stupid bruteforce question... global mult very different then increasing BF subdivs..

[Subburb]

Hello !

was playing with engine today (3.00.05) to find a good way for BF/LC or BF/BF renderings.
btw i figured a huge difference when using LC retrace with BF/LC method. Impressive.

here's the question :

if i leave the BF subdivs to default 8 and raise the global multiplier to, let's say, 8, i excepected a BF subdivs at 64.
But in fact the result is VERY different if if set the BF subdivs to 64 and let the global multiplier to 1.0.
There's something important here i missed, for sure.
Someone can explain me ?

thanks in advance,

[vlado]

8 BF subdivs = 64 rays times 8 = 512 rays max

64 BF subdivs = 4096 rays times 1 = 4096 rays max.

Best regards,
Vlado

[Subburb]

doh ! got it !
multiplication opposed to square... sounds very logical.

Huge thanks, Vlado !

[bennyboy]

Math. Love it.

[RockinAkin]

Math. Love it.

Might help for those who don't know about it: DMC Calculator

[Subburb]

Sounds like it's same behavior for irrmap subdivs.

if you want 100 Hsph subdivs then you need to put 71 with globl mult = 2. (71² * 2 = 10000 approx). And not 50, wich is bad value is a bad value. Right ?

was same in older Vray versions (hsph) ?

[vlado]

was same in older Vray versions (hsph) ?

Yes, it was exactly the same.

Best regards,
Vlado

[Subburb]

thanks for fast answer.
I have some things to rewrite now

[Morne]

8 BF subdivs = 64 rays times 8 = 512 rays max

64 BF subdivs = 4096 rays times 1 = 4096 rays max.

Best regards,
Vlado

Is that also for material and light subdivisions?
ie:
8 reflect subdivs = 64 rays times 8 = 512 rays max
64 reflect subdivs = 4096 rays times 1 = 4096 rays max.
?

[RockinAkin]

Is that also for material and light subdivisions?
ie:
8 reflect subdivs = 64 rays times 8 = 512 rays max
64 reflect subdivs = 4096 rays times 1 = 4096 rays max.
?

Yes - it's the same for Light, GI, Material, Fog subdivisions throughout your scene.

Subdivsions² * Global Subdivs Multiplier

[Dariusz Makowski (Dadal)]

Don't forget its also divided by your maximum AA samples and controlled by adaptive amount in dmc settings...

[3DMK]

Don't forget its also divided by your maximum AA samples and controlled by adaptive amount in dmc settings...

And is also affected by the orbit of the moon relative to Saturn.....that's where my brain goes when people start talking maths....
[Sorry to clutter the post with my ramble..it's Friday & I need to laugh..we all do]
cheers

[^Lele^]

Vlado, in that respect, the wording for it is somewhat misleading: if it was a SUBDIVS multiplier, the maths for obtaining rays would have to be (subdivs*GSM)^2.
In fact, as you pointed out, it's (subdivs^2)*GSM, so it multiplies the rays, not the subdivs.
Would you say it's worth a re-wording, at all?

[RockinAkin]

Vlado, in that respect, the wording for it is somewhat misleading: if it was a SUBDIVS multiplier, the maths for obtaining rays would have to be (subdivs*GSM)^2.
In fact, as you pointed out, it's (subdivs^2)*GSM, so it multiplies the rays, not the subdivs.
Would you say it's worth a re-wording, at all?

Agreed.

Global Subdivs Multiplier should be Global Samples Multiplier
Min Samples should be Min Adaptive Samples
Min Shading Rate should be moved from the Image Sampler group to the Global DMC group, as it has no effect on Image Sampling whatsoever... and perhaps should be named better as well.
At least they fixed the 'Time Dependent / Independent' madness.

...or just get rid of it all and allow us to directly set the DMC Sampler min and max.

[vlado]

Vlado, in that respect, the wording for it is somewhat misleading: if it was a SUBDIVS multiplier, the maths for obtaining rays would have to be (subdivs*GSM)^2. In fact, as you pointed out, it's (subdivs^2)*GSM, so it multiplies the rays, not the subdivs. Would you say it's worth a re-wording, at all?

This is correct, yes; I agree it will make more sense to change it.

Best regards,
Vlado