Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Video Encoding > MPEG-4 Encoder GUIs

Closed Thread
 
Thread Tools Search this Thread Display Modes
Old 20th January 2020, 09:21   #1321  |  Link
neo_sapien
Registered User
 
Join Date: Jan 2002
Location: USA
Posts: 249
Quote:
Originally Posted by Patman View Post
Change your x265.exe with a version from here and test again.
I tried using 3.2.1+3-b4b2eca VS 2019 AVX2 from the stable branch. I encountered this error. (I'm trying to modify the below text to keep from breaking the tables too bad)

Code:
--------- Error Video encoding using x265 3.2.1+3-b4b2ecac21f6 ---------

Video encoding using x265 3.2.1+3-b4b2ecac21f6 failed with exit code: -1073741674 (0xC0000096)

The exit code might be a system error code: {EXCEPTION}
Privileged instruction.


------------ Video encoding using x265 3.2.1+3-b4b2ecac21f6 ------------

C:\StaxRip-x64-2.0.6.2-beta\Apps\Support\avs2pipemod\avs2pipemod64.exe -y4mp "video_input.avs" 
| C:\StaxRip-x64-2.0.6.2-beta\Apps\Encoders\x265\x265.exe --crf 20 --preset slow --level-idc 4.1
 --output-depth 10 --csv-log-level 1 --ssim --psnr --frames 200000 --y4m --output "video_output.hevc" -

avs2pipemod[info]: writing 200000 frames of 24000/1001 fps, 1904x792,
sar 0:0, YUV-420-planar-8bit progressive video.
y4m  [info]: 1904x792 fps 24000/1001 i420p8 unknown frame count
raw  [info]: output file: video_output.hevc
x265 [info]: HEVC encoder version 3.2.1+3-b4b2ecac21f6
x265 [info]: build info [Windows][MSVC 1924][64 bit] 10bit
x265 [info]: using cpu capabilities: MMX2 SSE2Fast LZCNT SSSE3 SSE4.2 AVX FMA3 BMI2 AVX2
x265 [warning]: --psnr used with psy on: results will be invalid!
x265 [warning]: --tune psnr should be used if attempting to benchmark psnr!
x265 [warning]: Specifying a decoder level with constant rate factor rate-control requires
x265 [warning]: enabling VBV with vbv-bufsize=50000kb vbv-maxrate=50000kbps. VBV outputs are non-deterministic!
x265 [info]: Main 10 profile, Level-4.1 (High tier)
x265 [info]: Thread pool created using 16 threads
x265 [info]: Slices                              : 1
x265 [info]: frame threads / pool features       : 4 / wpp(13 rows)
x265 [info]: Coding QT: max CU size, min CU size : 64 / 8
x265 [info]: Residual QT: max TU size, max depth : 32 / 1 inter / 1 intra
x265 [info]: ME / range / subpel / merge         : star / 57 / 3 / 3
x265 [info]: Keyframe min / max / scenecut / bias: 23 / 250 / 40 / 5.00
x265 [info]: Lookahead / bframes / badapt        : 25 / 4 / 2
x265 [info]: b-pyramid / weightp / weightb       : 1 / 1 / 0
x265 [info]: References / ref-limit  cu / depth  : 4 / on / on
x265 [info]: AQ: mode / str / qg-size / cu-tree  : 2 / 1.0 / 32 / 1
x265 [info]: Rate Control / qCompress            : CRF-20.0 / 0.60
x265 [info]: VBV/HRD buffer / max-rate / init    : 50000 / 50000 / 0.900
x265 [info]: tools: rect limit-modes rd=4 psy-rd=2.00 rdoq=2 psy-rdoq=1.00
x265 [info]: tools: rskip signhide tmvp strong-intra-smoothing lslices=4
x265 [info]: tools: deblock sao
avs2pipemod[info]: finished, wrote 24000 frames [12%].
avs2pipemod[info]: total elapsed time is 12446.648 sec.
avs2pipemod[error]: only wrote 24000 of 200000 frames.

