Page 10 of 12
Port Cube
Posted: Thu Sep 22, 2022 7:19 am
by Rocky
Finding #5: Frame properties
When using std-b67, the _Transfer frame property is 18. When using st2084, the _Transfer frame property is 16. All other properties are the same.
So we will focus initially on the last stage and seek out the reason for the discrepancy. I'll have to bring up avsresize in the debugger. That will take a little time to get going.
Port Cube
Posted: Thu Sep 22, 2022 11:53 am
by Curly
@Guest 2
Just for fun, add this parameter to all z calls and see if things go faster.
approximate_gamma=true
And modify avsresize to support CUDASynth protocol. It's not difficult.
Port Cube
Posted: Thu Sep 22, 2022 1:24 pm
by Guest 2
Curly wrote: ↑Thu Sep 22, 2022 11:53 am
Just for fun
Will try, thanks!
Curly wrote: ↑Thu Sep 22, 2022 11:53 am
And modify avsresize to support CUDASynth protocol. It's not difficult.
Port Cube
Posted: Thu Sep 22, 2022 3:11 pm
by Rocky
Curly wrote: ↑Thu Sep 22, 2022 11:53 am
And modify avsresize to support CUDASynth protocol. It's not difficult.
No. To gain anything you need to port z algorithms to CUDA.
Port Cube
Posted: Wed Sep 28, 2022 3:24 pm
by Guest 2
Rocky wrote: ↑Thu Sep 22, 2022 3:11 pm
No.
Hello? Everything is fine?
Port Cube
Posted: Wed Sep 28, 2022 4:01 pm
by Bullwinkle
Rocky is ill and may not get back to coding for some days.
Any luck with the approximate_gamma hack?
Port Cube
Posted: Wed Sep 28, 2022 4:06 pm
by Guest 2
Port Cube
Posted: Fri Sep 30, 2022 11:08 am
by Bullwinkle
Thank you, big t.
BTW, I slipstreamed a change to DGCube to map negative values to 0. People should not create cubes with negative values but I'll be nice and try to help them. Redownload to get the change.
Any luck with the approximate_gamma hack?
Port Cube
Posted: Sat Oct 01, 2022 1:08 pm
by Guest 2
Bullwinkle wrote: ↑Fri Sep 30, 2022 11:08 am
BTW, I slipstreamed
I was about sending you the url of the message
Bullwinkle wrote: ↑Fri Sep 30, 2022 11:08 am
Any luck with the approximate_gamma hack?
I need to empty the encoding queue before trying. Will do ASAP.
Port Cube
Posted: Tue Oct 04, 2022 10:20 am
by Guest 2
I had some time to test approx gamma.
UHD: Thor – Love and thunder, mixed frames (1722 frames in total)
Script:
Code: Select all
SetMemoryMax()
LoadPlugin("D:\Eseguibili\Media\DGDecNV\DGDecodeNV.dll")
LoadPlugin("D:\Eseguibili\Media\DGCube\DGCube.dll")
DGSource("F:\In\1_58 Thor Love and thunder\thor.dgi",ct=280,cb=280,cl=0,cr=0)
propClearAll()
z_ConvertFormat(pixel_type="RGBP16", colorspace_op="2020:st2084:2020:limited=>rgb:st2084:2020:full", resample_filter_uv="spline64", dither_type="error_diffusion")
DGCube("D:\Programmi\Media\AviSynth+\cube\1a_PQ1000_HLG_mode-nar_in-nar_out-nar_nocomp.cube", in="full", lut="full", out="full")
z_ConvertFormat(pixel_type="YUV420P10", colorspace_op="rgb:std-b67:2020:full=>2020:std-b67:2020:limited", resample_filter_uv="spline64", dither_type="error_diffusion")
fmtc_bitdepth (bits=10, dmode=8)
Prefetch(8)
Code: Select all
AVSMeter64.exe "Thor 4K HLG BBC ext.avs"
AVSMeter 3.0.8.0 (x64), (c) Groucho2004, 2012-2021
AviSynth+ 3.7.3 (r3689, 3.7, x86_64) (3.7.3.0)
Number of frames: 1722
Length (hh:mm:ss.ms): 00:01:11.822
Frame width: 3840
Frame height: 1600
Framerate: 23.976 (24000/1001)
Colorspace: YUV420P10
Frames processed: 1722 (0 - 1721)
