PDA

View Full Version : Maya 6.5/Mentalray 3.4/secondary diffuse bounces


slipknot66
02-26-2005, 08:05 AM
Here is an image i did to test the new feature of mentalray 3.4, secondary diffuse bounces,using FG, 100 rays min 0.6 max 6, GI accuracy 100, radius 3. 1 point light converted into a mentalray area light with a square projection, emtting 10.000 photons, and 1 directional light, both lights using the new mentalray light shader mib_cie_d.
1 hour and 17 min, to render the scene, with an Athlon XP 1.6, 512ram dim(not ddr).. hehehe (from hell).
http://img148.exs.cx/img148/7797/secfgbounce7ci.th.jpg (http://img148.exs.cx/my.php?loc=img148&image=secfgbounce7ci.jpg)

twistedpoly
02-26-2005, 09:24 AM
if you would like to share the scene file, please go ahead :):)

floze
02-26-2005, 11:55 AM
Looks nice! Is the gloom made by photoshop or is it 'built in'? Or is my crappy tft here at home cheating my eyes? ;)
One question about the cie_d: Did you use an approximate rgb value for the photon color or did you manage to hook it up somehow?
Another question: How far away (from the window) did you place the area light? May you post a wireframe? :drool:

slipknot66
02-26-2005, 02:34 PM
slc, yes i can share the scene, but i dont know how?! if you know of a ftp or something like that..?!
floze, yes the glow is photoshop, but there is no color correction or any other adjustment, here is the image without the glow effect and another one with the wireview. About the photons the only thing i had to do, was to decrease the intensity.
http://img216.exs.cx/img216/234/wireview9ia.th.jpg (http://img216.exs.cx/my.php?loc=img216&image=wireview9ia.jpg)

http://img237.exs.cx/img237/5856/noglow8pm.th.jpg (http://img237.exs.cx/my.php?loc=img237&image=noglow8pm.jpg)

twistedpoly
02-26-2005, 04:18 PM
tnx slipknot66, can you send it to this (nejc.polovsak@gmail.com) address.

Lately i'm studying interior lightning and i hope it will be helpfull.

Once again, thank you.

jooji
02-26-2005, 06:36 PM
slipknot66 thats a really neat render....mental Ray 3.4 is really reshaping things in here :thumbsup: ...if its not too much to ask would u be kind enough to forward your scene file on my email address as well (s_arooj_azam@hotmail.com). I am really looking forward to study your scene. Thankyou :rolleyes:

jooji
02-26-2005, 06:53 PM
slipknot66 thats a very neat render. MentalRay 3.4 is really reshaping things in here :thumbsup: . If its not too much to ask would u be kind enough to forward your scene file on this address as well (s_arooj_azam@hotmail.com). I am really looking forward to study your scene. Thankyou :rolleyes:

cpan
02-26-2005, 07:42 PM
or better post it somewhere on highend3d :)
oopss
you can zip that scene and attach-it directly here... indeed that scene helps a lot. It looks incredebly smooth and nice ( ofcourse there are some materials and textures that could be better but it looks fantastic)

T-R
02-26-2005, 07:54 PM
looks ver nice slip. Just wondering what unit size you set for your interior scenes. I would like to take a look at this scene as well because I would like to see the impact that the secondary bounces have. I have been trying lately but havent noticed much of a difference(even with 100 bounces!).

slipknot66
02-26-2005, 09:27 PM
Here is the link to the interior scene, thanks to Mr BillSpradlin:thumbsup:
http://www.billspradlin.com/files/interior.rar

T-R
02-26-2005, 10:03 PM
Thanks for sharing Slipknot66!

Now maybe im crazy but if I turn off seconday diffuse bounces I cant see the difference between the 2 renders. In which areas is this function suppose to have an effect on? Sorry I'm just trying to get a better understanding of this.

rwijaya
02-26-2005, 11:16 PM
dude, thx alot man

slipknot66
02-27-2005, 01:09 AM
T_R, what ive noticed, is that when you are using GI, the secondary diffuse bounce have a small effect over the illumination, but if you turn off the photons emission(GI), and use just FG, you will see a huge difference, when using the secondary diffuse bounce.

jooji
02-27-2005, 07:28 AM
I do agree with slipknot66 and T_R. After some experimentation i have noticed ...while using final gather with GI ....there is no secondary bounce diffuse. I think alias should have disabled it anywayz while using GI. In my opinion u dont really require a secondary bounce when using FG with GI. The secondary bounce feature is used when FG is lighting the scene standalone and there is no GI (indirect illumination). By the way slipknot66 thankyou very much for the scene file much appreciated :).

cpan
02-27-2005, 02:12 PM
hmmm, somehow this is logical since GI already makes the multibounce effect, and that's why with FG multibounce and GI enabled you won't see big differences than FG multibounce only. As soon as I get my maya 6.5 copy i'll do some speed tests to see what is more convenable.

twistedpoly
02-27-2005, 06:36 PM
well, i've done i a scene myself too today, using onyl final gather, and no GI.

