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 > New and alternative video codecs

Reply
 
Thread Tools Search this Thread Display Modes
Old 12th June 2024, 22:33   #1001  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,829
Quote:
Originally Posted by kurkosdr View Post
You mean like how Hollywood studios picked the winner of the HD-DVD vs Blu-Ray format war by deciding one day to only throw content at one format (Blu-Ray) while letting the other rot (HD-DVD)? This happened shortly after BD+ gave studios the hope of unbreakable DRM. Whatever studio support HD-DVD had dried up right there and then.
It was more that Sony, who had thought they bet Blu-ray on the success of the PS3, realized they'd also bet the PS3 on the success of Blu-ray, and spent about $1B paying studios to adopt Blu-ray and drop HD-DVD.

Quote:
Or how Hollywood studios cut out MacOS X from Blu-Ray and HD-DVD support because MacOS X didn't provide the DRM infrastructure Hollywood studios demanded for Blu-Ray and HD-DVD?
Given that better content security than DVD was a core motivation for new disc formats, it was utterly expected that they weren't going to support a "just buy a Mac!" sized hole for pirates to drive through.

Expected by Apple as well, who knew what they needed to do to support HW encryption, and got it as part of their Intel transition.

Quote:
It doesn't matter what happened in the distant past, today pre-recorded content is king and DRM is legally empowered by the DMCA. Hollywood studios can cut out from their encrypted content whatever OS and hardware combination doesn't comply with their arbitrary demands.
Studios could only invalidate DRM keys for content on their own wholly-owned streaming services, and I am not aware of them ever doing so. Content distributed by other streaming services control their own DRM keys.

And their demands for secure video paths and HW license stores are hardly "arbitrary" - they have been well documented and understood for years.

Quote:
The weird thing in this case is that HDMI 2.1 is required only if you want 4K@120fps, and Hollywood doesn't serve any 4K@120fps content, but they can still mandate all kinds of DRM requirements by acting as gatekeepers to the HDMI spec. DisplayPort is technically a thing but isn't a thing on TVs (due to lack of eARC support), so you want at least one HDMI 2.1 port if you want to output 4K@120fps to a TV (which means GPU vendors can't boycott HDMI 2.1). And then there is the open question of whether encrypted 8K content will be allowed on DisplayPort or be HDMI 2.1-only.
I'm missing your point here. Pretty much the only things that can play out HDMI 2.1 so far are recent GPUs and game consoles. Most streaming media players and all disc players are still on HDMI 2.0, and some other than that.

Also, you can do DisplayPort to HDMI 2.1 with an adaptor. Nvidia's professional A-series GPUs have 4 DP and no HDMI ports, so I've used an external adaptor to get 120 fps, which has worked well.

The lack of eARC isn't why TVs don't have DisplayPort input. It's more than TV SoCs don't have support for it, so it would be quite expensive to add it.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 14th June 2024, 13:39   #1002  |  Link
kurkosdr
Registered User
 
Join Date: Aug 2009
Posts: 356
Quote:
Originally Posted by benwaggoner View Post
It was more that Sony, who had thought they bet Blu-ray on the success of the PS3, realized they'd also bet the PS3 on the success of Blu-ray, and spent about $1B paying studios to adopt Blu-ray and drop HD-DVD.
HD-DVD was also paying studios, but AACS had already been cracked by then, so the very moment BD+ arrived of Blu-Ray (giving studios the hope of uncrackable DRM), all studios dropped HD-DVD immediately.

Quote:
Originally Posted by benwaggoner View Post
Given that better content security than DVD was a core motivation for new disc formats, it was utterly expected that they weren't going to support a "just buy a Mac!" sized hole for pirates to drive through.
My point is, studios will cut out more popular platforms than Desktop Linux if they don't get the DRM infrastructure they want, so the "boycott" suggested by another person won't work.

Quote:
Originally Posted by benwaggoner View Post
Studios could only invalidate DRM keys for content on their own wholly-owned streaming services, and I am not aware of them ever doing so. Content distributed by other streaming services control their own DRM keys.

And their demands for secure video paths and HW license stores are hardly "arbitrary" - they have been well documented and understood for years.
My point is, the DMCA's doesn't define what copyright holders can demand as part of the DRM infrastructure. If studios decide one day to require a DNA sample, all hardware vendors out there will have to incorporate it into their hardware to gain access to Hollywood content. I am absurd on purpose, but it shows the absurdity of the DMCA's anticircumvention provisions.