Last edited by neo_sapien; 20th January 2020 at 09:26. Reason: prevent breaking tables
neo_sapien is offline  
Old 20th January 2020, 18:15   #1322  |  Link
Patman
Registered User
 
Patman's Avatar
 
Join Date: Jan 2015
Posts: 286
Quote:
Originally Posted by neo_sapien View Post
I tried using 3.2.1+3-b4b2eca VS 2019 AVX2 from the stable branch. I encountered this error. (I'm trying to modify the below text to keep from breaking the tables too bad)
Maybe a problem of avs2pipemod... go to x265 options, other and set piping tool to none... test again... really strange
__________________
Tools for StaxRip | x264 - x265
Patman is offline  
Old 20th January 2020, 22:27   #1323  |  Link
LazyNcoder
Registered User
 
Join Date: Feb 2015
Posts: 33
Hello guys,
I hope you can help me here.

What I'm trying to do is, to create a blank video in length of the %Video_Delay% and add it to the start of the encode.

I mean, if the source video has 5000 ms delay, I create this 5000 ms delay as blank black video.

I'm using Vapoursynth. What I already have is this:
clip2 = core.ffms2.Source(r"%source_file%", cachefile = r"%source_temp_file%.ffindex")
clip1 = core.std.BlankClip(clip2,length=%Video_Delay%)
clip = clip1+clip2


I know it has huge problems. So, What should I do here?

I also need to change the delay which is in ms to needed FPS here.
LazyNcoder is offline  
Old 21st January 2020, 15:55   #1324  |  Link
ukmark
Registered User
 
Join Date: Oct 2018
Posts: 33
Quote:
Originally Posted by ukmark View Post
Received the i7 10th gen laptop. Brief testing - very strange results with QSV HW encoding. The graphics driver version really makes a huge difference and has been flaky. The Lenovo driver (non-DCH) and the latest Intel driver (DCH) give different results. It seems that the drivers will have to be improved over the next few months. I had one occasion (doing a test on 2 mins of video) where the output was noticeably different between the first and second test - same input file, same driver, same settings etc - I can't figure that one out, but I swear nothing changed - apart from the output quality getting better. The speed increase is about 70% for HW encoding. The API version used by QSV is now 1.30, on KL it was 1.27 and there are some extra features auto plugged into the encode settings - tskip is new setting, sao is now set to "all" (was "none" on KL), CTU is now 64 (was 32 on KL).

Another observation is that ICQ 22 on Kaby Lake seems the same as ICQ 19 on Ice Lake(!) with the current Ice Lake driver. I used the same input file with the same settings and encoded a QSV HEVC video on Ice Lake with the exact same settings as Kaby Lake (ICQ 22 and all the extra settings identical), and the file was 35% smaller(!). At first I thought maybe quality had somehow been magically maintained with a 35% smaller file - but no, the quality was noticeably inferior on Ice Lake. To get the same video bit rate, I had to lower the ICQ value from 22 to 19 on Ice Lake (flaky drivers again??)

For SW encoding I'm getting about triple the speed (pleased with that) - and the quality is identical to Kaby Lake (which it should be as the graphics driver is not used in SW encoding AFAIK).

I think once the drivers get improved, then QSV HW encoding looks promising. I'm seeing some excellent quality (at times), but as I say, the graphics drivers seem flaky right now.
Longer testing - encoded over 2 hr action movie with my x265 CRF24 settings (720p 10bit HEVC) using SW. Then used latest Lenovo graphics driver (25.x version) and tried with QSV HEVC and set ICQ to 18. This resulted in very similar file size/bitrate between x265 and QSV.

Quality is excellent with QSV (I used the StaxRip "video comparison" tool). I haven't checked every frame , but I did check the usual suspects (dark scenes, fast moving action) and the QSV held up very well. Being very picky, there were some really very minor differences in the dark scenes, but I was impressed with QSV. During normal playback of these scenes I could not see any differences.

Settings for x265 (medium preset, CRF 24, 10bit HEVC, aq mode 3), rest of settings at default.
Settings for QSV (best quality, ICQ 18, 10bit HEVC, bframes 8, ref frames 5, mbbrc on (per macro block rate control), rest of settings at default.

