Overexposing vray render to reduce noise

[Crayox13]

But, Morbid, the thing is that 3, 6 gave me sufficient results when I wasn't using LWF, but with LWF dark areas were to noisy, again, with gamma 2.2 and (adaptation only) checked. So it seams that there is a certain problem with this. I'll let you know what Vlado comes up with, after seeing my scene.

[rivoli]

So it seams that there is a certain problem with this.

I don't think there's a problem as such. slamming the gamma obviously brightens previously really dark areas, so you can actually seen what's going on there.
this is something that you do really often when compositing digital elements with real footage, when you have to match film grain.

[Crayox13]

I don't think so. This is my thread about it, and you can see what should be "normal" noise, here it is very bad

http://www.chaosgroup.com/forums/vbu...ad.php?t=43714

LWF should give the same results with dark colors/materials, because you check (adaptation only) and set the gamma.

[rivoli]

I remember the thread, and I still don't see any bug or any way in which vray doesn't take gamma into account. I guess it's a matter of finding the right sampling, what works best in terms of time and quality. and I don't find it any odd that what works for non gamma corrected images might be off when switching to linear.

[Crayox13]

Since gamma gives a non linear values across 0 - 255, you can't simply convert the values but for instance - you have a dark gray color that you used to use for chrome materials, or you have a dark gray matte metal surface. Before, you put something like 25,25,25 RGB, and now with LWF you put 5,5,5 and in the end, both will get you a similar looking material/object in the scene. I mean similar dark gray color. So in LWF I won't get a almost black with 5,5,5. That's why vray has that 2.2 gamma for adaptation only option which should make vray see the new 5,5,5 (noise threshold concerning) like it used to see 25,25,25. It kinda doesn't. That's why I reported this. Will see what will Vlado think.

[sv]

It kinda doesn't. That's why I reported this. Will see what will Vlado think.

Would be interested to see, because in my experience there is no problems with Vray's sampling when using the adaptation only setting...

As rivoli says it's about finding the right sampling for each scene...I too can't see any problems with that link, especially your only using 6 dmc samples on your AA...

[dejaVu]

I also believe there is an issue with adaption only gamma sampling.
In some quick tests, the sampling of noise in dark areas is no better.
Could it be that it is not working properly?

[^Lele^]

gammaed sampling works a charme here.
Of course, it helps naught with overbrights in glossies.
That's still down to user massage

[Crayox13]

Yes, but like we said, very dark surfaces, of up to say 8,8,8 give problems.