Page 1 of 1

Test sSpeed vs noise & rendering time

Posted: Fri Jul 15, 2005 11:19 am
by jomaga
What do you think about this?
Tom?
Image

Posted: Fri Jul 15, 2005 12:04 pm
by Mihai
3 and 4 look exactly the same to me.... :?:

Posted: Fri Jul 15, 2005 12:18 pm
by rivoli
um, i can't tell any difference either.

Posted: Fri Jul 15, 2005 12:28 pm
by tom
As you see in 1 and 2, with higher shutter speeds, it needs more time to reach same sampling level.
And it's all right it doesn't affect so much...
4th is 0.5% better then 3rd if you avarage and compare pixel values.
But I didn't understand what settings you used for 4th exactly.
What do we have to understand from last 2 red lines, seems confusing.

Posted: Fri Jul 15, 2005 1:37 pm
by jomaga
This is what I´m trying to do:
Image 3
1. Set up the model in max with fStop 4 / sSpeed 30
2. Render
Image 4
1. Set up the model in max with fStop4 / sSpeed 500
2. Render until Maxwell reaches sLevel 13.24
3. Change sSpeed to 30 in the last sLevel pass
4. the image turn brighter
5. sLevel 14.51 reached with sSpeed 30

Images 4 & 5 are almost the same, but not exactly the same. They should be equal?
I´m trying to seek for a save of rendering time doing this, but the differences are minimal. But noise is perhaps a bit different?

Cheers

Posted: Fri Jul 15, 2005 5:34 pm
by albertMaxwell
Hi Jose,

Rendering 3th with shutterspeed 30 and get the image 3
Rendering 4th with shutterspeed 500 and then 30 is the same than rendering 3th. It only matter the final preview values.

The very little difference of noise between 3 and 4 is due a little difference of sampling level reached. With equal S.L both were identical.