In my previous post, I noted some discrepancies between the ICQ levels for KL and IL, but I was messing around so much with different driver versions and other things that those initial quick tests should probably be ignored. Thus I ended up doing a fresh install of the manufacturer's Win10 image for the above test.

It does appear that QSV compression has improved with IL. On KL, I was getting approx ICQ of 21 or 22 equivalent to x265 CRF24 (bit rate equivalents), whereas now on IL, I get the same bitrate with ICQ 18 or 19. The IL QSV encodes at ICQ 18/19 are better than my KL encodes at ICQ 21/22 yet achieving equivalent bit rates.

I hope to do a few more movies this way (or at least sections of more movies) and will post if I find anything of note.

UPDATE: Checked more dark scenes and QSV still lags a little behind x265 in the really poorly defined dark scenes, where details are not clear even in the original source. The differences on a per frame basis are noticeable but not terrible. Normal playback does not usually expose these differences (to my eyes). I guess for those archive quality videos that you are going to keep forever, then x265 is still the way to go, but IMHO the gap is small between QSV and x265 and seems to be getting smaller with each generation of processor. I'll revisit this if/when new graphics drivers are released (I'm sticking to OEM drivers as trying the latest from Intel made the output worse.)

Last edited by ukmark; 22nd January 2020 at 23:13.
ukmark is offline  
Old 22nd January 2020, 07:49   #1325  |  Link
neo_sapien
Registered User
 
Join Date: Jan 2002
Location: USA
Posts: 249
Quote:
Originally Posted by Patman View Post
Maybe a problem of avs2pipemod... go to x265 options, other and set piping tool to none... test again... really strange
When I set the piping tool to none, I get this error:

Quote:
x265 [error]: yuv: width, height, and FPS must be specified
So in addition to setting the piping tool to none, I also go to Input/Output and set the Frame Rate field to 24000/1001, for 23.976fps. And then I go down to Custom, and input in the Custom field "--input-res 1904x792" which is the resolution of my content after cropping. I then start encoding, and my resulting encode gets up to 2KB and then stops (pretty much immediately).

x264 seems to be working fine without crashing, so I'm doing a workaround - I'm encoding the heavy duty filters that really slow down your encode into x264 qp0 (lossless), then taking that lossless x264 pre-filtered file and running it through x265 using the medium preset and without any additional filters. So instead of spending 25-30 hours running x265, I'm spending 20 hours to build a lossless x264 file, and then spending another 5 hours or so to encode it to x265. This way, if my x265 encode crashes, I only lose up to 5 hours of work, not up to 25-30 hours of work.

Last edited by neo_sapien; 23rd January 2020 at 12:44. Reason: x264 working
neo_sapien is offline  
Old 27th January 2020, 20:21   #1326  |  Link
JKyle
App Digger
 
JKyle's Avatar
 
Join Date: Sep 2018
Posts: 411
@Patman

Hi. In your builds archive, I see a modified version of x264 named x264-159-r2991M-1771b55.
Can you please explain what makes it different from the regular version x264-159-r2991-1771b55?

Thanks
JKyle is offline  
Old 28th January 2020, 07:41   #1327  |  Link
Patman
Registered User
 
Patman's Avatar
 
Join Date: Jan 2015
Posts: 286
Quote:
Originally Posted by JKyle View Post
@Patman

Hi. In your builds archive, I see a modified version of x264 named x264-159-r2991M-1771b55.
Can you please explain what makes it different from the regular version x264-159-r2991-1771b55?