And yes, their demands are arbitrary. Just look how Microsoft jumped through every DRM hoop there is out there only for Disney to decree that Disney+ will only be available in 720p on Windows (on Edge and on the Microsoft Store app). Funny thing is, pirates are extracting CDMs from tablets and Smart TVs so the interwebs is still full of Disney WEB-DL torrents (in 4K HDR). But if you are a customer and have a Windows-centric UHD setup, you are screwed even if you've ensured your setup is compliant with the highest tier Widevines and PlayReadys. Did I say arbitrary? I did.

Last edited by kurkosdr; 14th June 2024 at 13:51.
kurkosdr is offline   Reply With Quote
Old 14th June 2024, 17:32   #1003  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,829
Quote:
Originally Posted by kurkosdr View Post
HD-DVD was also paying studios, but AACS had already been cracked by then, so the very moment BD+ arrived of Blu-Ray (giving studios the hope of uncrackable DRM), all studios dropped HD-DVD immediately.
HD-DVD and Blu-ray coexisted on the market for over a year, from October 2006 to January 2008.


Wikipedia has a decent overview: https://en.wikipedia.org/wiki/High-d...isc_format_war

I was working on Microsoft's HD-DVD throughout that period, and paid a lot of attention to the blow-by-blows of what was happening inside the industry as well as for the consumers.

Quote:
My point is, studios will cut out more popular platforms than Desktop Linux if they don't get the DRM infrastructure they want, so the "boycott" suggested by another person won't work.
I'm not sure what you think rights holders should do instead? Leave gaping security holes just so people on low market share operating systems with older hardware can play optical disks instead of just buying a player?

Quote:
And yes, their demands are arbitrary. Just look how Microsoft jumped through every DRM hoop there is out there only for Disney to decree that Disney+ will only be available in 720p on Windows (on Edge and on the Microsoft Store app). Funny thing is, pirates are extracting CDMs from tablets and Smart TVs so the interwebs is still full of Disney WEB-DL torrents (in 4K HDR). But if you are a customer and have a Windows-centric UHD setup, you are screwed even if you've ensured your setup is compliant with the highest tier Widevines and PlayReadys. Did I say arbitrary? I did.
Studio DRM rules are complex, but a lot hinges on software versus hardware DRM, secure license stores, and that sort of thing. The lower the security, the lower quality of content can be licensed. Pretty much anything can do 480p, most 720p, the large majority 1080p, with UHD and HDR the most limited, and almost never in web browsers to date (more feasible in desktop apps). While the issues are certainly complex, often confusing, and commonly have compromises to balance security and market size, they are absolutely not arbitrary.

IIRC, getting full SL3000 level security on Windows requires Windows 11, a GPU from the last few years, and TPM 2.0. Last I looked maybe 18 months ago, <10% of Windows PCs met those requirements (it's certainly higher now but still well less than 50%).

No one believes that any specific technical means would cause piracy to vanish! The goal is to make piracy more difficult to do and make it easier to potentially identify who pirated the video. That discourages piracy-by-default by a casual mass of customers. Piracy being a hassle means more people decide it's just a better use of their time and money to get the content legitimately. And we've seen that the combination of having piracy of the best quality version more challenging while having affordable and easy to use official ways to access the content is highly successful.

But piracy can't be stopped outright, and everyone involved in content protection knows it. The "photon exploit" of putting a really good camera on a tripod aimed at a really good TV in a dark room isn't closable. But that exploit is an expensive hassle, and runs the risk of forensic watermarking identifying who pirated the content.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 15th June 2024, 02:07   #1004  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,991
I couldn't have said it better myself

It's easy to judge the rules that Hollywood studios impose on the streaming of their content from the outside without context, but it all makes sense with some experience.
__________________
These are all my personal statements, not those of my employer :)
Blue_MiSfit is offline   Reply With Quote
Old 15th June 2024, 16:55   #1005  |  Link
ShortKatz
Registered User
 
Join Date: Aug 2018
Location: Germany
Posts: 126
FFmpeg has today added the possibility to add VVC encoding support using
Code:
--enable-libvvenc
via Fraunhofer VVenC.

Edit: It works, but is very slow, about speed=0.012x for 4k on my MacBook Pro.

Last edited by ShortKatz; 15th June 2024 at 17:09.
ShortKatz is offline   Reply With Quote
Old 15th June 2024, 19:44   #1006  |  Link
birdie
Artem S. Tashkinov
 
birdie's Avatar
 
Join Date: Dec 2006
Posts: 363
Indeed.

