Page 1 of 1

Blue noise in dark areas is difficult to remove.

Posted: Sat Jul 11, 2020 5:03 am
by DrOmalpalli
Hello,

I'm having trouble removing blue chroma noise in dark parts of the image. I first tried auto-profile, and then tried selecting just the area with the noise, but it appeared to make little difference.

I've attached 3 images: The Noise profile, the Temporal Settings and the Spatial Settings.

Any suggestions?

Thanks in advance.

Regards,
Rajnesh

Noise Profile:

Noise Profile
Noise Profile
Noise Profile.jpg (243.88 KiB) Viewed 2891 times

Temporal Settings:

Temporal Settings
Temporal Settings
Temporal.jpg (238.25 KiB) Viewed 2891 times

Spatial Settings:

Spatial Settings
Spatial Settings
Spatial.jpg (255.67 KiB) Viewed 2891 times

Re: Blue noise in dark areas is difficult to remove.

Posted: Mon Jul 13, 2020 9:30 pm
by NVTeam
I recommend to try preparing a noise profile using the method shown in this video tutorial. Once you build a good noise profile using that method, try additionally to increase the temporal filter radius. That may help.

Do not enable "Adapt to changing noise" if your clip shows the same noise from the beginning to the end.

Hope this helps,
Vlad

Re: Blue noise in dark areas is difficult to remove.

Posted: Fri Nov 13, 2020 10:51 am
by themaster
Neat video is not really suited for this kind of noise, instead use avisynth and the script below:

AVISource("MyVideo.avi")
assumetff()
ConvertToYV16(interlaced=true)
orig=last
ev=orig.assumetff().separatefields().selecteven()
od=orig.assumetff().separatefields().selectodd()
ev
ue_chroma = UToY(ev).blur(0,1.5).binomialblur(5).ttempsmooth(maxr=6,lthresh=150, strength=6).KNLMeansCL(d=3,a=8,s=2,h=6)
ve_chroma = VToY(ev).blur(0,1.5).binomialblur(5).ttempsmooth(maxr=6,lthresh=150, strength=6).KNLMeansCL(d=3,a=8,s=2,h=6)
YToUV(ue_chroma, ve_chroma)
MergeLuma(ev)
ev_filtered=last
od
uo_chroma = UToY(od).blur(0,1.5).binomialblur(5).ttempsmooth(maxr=6,lthresh=150, strength=6).KNLMeansCL(d=3,a=8,s=2,h=6)
vo_chroma = VToY(od).blur(0,1.5).binomialblur(5).ttempsmooth(maxr=6,lthresh=150, strength=6).KNLMeansCL(d=3,a=8,s=2,h=6)
YToUV(uo_chroma, vo_chroma)
MergeLuma(od)
od_filtered=last
interleave(ev_filtered,od_filtered)
assumefieldbased().assumetff().weave()

Note filters required: ttempsmooth, KNLmeans, variablur