VRay-like interior renders with mental ray???

#501

dragon, so maya doesn’t make adaptive sampling with per obj sampling? I’ve tested it in maya 6.0 and the results were faster and better-looking than the usual global settings.

BTW here’s the script for maya 6, as I promissed :slight_smile:

#502 
 --> 2min 57sec (Athlon XP 3200)
 [[img]http://img153.exs.cx/img153/1739/roomtestthgion012ix.jpg[/img]](http://www.imageshack.us)

Thanks to ImageShack for Free Image Hosting

#503

dude…thx for a AIR demo but…posting a pic with no caustics…way less glass objects and way less agressive reflective objects should not be a way to cut rendering times when it comes to comparsion. Another thing…nice idea to add some colors and textures but it makes really hard to compare direct shading. You are bringing in too many different values so that it becomes pretty impossible to compare rendering engines. Sry.

#504

agree with you Splin and beside the mray render looks way realistic than the AIR one. Also the image resolution is about double in the mray image.

Even with maya software would look better
just joking :slight_smile:

Ok here’s a question:

I’ve read somewhere (maybe mray docs, don’t remeber) that enabling GI will disable FG multibounce cause the photons already bounce. Now, is FG multibounce creating caustic effects?
Another thing I’ve noticed is that the FG multibounce doesn’t have something like an exponent to set(like quadratic, linear, etc) and it seems that it’s set to linear, if I increase the number of bounces the images get whiter and whiter and this isn’t how this should be. Is there an setting to correct this? or maybe it’s an dynamic attribute, who knows…

cheers.yashu.

#505

and way less agressive reflective objects

You ment to say way more reflective objects !

a. The entire floor surface area is reflective
b. The screen is reflective
c. The fixture and the sphere on the table are reflective
d. The stereo has some reflection.

The aggressive aspect is irrelevant for AIR. Once reflections are tured on for an object it takes the same amount time regardless of its strength.

The only thing different is caustics because the demo version is limiting.

(If you look back in this thread there were other samples with no materials applied)
(The size is limited in the Demo version to 640X480, I did not shrink it on purpose)

#506

I’ll try to explain the object min max samples issue in Maya 6 and Maya 6.5.
When rendering from within Maya, the object max samples value takes on the min samples value.

Ex. case: Say you wanted to increase the minimum samples value for a certain object.
Your whole scene is samples -1 2 (min max), and you wanted to increase your samples on this object to 1 2 (min max). Well, from within Maya, it would be passed to an interactive render as 1 1. Typically, this would mean that your object won’t get enough samples in its higher contrast parts. However, if you wrote out to mi file, and rendered standalone, it would use the correct values.

If rendering interactively, I’d recommend fixing the object samples (eg, 2 2) until this is remedied.

When you are learning about this stuff, its always useful to use some simple objects and the diagnostic samples flag.

Also, see the Object Samples section at the end of this tips page on samples: http://www.lamrug.org/resources/samplestips.html

#507

it look like a nice occlusion render, but i dont see GI shadows in this render
i think air is very fast for this type of rendering, i remember my test with occlusion in this scene take 3/4minutes, but with only diffusive materials!
what i want to see is a comparaison with my last render, with caustics, glossy reflections and gi shadows… to see the quality of air, not only the great rendertime :wink:

for the multibounces
if u turn on the GI the multibounces doesn’t work, but only the diffusive multibounces
the bounces in reflection and in refraction works fine…
in this render i’ve used only caustics photons and fg multibounces, not gi
for the multibounces exponent, i think u have to increase the ray depth of your raytrace light (in the attribute of the light and in the renderglobal) to balance the diffuse depth of fg

ciau

mat

#508

Ok nice to know ThomasAn, thx! That makes the picture a bit more clear!

I may be hitting a blind spot here. yashu, test out if it works what dragon suggested. Otherwise I can recall that dgs materials should be used in case of GI cause that they transmit light energy way better than normal phong, lamberts etc. Is this also the case with new mulltibounce FG?

#509

dagon1978: I’d be interested if you had a performance comparison between Maya6 and Maya6.5. In mr 3.4 in Maya6.5, you should be able to use less FG samples/accuracy for the same quality. Hence, the render should be faster.

#510

Please browse throught the thread first…as I recall there are many examples with that teapot scene a while ago. I agree that it is tedious process but you will learn alot :wink:

#511

Quote:
Originally Posted by bgawboy
dagon1978: I’d be interested if you had a performance comparison between Maya6 and Maya6.5. In mr 3.4 in Maya6.5, you should be able to use less FG samples/accuracy for the same quality. Hence, the render should be faster.

Please browse throught the thread first…as I recall there are many examples with that teapot scene a while ago. I agree that it is tedious process but you will learn alot

Splin : You are don’t know who your talking to, bgawboy is not a person who need to learn MR more than you do.

bgawboy : Sorry missed the user meeting last time, my schedule are full on tuesday, and no one from the studio has free time that day. :slight_smile:

#512

I did browse through the thread, and just again. The non-beta Maya 6.5 has only been out a short time. Sorry, I guess I’m missing the information.

I’m interested in exactly the scene just above, not the teapots, only using different FG accuracy. I thought it would be nice to have a post with this same scene and two compared times together, using released versions of 6.0.1 and 6.5.

However, now that I think about it, it may not be an apples-to-apples comparison visually, because of the differences in FG multibounce between the two releases.

#513

rw: Thanks so much for the support, but I can always learn:-)

splin: Indeed, I did get more out of it, by reviewing.

dagon1978: I want to make FG the focus of the next meeting. Is it possible I could get a hold of that scene as you’ve tuned it? I’m also going to test with a satellite.

#514

Yes, more reflective, but not glossy reflective :slight_smile: Simple reflections like that take nothing in mr too, and like air, it’s not dependant on the “strengh” of the reflection., the ray is throw, you are free to decrease or increase the result :slight_smile:

No render war here, but your comparaison is irrelevant (as a reply for dagoon test)

#515

Yes, more reflective, but not glossy reflective

Not sure what this means.
Could you show me an example of “glossy reflective” ?
(and how is it different than the floor surface as shown) ?

thx

#516

bgawboy are you a lamrug member?
sorry i really don’t understand all, because my english is very limited
by the way,this is a thread in treddi.com that i’ve started with maya 6.0.1, but i make the initial test obviously with gi+fg
if u want i can send u my scene or my settings :wink:

ops

the link
http://www.treddi.com/forum/index.php?showtopic=2608

#517

look at this one

the floor, the screen frame and the stereo have glossy reflections

and the table too (but it’s too dark too see it :stuck_out_tongue: )

oh and the heaters too :wink:

#518

oh and this is my last test with multibounces :slight_smile:

i’ve solved the problem of the glass using the reflective/refractive bounces of final gather the same of reflective/refractive raytrace depth

#519

But where are the glossy reflections that are supposed to be different than the floor surface in the AIR image ?

I honestly like to know the distinction.

#520

in the air image i dont see glossy reflections, u can see it?