Thanks
Hi JKyle,

Some time ago the code for the progress feedback was changed. Original code is
Code:
    if( last_ts )
        bitrate = (double) i_file * 8 / ( (double) last_ts * 1000 * param->i_timebase_num / param->i_timebase_den );
    else
        bitrate = (double) i_file * 8 / ( (double) 1000 * param->i_fps_den / param->i_fps_num );
    if( i_frame_total )
    {
        int eta = i_elapsed * (i_frame_total - i_frame) / ((int64_t)i_frame * 1000000);
        sprintf( buf, "x264 [%.1f%%] %d/%d frames, %.2f fps, %.2f kb/s, eta %d:%02d:%02d",
                 100. * i_frame / i_frame_total, i_frame, i_frame_total, fps, bitrate,
                 eta/3600, (eta/60)%60, eta%60 );
    }
    else
        sprintf( buf, "x264 %d frames: %.2f fps, %.2f kb/s", i_frame, fps, bitrate );
    fprintf( stderr, "%s  \r", buf+5 );
    x264_cli_set_console_title( buf );
    fflush( stderr ); // needed in windows
    return i_time;
my changes are
Code:
    if( last_ts )
        bitrate = (double) i_file * 8 / ( (double) last_ts * 1000 * param->i_timebase_num / param->i_timebase_den );
    else
        bitrate = (double) i_file * 8 / ( (double) 1000 * param->i_fps_den / param->i_fps_num );
    if( i_frame_total )
    {
    int eta, eta_hh, eta_mm, eta_ss, fps_prec, bitrate_prec, file_prec, estsz_prec;
    double percentage, estsz, file_num, estsz_num;
    char *file_unit, *estsz_unit;
    fps_prec     = fps > 999.5 ? 0 : fps > 99.5 ? 1 : fps > 9.95 ? 2 : 3;
    bitrate_prec = bitrate > 9999.5 ? 0 : bitrate > 999.5 ? 1 : 2;
    file_prec    = i_file < 1048576000 ? 2 : i_file < 10485760000 ? 1 : 0;
    file_num     = i_file < 1048576 ? (double) i_file / 1024. : (double) i_file / 1048576.;
    file_unit    = i_file < 1048576 ? "K":"M";
    if( i_frame_total )
    {
        eta        = i_elapsed * (i_frame_total - i_frame) / ((int64_t)i_frame * 1000000);
        percentage = 100. * i_frame / i_frame_total;
        eta_hh     = eta / 3600;
        eta_mm     = ( eta / 60 ) % 60;
        eta_ss     = eta % 60;
        estsz      = (double) i_file * i_frame_total / (i_frame * 1024.);
        estsz_prec = estsz < 1024000 ? 2 : estsz < 10240000 ? 1 : 0;
        estsz_num  = estsz < 1024 ? estsz : estsz / 1024;
        estsz_unit = estsz < 1024 ? "K" : "M";
        sprintf( buf, "x264 [%.1f%%] %d/%d frames, %.*f fps, %.*f kb/s, %.*f %sB, eta %d:%02d:%02d, est.size %.*f %sB",
                 percentage, i_frame, i_frame_total, fps_prec, fps, bitrate_prec, bitrate,
                 file_prec, file_num, file_unit,
                 eta_hh, eta_mm, eta_ss,
                 estsz_prec, estsz_num, estsz_unit );
    }
    else
        sprintf( buf, "x264 %d frames: %.*f fps, %.*f kb/s, %.*f %sB",
                 i_frame, fps_prec, fps, bitrate_prec, bitrate,
                 file_prec, file_num, file_unit );
estimated size reappears

At the moment, i do some tests in compiling x265 with msvc2019 and a filesize of 7 MB. Stay excited

Short question: Does anyone use the -bStylish option of my x265 build? Or is it enough for you to see the estimated size?
__________________
Tools for StaxRip | x264 - x265

Last edited by Patman; 28th January 2020 at 08:28.
Patman is offline  
Old 28th January 2020, 08:26   #1328  |  Link
JKyle
App Digger
 
JKyle's Avatar
 
Join Date: Sep 2018
Posts: 411
Quote:
Originally Posted by Patman View Post
At the moment i do some tests in compiling x265 with msvc2019 and an filesize of 7 MB. Stay excited
Thanks for the info.