FPS (min | max | average): 0.839 | 322580 | 16.01
Process memory usage (max): 2840 MiB
Thread count: 27
CPU usage (average): 71.5%
Time (elapsed): 00:01:47.571
Replacing the cube part with
Code: Select all
z_ConvertFormat(pixel_type="RGBP16", colorspace_op="2020:st2084:2020:limited=>rgb:st2084:2020:full", resample_filter_uv="spline64", dither_type="error_diffusion",approximate_gamma=true)
DGCube("D:\Programmi\Media\AviSynth+\cube\1a_PQ1000_HLG_mode-nar_in-nar_out-nar_nocomp.cube", in="full", lut="full", out="full")
z_ConvertFormat(pixel_type="YUV420P10", colorspace_op="rgb:std-b67:2020:full=>2020:std-b67:2020:limited", resample_filter_uv="spline64", dither_type="error_diffusion",approximate_gamma=true)
I get
Code: Select all
Frames processed: 1722 (0 - 1721)
FPS (min | max | average): 0.606 | 344828 | 16.62
Process memory usage (max): 2697 MiB
Thread count: 27
CPU usage (average): 70.5%
Time (elapsed): 00:01:43.608
Not so much to risk issues with gamma approx.
Port Cube
Posted: Wed Oct 05, 2022 12:52 pm
by Rocky
OK, nowhere near the speedup hoped for. Thank you for your testing.
Port Cube
Posted: Wed Oct 05, 2022 1:48 pm
by Guest 2
Rocky wrote: ↑Wed Oct 05, 2022 12:52 pm
OK, nowhere near the speedup hoped for. Thank you for your testing.
Now that you have stocked with acorns, I bet you will pull a rabbit out of the hat.
Port Cube
Posted: Wed Oct 05, 2022 5:43 pm
by Rocky
Ha ha, clever frog. Croak!
Port Cube
Posted: Thu Oct 06, 2022 5:32 pm
by Rocky
Completed my last coaching cert today and Sherman is taking over the radios. I'll try to make sense of z source code to extract the matrices they are using for comparison to mine.
Port Cube
Posted: Mon Oct 10, 2022 8:52 am
by Guest 2
Rocky wrote: ↑Thu Oct 06, 2022 5:32 pm
I'll try to make sense of z source code to extract the matrices they are using for comparison to mine.
Eager to test your wonders.
Port Cube
Posted: Mon Oct 10, 2022 1:07 pm
by Rocky
Me too.
Port Cube
Posted: Tue Oct 11, 2022 9:39 am
by Rocky
Ooh, I think I found the place in the code where the coefficients are generated. Let's see if I can hit it in the debugger.
Port Cube
Posted: Tue Oct 11, 2022 3:34 pm
by Guest 2
Rocky wrote: ↑Tue Oct 11, 2022 9:39 am
Ooh, I think I found the place in the code where the coefficients are generated.
Remember that conversion from PQ to SD works fine, it's the HLG part that has issue.
Port Cube
Posted: Tue Oct 11, 2022 3:55 pm
by Rocky
I don't follow that because the HLG is part of the 3D LUT. I've shown the discrepancies for the two operations before and after the 3D LUT application, i.e., the YUV -> RGB and RGB -> YUV conversions.
Port Cube
Posted: Tue Oct 11, 2022 3:56 pm
by Guest 2
I was reading
https://github.com/WolframRhodium/VapourSynth-BM3DCUDA and I noticed:
Code: Select all
fast:
Multi-threaded copy between CPU and GPU at the expense of 4x memory consumption.
Default True.
Perhaps you can find some trick in the source to increase transfer speed.
P.S: I really would like to see its AVS+ native by your talented hands.
Port Cube
Posted: Tue Oct 11, 2022 7:11 pm
by Sherman
And now, it's time for a message from our sponsor.
Port Cube
Posted: Tue Oct 11, 2022 7:15 pm
by Britney
Port Cube
Posted: Tue Oct 11, 2022 7:18 pm
by Baltasar
Port Cube
Posted: Tue Oct 11, 2022 7:21 pm
by Albert
Port Cube
Posted: Tue Oct 11, 2022 7:26 pm
by Boris