Quote:
Add external encoder VVenC for H266/VVC encoding.
Register new encoder libvvenc.
Add libvvenc to wrap the vvenc interface.
libvvenc implements encoder option: preset,qp,qpa,period,passlogfile,stats,vvenc-params,level,tier.
Enable encoder by adding --enable-libvvenc in configure step.
birdie is offline   Reply With Quote
Old 15th June 2024, 22:46   #1007  |  Link
ShortKatz
Registered User
 
Join Date: Aug 2018
Location: Germany
Posts: 126
Are there VVenC options to speed encoding up a bit? I did my first test with
Code:
ffmpeg -i input.mp4 -acodec copy -vcodec libvvenc -preset medium -q 30 output.mp4
Which is really very slow.
ShortKatz is offline   Reply With Quote
Old 16th June 2024, 00:21   #1008  |  Link
benwaggoner
Moderator
 
Join Date: Jan 2006
Location: Portland, OR
Posts: 4,829
Quote:
Originally Posted by ShortKatz View Post
Are there VVenC options to speed encoding up a bit? I did my first test with
Code:
ffmpeg -i input.mp4 -acodec copy -vcodec libvvenc -preset medium -q 30 output.mp4
Which is really very slow.
It's still a pretty early encoder, and doesn't seem to actively being developed to be a practical high-volume production encoder.

A lot of speed optimization is both needed and very achievable.
__________________
Ben Waggoner
Principal Video Specialist, Amazon Prime Video

My Compression Book
benwaggoner is offline   Reply With Quote
Old 16th June 2024, 11:56   #1009  |  Link
birdie
Artem S. Tashkinov
 
birdie's Avatar
 
Join Date: Dec 2006
Posts: 363
Quote:
Originally Posted by ShortKatz View Post
Are there VVenC options to speed encoding up a bit? I did my first test with
Code:
ffmpeg -i input.mp4 -acodec copy -vcodec libvvenc -preset medium -q 30 output.mp4
Which is really very slow.
vvenc can actually use all the cores unlike e.g. libaom which is ... single-threaded for all I know.

You really need a device with as many cores as possible to encode into VVC. The standard itself is at the very least 10 times more computationally expensive than HEVC to encode.

And unfortunately in my tests x265 is better than vvenc at lower resolutions (1080p and below) and decent bitrates. Don't ask me why or how. The last time I compared vvenc-1.11.1 (preset=slower) and x265 3.5 (preset=veryslow) and I was unpleasantly surprised.

Looks like new codecs (AV1, VVC) are heavily optimized for 1440p and higher resolutions and fancy features like HDR, 10/12 bit encoding and VR and H.264/x264 and H.265/x265 still beat them at 1080p and below for visually lossless encoding.
birdie is offline   Reply With Quote
Old 16th June 2024, 14:58   #1010  |  Link
Jamaika
Registered User
 
Join Date: Jul 2015
Posts: 737
Quote:
Originally Posted by ShortKatz View Post
Are there VVenC options to speed encoding up a bit? I did my first test with
Code:
ffmpeg -i input.mp4 -acodec copy -vcodec libvvenc -preset medium -q 30 output.mp4
Which is really very slow.
What's the problem? Let's see
Code:
static const AVOption options[] = {
    { "preset",       "set encoding preset", OFFSET(preset), AV_OPT_TYPE_INT, {.i64 = 2}, 0, 4, VE, "preset"},
    { "faster",       "0", 0, AV_OPT_TYPE_CONST, {.i64 = VVENC_FASTER}, INT_MIN, INT_MAX, VE, "preset" },
    { "fast",         "1", 0, AV_OPT_TYPE_CONST, {.i64 = VVENC_FAST},   INT_MIN, INT_MAX, VE, "preset" },
    { "medium",       "2", 0, AV_OPT_TYPE_CONST, {.i64 = VVENC_MEDIUM}, INT_MIN, INT_MAX, VE, "preset" },
    { "slow",         "3", 0, AV_OPT_TYPE_CONST, {.i64 = VVENC_SLOW},   INT_MIN, INT_MAX, VE, "preset" },
    { "slower",       "4", 0, AV_OPT_TYPE_CONST, {.i64 = VVENC_SLOWER}, INT_MIN, INT_MAX, VE, "preset" },
    { "qp",           "set quantization",          OFFSET(qp), AV_OPT_TYPE_INT,  {.i64 = 32}, -1, 63, VE },
    { "qpa",          "set subjective (perceptually motivated) optimization", OFFSET(qpa), AV_OPT_TYPE_BOOL, {.i64 = 1},  0, 1, VE},
    { "passlogfile",  "Filename for 2 pass stats", OFFSET(stats), AV_OPT_TYPE_STRING, {.str = NULL}, 0, 0, VE},
    { "stats",        "Filename for 2 pass stats", OFFSET(stats), AV_OPT_TYPE_STRING, {.str = NULL}, 0, 0, VE},
    { "period",       "set (intra) refresh period in seconds", OFFSET(intra_refresh_sec), AV_OPT_TYPE_INT,  {.i64 = 1},  1, INT_MAX, VE },
    { "vvenc-params", "set the vvenc configuration using a :-separated list of key=value parameters", OFFSET(vvenc_opts), AV_OPT_TYPE_DICT, { 0 }, 0, 0, VE },
    { "level",        "Specify level (as defined by Annex A)", OFFSET(level), AV_OPT_TYPE_STRING, {.str = NULL}, 0, 0, VE},
    { "tier",         "set vvc tier", OFFSET(tier), AV_OPT_TYPE_INT, {.i64 = 0},  0, 1, VE, "tier"},
    { "main",         "main", 0, AV_OPT_TYPE_CONST, {.i64 = 0}, INT_MIN, INT_MAX, VE, "tier"},
    { "high",         "high", 0, AV_OPT_TYPE_CONST, {.i64 = 1}, INT_MIN, INT_MAX, VE, "tier"},
    {NULL}
};
Code:
    if (avctx->thread_count > 0)
        params.m_numThreads = avctx->thread_count;
So it should be
ffmpeg -i input.mp4 -acodec copy -vcodec libvvenc -vvenc-params preset=4 -threads 4 -q 30 output.mp4