And... Wow! That's good news.
I'm pretty much looking forward to it.
JKyle is offline  
Old 29th January 2020, 00:30   #1329  |  Link
Spunkmeyer
Registered User
 
Join Date: Jan 2020
Posts: 12
Quote:
Originally Posted by Taurus View Post
@all:
If someone is doing a manual update:
NVEnc 4.61 & NVEnc 4.60 are throwing an exception on my machine.
4.59 and former are doing nice.
Just to be warned.....
No time for troubleshooting at the moment.
This is on Win7 64bit (what else?)
Win 10 64 bit and it throws the exception out at the very end of encoding. All is not lost though as I been able to manually mux the video and audio back from the temp files.

Definitely Nvenc is the problem, as I have tried it in other ways as well.
Spunkmeyer is offline  
Old 1st February 2020, 07:36   #1330  |  Link
craigpro
Registered User
 
Join Date: Feb 2003
Posts: 68
Hi all, hoping someone can help me please.

I'm using Staxrip 2.0.6.0 stable and I've noticed that subtitles are not being copied exactly from the source MKV.

eg: looking at the source file

Text #1
ID : 4
Format : ASS
Codec ID : S_TEXT/ASS
Codec ID/Info : Advanced Sub Station Alpha
Duration : 23 min 27 s
Bit rate : 15.2 kb/s
Count of elements : 10888
Compression mode : Lossless
Stream size : 2.55 MiB (0%)
Title : Signs [kdfss / FFF]
Language : English
Default : Yes
Forced : No

Text #2
ID : 5
Format : ASS
Codec ID : S_TEXT/ASS
Codec ID/Info : Advanced Sub Station Alpha
Duration : 23 min 38 s
Bit rate : 15.3 kb/s
Count of elements : 11334
Compression mode : Lossless
Stream size : 2.58 MiB (0%)
Title : Subtitles [kdfss / FFF]
Language : Japanese
Default : No
Forced : No


looking at the output file:

Text #1
ID : 4
Format : ASS
Codec ID : S_TEXT/ASS
Codec ID/Info : Advanced Sub Station Alpha
Duration : 23 min 27 s
Bit rate : 15.2 kb/s
Count of elements : 10888
Compression mode : Lossless
Stream size : 2.55 MiB (1%)
Title : English
Language : English
Default : No
Forced : No

Text #2
ID : 5
Format : ASS
Codec ID : S_TEXT/ASS
Codec ID/Info : Advanced Sub Station Alpha
Duration : 23 min 38 s
Bit rate : 15.3 kb/s
Count of elements : 11334
Compression mode : Lossless
Stream size : 2.58 MiB (1%)
Title : Japanese
Language : Japanese
Default : No
Forced : No


MPC-HT will not display any subtitle text from the output file when subtitles for stream 1 are enabled. If I select stream 2 on the output file then the ENGLISH subtitles are displayed (not the Japanese).

my Subtitle settings in Staxrip are set to:

Preferred Languages: all
Track Name: %language_english%
Demux Subtitles: All
Default Subtitle: None
convert sup - disabled
Add hardcoded - disabled

I believe those settings are default as I haven't messed with them that I can remember.

What should the settings be to directly copy the existing subtitles from the source file be please? Thank you.
craigpro is offline  
Old 1st February 2020, 15:23   #1331  |  Link
Spunkmeyer
Registered User
 
Join Date: Jan 2020
Posts: 12
Quote:
Originally Posted by Taurus View Post
@all:
If someone is doing a manual update:
NVEnc 4.61 & NVEnc 4.60 are throwing an exception on my machine.
4.59 and former are doing nice.
Just to be warned.....
No time for troubleshooting at the moment.
This is on Win7 64bit (what else?)
Seems to be fixed now with Nvenc 4.62.

@craigpro the only thing I can see is that subtitle 1 is set as default in the original file and not so in the output.

Last edited by Spunkmeyer; 1st February 2020 at 15:26.
Spunkmeyer is offline  
Old 1st February 2020, 22:05   #1332  |  Link
ukmark
Registered User
 
Join Date: Oct 2018
Posts: 33
Quote:
Originally Posted by craigpro View Post
Preferred Languages: all
Track Name: %language_english%
Demux Subtitles: All
Default Subtitle: None
convert sup - disabled
Add hardcoded - disabled

I believe those settings are default as I haven't messed with them that I can remember.

