to find out how other people are using mental ray. And yes, I exported a similar scene to xsi 3 (via xsi dot) and rendered using same/similar settings. Maya was on par or slightly faster which is good.

Dedicated forum would be great, we could have lots of fun, just the two of us.

re Jim


Mayan elder......only in years.....
# 10 14-01-2003 , 10:42 PM
pitbull's Avatar
Registered User
Join Date: Dec 2002
Posts: 125
thanks for the info :-)


:-)
Posting Rules Forum Rules
You may not post new threads | You may not post replies | You may not post attachments | You may not edit your posts | BB code is On | Smilies are On | [IMG] code is On | HTML code is Off

Similar Threads