MCTemporalDenoise [v1.4.20 - Update 2010/07/02]

[foxyshadis]

Curious, if it's titled Ultimate Degrain, why not just call it UltimateDegrain() instead of UDegrain? UDegrain kind of connotates something that works with the chroma channels.

[LaTo]

Originally Posted by foxyshadis

Curious, if it's titled Ultimate Degrain, why not just call it UltimateDegrain() instead of UDegrain? UDegrain kind of connotates something that works with the chroma channels.

Ultimate? Xtrem? Awesome? Of Death?

UXADdegrain

if someone have a better name...

(MCdegrain/MCttempsmooth?)

[sarmano]

LaTo
Good job!
This script much better then another in public zone, but a few moment:
- too slow on heavy mode
- didn't working with 1080p video with radius more then 2 (with x264)

[LaTo]

Originally Posted by sarmano

- didn't working with 1080p video with radius more then 2 (with x264)

maybe you don't have enough GPUram (memory) for 1080p, try with gpu=false...

from fft3dgpu readme:

Q: What does it mean when I get a popup box Unexpected error encountered with Error Code: D3DERR_OUTOFVIDEOMEMORY.

A: It means that fft3dgpu needs more memory than there are availebol on the graphics card. So either you will have to upgrade or try lowering the resolution, precision, bt,bh,bw,ow,oh or use usefloat16=true or mode 0 or 2

[LaTo]

Originally Posted by martino

I think that's just how MVDegrain works. Try to apply it on some anime with scenes from space and you'll see most of the stars gone.

can you provide a sample to test this?

[martino]

http://www.mediafire.com/?ve3lxmm9j1f (50MB, telecined)

Enjoy.

[Ranguvar]

Feature request, a simple parameter to control how many motion vectors are used Also parameters for controlling bt, precision, and ncpu in FFT3DFilter/FFT3DFilter, but those are easy to set manually.

[LaTo]

v7.29a:
added star & bright points protection (default: off, see "protect")
changed some default options
added more options for FFT3D
added YV12 checking

For star & bright points protection:

Source
Protect=false (default)
Protect=true

I haven't tested much this option (only on this source), so default settings need to be tweaked a little... I haven't be abble to reproduce the star loss with martino's sample, maybe with very high settings
Moreover I have not watched at MC_Spuds starfield parameter, maybe it's different but it works well (screenshot above)

LaTo,

I see you added the temporal Min/Max limiting to the sharpening. I don't think you should limit with both "contra" and the temporal min/max, it really should be one or the other. Otherwise, after "contra" limiting, the temporal min/max limit rarely does anything because the contra limits much more than the min/max.

-G

[Nikos]

LaTo, with MoComp'ed Contra-Sharpening we must use advanced sharpening routines, not just the simple 3x3 kernel Contra-Sharpening.
Read Didie's long explanation here

In my opinion, the Smode 4 from LSF is more suited.

strength = default(strength, 100)
Str=string(float(strength)/100.0)
mt_lutxy(clp, clp.RemoveGrain(11,-1), "x y == x x x y - abs 16 / 1 2 / ^ 16 * "+Str+"
 \  * x y - 2 ^ x y - 2 ^ "+Str+" 100 * 25 / + / * x y - x y - abs / * + ?")

In any case thanks for your hard work.

[LaTo]

Originally Posted by Nikos

LaTo, with MoComp'ed Contra-Sharpening we must use advanced sharpening routines, not just the simple 3x3 kernel Contra-Sharpening.
Read Didie's long explanation here

In my opinion, the Smode 4 from LSF is more suited.

strength = default(strength, 100)
Str=string(float(strength)/100.0)
mt_lutxy(clp, clp.RemoveGrain(11,-1), "x y == x x x y - abs 16 / 1 2 / ^ 16 * "+Str+"
 \  * x y - 2 ^ x y - 2 ^ "+Str+" 100 * 25 / + / * x y - x y - abs / * + ?")

In any case thanks for your hard work.

Yes, it was already on my todo list... Thanks

[Honeyko]

For a real torture test, try it on this clip.

...Can it remove all the grain in the second scene without introducing artifacts in the dark color-gradient area over the guy's left shoulder in the first scene?

[LaTo]

Originally Posted by Honeyko

For a real torture test, try it on this clip.

...Can it remove all the grain in the second scene without introducing artifacts in the dark color-gradient area over the guy's left shoulder in the first scene?

udegrain(radius=3,settings="heavy") remove most of the grain...

[Honeyko]

Originally Posted by LaTo

udegrain(radius=3,settings="heavy") remove most of the grain...

After finishing, is there any garbage over the guy's left shoulder in the first scene?

(How about radius=7? That is some fat-ass grains.... ;-)

[LaTo]

Originally Posted by Honeyko

After finishing, is there any garbage over the guy's left shoulder in the first scene?

(How about radius=7? That is some fat-ass grains.... ;-)

I don't know, test by yourself...

[LaTo]

v8.08a:
added mocomp contra-sharpening (basic contra-sharpening always usable) -> see readme for usage

[Nikos]

In my opinion the s4R = (post==2) ? repair(s4,smB,1,-1,-1).Mergechroma(s4) : NOP() line is not suited because now we have almost the same effect as simple contra-sharpening.
If you want to repair a little try another mode, not the 1.
The soothe(xx) is more appropriate here.

[LaTo]

Originally Posted by Nikos

In my opinion the s4R = (post==2) ? repair(s4,smB,1,-1,-1).Mergechroma(s4) : NOP() line is not suited because now we have almost the same effect as simple contra-sharpening.

fixed.

[Nikos]

Lato according to Didie's explanation i think for the "mocomp contra-sharpening" radius=2 it's enough.

[buzzqw]

Hi LaTo

i would like to add your Udegrain script to automkv, so first of all i will ask permission to include

second question: could you suggest 4 default parameters usage (light grain/medium/heavy/sharp/... whatelse this script is capable...) ?

thanks

BHH