What should the settings be to directly copy the existing subtitles from the source file be please? Thank you.
Could you try setting Default Subtitle: None to 'eng'?
ukmark is offline  
Old 4th February 2020, 00:43   #1333  |  Link
JoyBell
Registered User
 
Join Date: Jun 2017
Posts: 18
How to add new AVISynth Filters to StaxRip

How to add new AVISynth Filters to StaxRip?
I have spent several hours reading around and I have to admit that I simply am stumped on how to add a new AVISynth script/filter to Stax. I know how to make new filters in Stax in Profiles I have a few dozen customized filters I use all the time under MyName grouping.

I am trying to add AviSynthShader v1.6.5 into Stax so that I can try using many of the GPU shader real time filters that keep coming out that look interesting.
https://github.com/mysteryx93/AviSynthShader/releases/tag/v1.6.5

I put the .dll and autoload file in the /plugins64+ folder and then I am lost.

I am more lost on how to import the HLSL shaders after.
https://github.com/bloc97/Anime4K

The new Nvidia sharpen I would like to try and is available as well as Anime4K filters, both of which are real time and of great interest.
That being said (as every other ask gets bogged down on what HLSL shaders I use as an example) the main thing I am asking is using AviSynthShader for all those interesting HLSL filters
JoyBell is offline  
Old 8th February 2020, 10:54   #1334  |  Link
Pulp Catalyst
Registered User
 
Join Date: May 2006
Posts: 297
Hello,

anyway of staxrip having the option to automatically remuxing the subtitles files back into the final job, so it doesn't have to be remuxed by manual methods.

what i want is the original subtitles string to just be remuxed in the final MKV, as most if not all of my players/streamers can handle native subtitles files now (dvd and bluray pgs).

thanks,

i'm doing a huge job of many dvd's (episodic tv shows and my wife is italian, so keeping the english subtitle files is a great help for her understanding).
Pulp Catalyst is offline  
Old 9th February 2020, 22:51   #1335  |  Link
ukmark
Registered User
 
Join Date: Oct 2018
Posts: 33
Just an observation and an FYI.................

Just thought I'd pass this along.

During testing of StaxRip in conjunction with Intel QSV 10bit HEVC, I accidentally found a way of drastically reducing CPU usage when encoding the video, and not lose any encoding speed. I had already changed in the encoder settings, the "Decoder" (under the "Basic" section), from "Avisynth/VapourSynth" to "QSVEnc Hardware", and that did reduce the CPU usage a little, but not by much. CPU usage was still around 30-35% with no other activity occurring on the laptop.

However, by chance, I changed a setting "Async Depth" - this is under the "Performance" section. I changed the value from the default 0 (this appears to set the value at runtime to some random number - often 6 - in my case), to a value of 1. Now my CPU usage ranges between 4-7%!!!

Hard to believe the difference. I have done this a few times now just to confirm what I'm seeing. The bottom of the laptop is barely lukewarm. Makes a change from 100% CPU usage with SW encoding.

There's not much info out there I could find about Async Depth, but just thought I'd pass this finding on.

UPDATE: Just had one file that was encoding at half-speed. I increased the async depth to 2 from 1 and it is now encoding at full speed. CPU usage is still between 4-7%.

UPDATE 2: After some quick testing re "async depth", this is what I found on my laptop. It appears that setting this value explicity from 2 to 5 keeps the CPU usage between 4-7%. Also fps encoding speed is identical (so far) regardless of which value you choose between 2 to 5. Setting the value to 1 can cause speed to drop about 50% (but not always).

Setting the "async depth" to 0 (default), caused the CPU usage to increase to approx. 35%. There was no fps encoding speed increase. The "async depth" 'chosen' by QSVEnC (or the Media SDK API??) for this video was 6. All of this stuff is new to me, but it would appear a safe bet to set "async depth" to 3 or 4 and leave it there. That's what I plan on doing and if anything changes I'll let you know.

Good to know that you can get very low CPU usage with QSV and not lose any speed.