Here is what i came up with: http://www.3delavnica.com/@script/galerija/goto.php?image=10382

2 point lights converted into mr area lights.
min/max radius left at zero
render time : 35 minutes on athlon 2.2ghz, 1gb ram.

rwijaya
02-27-2005, 06:46 PM
damn, 35 minutes is too long slc. the scene that slipknot attached, rendered in 15 minutes in my laptop.

ajsfuxor
02-27-2005, 11:18 PM
of course im still waiting on my copy of 6.5, but it looks like the secondary bounces featrue really darkens up the corners in renders, something which i thought was lacking in standard FG. It really brings the renders to life.

Still, i really want to see it in action with an animation. Anyone wanna take the challenge?
Another intersting comparison would be to render the same image with single bounce FG to show us what the major diff is for those without Maya 6.5.

Great renders guys http://www.cgtalk.com/images/icons/icon13.gif

slipknot66
02-28-2005, 12:35 AM
ajsfuxor, here are 2 images with and without the diffuse multi bounces. No GI, just FG
http://img188.exs.cx/img188/7772/fgbounces6df.jpg
http://img188.exs.cx/img188/2226/nofgbounce4ct.jpg

Sphere♥
02-28-2005, 02:04 AM
[...]secondary bounces featrue [...] i really want to see it in action with an animation. Anyone wanna take the challenge? [...]
Perhaps you want to see one of these nice interiors animated, but just in case you missed it I previously did a simple FG secondary diffuse bounces animation test (http://www.cgtalk.com/showthread.php?t=210291#post1994305).

ajsfuxor
02-28-2005, 02:11 AM
Just checked out the anim, very nice sphere. Acceptable time on those frames too.

It look like GI has quite a large effect on your render slipknot. Although maybe you toned down the FG to get acceptable render times for the test?

Cant wait to go through the scene when i get 6.5....counting the seconds....

twistedpoly
02-28-2005, 10:09 AM
any sugestions what to do on my long render time?

i guess it would help if i get into min/max radious and tune that up? or is the problem with two mr area lights with 8 samples each?

cpan
02-28-2005, 10:11 AM
both of them influence render time, but the area light sampling does the most.
Anyway, the mr docs says that using pointlight arealight is much faster than a spot arealight
also if ya don't know very good how to set the Adaptive Sampling see szalbocs tutorial.

floze
02-28-2005, 11:23 AM
Anyway, the mr docs says that using pointlight arealight is much faster than a spot arealight
May you point me to that chapter/location? I can hardly believe this.

cpan
02-28-2005, 11:27 AM
neither do I, but they say that the point light requires fewer samples than a spot one, have no idea why... search there cause I don't remeber where i've read this

floze
02-28-2005, 11:38 AM
Hmm.. I did search but couldnt find anything about it. The sampling/overall behavior is pretty much the same, except that spot arealights are terminable to the given cone angle -which should result in an optimized (since limited) calculation. Could someone share some experience on this?

MunCHeR
02-28-2005, 12:28 PM
s!c, maybe use spotlights instead of point lights, point lights are actually 6 spotlights anyway facing different directions, and area lights are expensive enough to calculate, my suggestion anyhow.:thumbsup:

Cheers

MunCH

T-R
02-28-2005, 04:20 PM
T_R, what ive noticed, is that when you are using GI, the secondary diffuse bounce have a small effect over the illumination, but if you turn off the photons emission(GI), and use just FG, you will see a huge difference, when using the secondary diffuse bounce.

OK I get it. I was a little thrown off thinking that the secondary bounces were the main source of fill. I think the thread title threw me off a little:curious:. Scene looks great and renders fast!

I might be wrong but the reason s!c scene renders so slow is because FG is his main source of fill and it has to calculate multiple bounces (assuming he has used a high number). So this process is going to take longer. In order to shave time off the render you would need to adjust FG radius, bsp, convert to .map files. But naturaly it would take longer because of the lighting approach.

Sphere♥
02-28-2005, 05:19 PM
[...]point lights are actually 6 spotlights anyway facing different directions[...]
:eek: I don't know about other methods, but coming from ray tracing that doesn't make any sense at all to me.

twistedpoly
02-28-2005, 05:37 PM
T_R i used trace depth of 4, just like slip did in his scene i think. i've adjusted min/max values also and got only something like 3 minutes less render time,but i think i could then decrease fg rays to around 50 only, got to try that out too,will see how will it look.

i'm gona also try using gi and fg together and we'll see than.

CGTalk Moderation
02-28-2006, 06:00 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.