new vray tutorial

[natty]

rhumm rhumm !!!

lololololol ... i thought the same ..

[studioDIM]

I think part 2 is only available if you sign up for a course - cant blame them though, its a lot of work to write that
the best tip in there I found is to only use 2-3 steps in the QMC sampler so min 4 max 6 instead of min 1 max 6

>> ok after extensive testing, that qmc setting does not make sense - its faster to do min 1 max 6 and the quality does not suffer which is what the document suggests

I totally agree about with that as well. I was using Min 1 max 10 with the to get my glossies nice and smooth, and to give the QMC sampler enough room to sample adaptively for good results.
After reading the PDF I tried Min 4 Max 6 settings and it increased my rendertimes with no visible difference in quality

Indeed, what would the poor guys at chaos have coded an adaptive sampler for, if people then tend to use it as a brute force tool?
It seems a common enough situation though, to just bash the settings up madly, and wait.
I'd rather spend more time investigating how the adaptive sampler works, myself...

Lele

[R_Cyph]

Indeed, what would the poor guys at chaos have coded an adaptive sampler for, if people then tend to use it as a brute force tool?
It seems a common enough situation though, to just bash the settings up madly, and wait.
I'd rather spend more time investigating how the adaptive sampler works, myself...

Lele

Maybe they like waiting?. I , however, am much more impatient
The Universal settings method also uses this same principal for those who are questioning our sanity

>> I just pulled this from the online Vray guide :

" Min subdivs - determines the initial (minimum) number of samples taken for each pixel. You will rarely need to set this to more than 1, except if you have very thin lines that are not captured correctly."

[priad]

This will be great if someone could definitly clarify this thing.

For me the resume of adaptive mean: less detail less rays more detail more rays, so if I understand well when you put the 1 / 6, 1 is very fast for plane thing without glossy or detail and 6 is more than good for molding (for example).
after that you have to catch the good balance between the two.

Someone could confirm or explain

[R_Cyph]

This will be great if someone could definitly clarify this thing.

For me the resume of adaptive mean: less detail less rays more detail more rays, so if I understand well when you put the 1 / 6, 1 is very fast for plane thing without glossy or detail and 6 is more than good for molding (for example).
after that you have to catch the good balance between the two.

Someone could confirm or explain

Yes, thats the pretty much it. Adaptivity simply means the ability change given different situations ( flat plane - needs few samples, crown moulding - needs more ) In the case of the univesal settings method. The Image sampler will use as many samples neccessary to reach the specified noise threshhold. If anybody can explain that better please jump in

[glyph]

yup - that's what I understood: min 1 and max a lot more - but then I read this tut and it recommends min value much closer to high value
I tested a scene I had had problems with (and which I'd gone to adaptive sub-ds for..) and for a while the qmc seemed better at 4/6

But then I started thinking (must remember to engage brain before fiddling) and looked again - so the upshot is the guys at cgarchitect need informing they are wrong on that point! - though I guess it depends a bit on what their qmc adaptive settings were

[cecil gittens]

I dont think we can say that the guys at CGarchitect are wrong, but we have to look at it in the context of edge enhancement vers time. If i render an image (and someone please test this) with qmc 4 6 with detail enhancement off in the IR map will render in less time if i rendered the same image with qmc 1 6 with detail enhancement on in the IR, using the default values. if you start increasing the subdiv to 0.4 you will see an increase in time.
I argree the using qmc 4 6 is almost the same as using fixed , but in the context of getting good egdes vers time i will go with qmc 4 6 and DE off.

[studioDIM]

of course, as detail enhancement is a primary QMC with a shortened ray path.
So you're adding a primary QMC GI pass of sorts to the render.
Nothing to do with AA in itself.

Lele

[ene.xis]

of course, as detail enhancement is a primary QMC with a shortened ray path.
So you're adding a primary QMC GI pass of sorts to the render.
Nothing to do with AA in itself.

Lele

i m so seconding this!