I am using ICQ setting (1=best & quality=21), HEVC 10bit, output 720p, and using latest version of QSVEnC (v3.30) and Media SDK API version of 1.30. Laptop is i7-1065g7/12gb ram and 4 cores/8 threads running Win 10 v1909.

UPDATE 3: I find that setting "1=best" is required to keep the CPU usage in that 4-7% range, setting it to "balanced" or "fast" sends it back up to 25% or higher (no idea why that is). That's fine by me as I always use that setting. I'm still getting around 90-100 fps, and as I have alluded to in a previous post, I'm finding video quality to be very good with the settings I am using.

That excellent "video comparison" tool works really well, and I am seeing very little or no difference between the QSV encode and ones done with SW encoding using the "medium" preset with CRF 24, x265 720p 10bit HEVC, (also set aq-mode to 3 and aq-strength between 0.7 and 1). I have compared about five movies so far and every comparison is extremely close between HW and SW encoding. Also, bitrates are reasonably close between HW and SW and I have had 2 movies so far where the HW bitrate was lower than the SW with equivalent (or very similar) quality. These were Captain Marvel and Avengers Infinity War. QSV seems to do a really good job with the modern digital format movies, where the image is very clean and has very little or no grain.

Considering the speed difference between HW and SW (about 3 to 4 times faster with HW), the drastically reduced CPU usage (less than 10% vs 100% with SW encoding) and the very acceptable quality of current HW encodes I am getting, then this will be default from now on. I'll only revert to SW encoding if the quality is poor for a HW encode.

Last edited by ukmark; 14th February 2020 at 15:53.
ukmark is offline  
Old 10th February 2020, 00:49   #1336  |  Link
Th3RadMan
Registered User
 
Join Date: Jan 2020
Posts: 2
I'm brand new to Staxrip, so when I took a look at the options list and got hit with... where do I start

First I'll start with my goal, transcode 4k blu rays to H.265 for space storage with as little video degradation as possible. Of course saying 'as little degradation' will make some people say then just don't transcode, however, other than the loss of HDR, when I was using handbrake, I managed to get extremely good quality with H.265 file sizes. You couldn't tell which video looked worse with my settings. That was comparing using a sdr monitor, so the color looked the same. But now that I have seen the light of a 10 bit pipeline, I want to take advantage.

So my question is this, what needs to be adjusted in the huge list of options to get the best quality to size videos. I'm not afraid of having 2 systems run 24/7 as I'm used to having 24-34hr transcode times per 4k movie on HB.

So far I have changed the preset to slow and mode to quality, with a value of 18. For video I don't know what else needs to be touch. Additionally for audio, I don't want anything to be transcoded, just pass through. Do I just set it to Copy/Mux the audio I want? I swear somewhere I read you want all or non under 'Demux Audio'... but I'm not sure.
Th3RadMan is offline  
Old 14th February 2020, 18:20   #1337  |  Link
NikosD
Registered User
 
Join Date: Aug 2010
Location: Athens, Greece
Posts: 2,901
Major VCEEncC v5.02 update for AMD cards.


-Only Win10 is supported.

-Updated the version of AMF to 1.4.14-> 1.4.16.
Requires AMD Radeon Software Adrenalin Edition 20.2.1 or later.

-Added an option for pre-anaysis added in AMF 1.4.16.
(Valid only in VBR mode, VCEEncC only)

(--pa, --pa-sc, --pa-ss, --pa-activity-type, --pa-caq-strength, --pa-initqpsc,- -pa-fskip-maxqp)

-Added the option to calculate ssim / psnr. (--ssim /-psnr)

...and more.

https://onedrive.live.com/?authkey=%...%21516&o=OneUp
__________________
Win 10 x64 (19042.572) - Core i5-2400 - Radeon RX 470 (20.10.1)
HEVC decoding benchmarks
H.264 DXVA Benchmarks for all
NikosD is offline  
Old 14th February 2020, 20:19   #1338  |  Link
Taurus
Registered User
 
Taurus's Avatar
 
Join Date: Mar 2002
Location: Krautland
Posts: 903
Rigaya released NVEncC 4.64 a few days ago.
And for what it's worth:
QSVEnc 3.30 is available.
Taurus is offline  
Old 15th February 2020, 23:35   #1339  |  Link
ukmark
Registered User
 
Join Date: Oct 2018
Posts: 33
Is this an Intel bug?????

