Experiment: Noise levels

[Siahpoosh]

look at this

[elindell]

1 also feels cleaner if I scrub back and forth between them, but honestly just eyeballing it they look of equal noisiness

[maru]

I think the two images are extremely similar in terms of noise but if I had to choose, I would pick #1. Looks like it has (at least perceptually) less noise in the direct lighted area near the window. But I wouldn't be surprised if Keymaster was testing us and the two images were identically noisy. ;)

[Ondra]

#1 is less noisy. It uses one trick from recent paper about sampling. It will be probably included in daily build soon

[Siahpoosh]

"LtWCJTK.png" is marginally noisier. I measured. It's something like 2.4 % more noisy.

edit: So that's image 1 that's cleaner.

How did you measure it without the reference image? If you have some reliable metric for estimating noise tell me, I can use it for adaptivity ;)

Maybe this could help :
http://homepages.inf.ed.ac.uk/rbf/CVonline/LOCAL_COPIES/VELDHUIZEN/node18.html

[Chakib]

I can see that the first image had a liiiittle bit less noise  than the second one, especially on the back of the chair.

[zzubnik]

Well, I guess I'm going blind. I really struggled to make out much overall difference. I think I need glasses!

[filipskrzat]

#1 is less noisy

[juang3d]

Awesome :D

Will this make the images look cleaner sooner?

Cheers

[cecofuli]

Maybe... are you working also on "noise pattern flickering" problem?  =^___^=

[Captain Obvious]

"LtWCJTK.png" is marginally noisier. I measured. It's something like 2.4 % more noisy.

edit: So that's image 1 that's cleaner.

How did you measure it without the reference image? If you have some reliable metric for estimating noise tell me, I can use it for adaptivity ;)

I just measured the maximum deviance between each pixels and its neighbours. It doesn't work for calculating the error in a single image (you need the reference image for that), but it works for checking which of two images have the greater variance.

[gabrielefx]

try to reduce the noise putting 50 ies or tiny emitters inside the room.

[Zom-B]

#1 is less noisy. It uses one trick from recent paper about sampling. It will be probably included in daily build soon

There is a very simple way to tell what image has less noise: File size for the PNG!
PNGs are super mega sensible to noise vs filesize ratio.

[cecofuli]

try to reduce the noise putting 50 ies or tiny emitters inside the room.

Can you explain better?

[boumay]

try to reduce the noise putting 50 ies or tiny emitters inside the room.

Yes it would be interesting to know more