Came across something I've never used before on my new shiny i7-1065g7 laptop - "fixed function encoding" setting in StaxRip.

Thought I'd give it a try and....!!!

Tried h264 encoding with the best (slowest) preset, output to 720p, ICQ method, and used QSVEnc decoding. Encoding speeds reached ..... over 500fps!!

The resulting file is of very good video quality and file sizes are not excessive.

Can't find much on the web regarding what settings are changeable (if any), when using fixed function encoding (I assume the term "fixed function" is self-explanatory). I can see that it ignores the number of b-frames selected and sets this to 0 in the encode log.

Looks like you choose the decoder, preset, profile, level, encoding method and quality setting, and output resolution. With h264 if I change the ICQ setting from say 24 to 26, the resulting file size is reduced as you would expect. However, with h265, changing the ICQ level has no effect on the file size produced.

Does this seem like an Intel bug with h265 ICQ on fixed function encoding, or is it a QSVEnc bug (rigaya) or even something in StaxRip??

BTW, speed on h265 10bit is 350fps using same settings as for h264 and video is very good quality. File sizes are approx 30-50% smaller than h264.

Any feedback welcome - also if anybody can shed more light on fixed function encoding and what settings are available.

I've also asked the question on rigaya's QSVEnc github.

Ta!

UPDATE: If encode method is VBR or QVBR-q, then changing the quality (or bitrate for VBR) does affect output file size for h265 when "fixed-func" is activated. Just does not currently work for ICQ and h265. I did find some notes on rigaya's github that "fixed-func" gave a lower quality output when activated. TBH, on Ice Lake, quality looks very good to me. I know that Intel touted improved QuickSync performance on Ice Lake - maybe this is why the quality has improved for fixed-function encoding? (I don't have anything to compare it with - but to my eyes, quality is very comparable to SW encoding).

I've uploaded 4 images from the "video comparison" tool showing 2 versions of "Star Trek Into Darkness". First version is 10bit HEVC SW encoded with CRF 24, x265, medium preset (encode used 100% CPU and achieved around 25-30fps). Second version is 10bit HEVC HW encoded with ICQ fixed-function (unknown quality setting due to issue above, used 25% CPU @ 350fps!). There is a tiny bit more detail on the face of Spock and Khan. However, watching the footage, you'd be very hard pressed to see any differences. Fixed-function file size is 20% bigger (bit rate ~1200kbps vs ~1000kbps on x265). I also encoded fixed-function to h264 8bit with ICQ setting. The file size is 40% larger than the h265 fixed-function version (bit rate ~1700kbps). Quality on this is also very good and encode achieved over 500fps. I've uploaded the freeze frame of Khan from h264 fixed-function encode also.

It would be difficult to go back to SW encoding. I can accept a very tiny drop in quality (and 20% larger files), in return for vastly improved encoding speed and much reduced CPU usage.

Last edited by ukmark; 16th February 2020 at 18:00.
ukmark is offline  
Old 19th February 2020, 01:39   #1340  |  Link
MrBrownCow
Registered User
 
Join Date: Feb 2019
Posts: 29
I can't figure out why my encodes with AAC audio keep getting an error. AC3 works fine but AAC does not. I think it has something to do with "CoreAudioToolbox 7.10.9.0" but I've followed the instructions for extracting the dll files from the itunes downloader with the make portable script and moved those dll files into the same folder as the qaac64.exe file yet i still get an error.


qaac 2.68, CoreAudioToolbox 7.10.9.0
ERROR: Not available input file format

Anyone having success using qaac and AAC audio?
Attached Files
File Type: txt sourcegoeshere_staxrip.txt (17.8 KB, 25 views)
MrBrownCow is offline  
Closed Thread

Tags
aac, hdr, hevc, nvenc, staxrip, x264, x265

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 23:42.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.