Last edited by Jamaika; 16th June 2024 at 15:00.
Jamaika is offline   Reply With Quote
Old 16th June 2024, 15:23   #1011  |  Link
birdie
Artem S. Tashkinov
 
birdie's Avatar
 
Join Date: Dec 2006
Posts: 363
"VVC is dead", VideoLAN president:

https://www.youtube.com/watch?v=6xUhpZXPbBM
birdie is offline   Reply With Quote
Old 16th June 2024, 15:28   #1012  |  Link
kurkosdr
Registered User
 
Join Date: Aug 2009
Posts: 356
Quote:
Originally Posted by benwaggoner View Post
HD-DVD and Blu-ray coexisted on the market for over a year, from October 2006 to January 2008.


Wikipedia has a decent overview: https://en.wikipedia.org/wiki/High-d...isc_format_war
If you look at the dates closely, any studios that used HD-DVD (even exclusively) dropped it the moment BD+ became a thing. Some of them admit it openly and some not, but it played a huge role. More information: https://web.archive.org/web/20080301...rypto-won.html

Quote:
Originally Posted by benwaggoner View Post
I'm not sure what you think rights holders should do instead? Leave gaping security holes just so people on low market share operating systems with older hardware can play optical disks instead of just buying a player?
Studios? Nothing, they are the perpetrators here. The only solution is to repeal DMCA's anti-circumvention provisions or replace them with something that protects the users' rights to fair use, format-shifting, and content ownership. And yes, I consider DRM an attack on fair use, format-shifting, and content ownership. Just look at the recent example of Warner Bros deleting content that users had purchased. Or not being able to rip a DVD or Blu-Ray you've bought and having to pay again to stream the same movie on your tablet.

Also keep in mind that Audio CDs don't have DRM (any attempt to add it like XCP and CDS ended in failure and dropped), Amazon's MP3 store has no DRM, and iTunes dropped its DRM, and yet record labels are making tons of money. Not as much money as if they had successfully coerced you into re-buying your Audio CDs as DRM-ed iTunes songs, but they are still making tons of money.

Anyway, you are on a website that has a "Decrypting" sub-forum and has published multiple articles criticizing DRM and the DMCA's anti-circumvention provisions (inside and outside the forums), so best stop it here before we fill 10 pages discussing this, see those other resources instead.

The gist of the story is that, with the DMCA's anti-circumvention provisions in place, "boycotts" won't work, the studios will simply cut out any platform that doesn't comply with their arbitrary demands.

Quote:
Originally Posted by benwaggoner View Post
IIRC, getting full SL3000 level security on Windows requires Windows 11, a GPU from the last few years, and TPM 2.0. Last I looked maybe 18 months ago, <10% of Windows PCs met those requirements (it's certainly higher now but still well less than 50%).
If I build a PC that complies with SL3000, will Disney+ work on 1080p and UHD (because that's what SL3000 was supposed to be for)? Nope. Nobody has managed to get it working, regardless of configuration. See why I say "arbitrary demands"?

Quote:
Originally Posted by benwaggoner View Post
The "photon exploit" of putting a really good camera on a tripod aimed at a really good TV in a dark room isn't closable.
Recording movies playing on a flat-screen TV is what the MPAA proposed in the past as a way for teachers to exercise their fair-use rights:
https://arstechnica.com/tech-policy/...-not-rip-dvds/
but now it's apparently a "piracy exploit" (and is been targeted by schemes like Cinavia as a "loophole" to be closed). See how DRM in its current form slowly redefines fair-use rights as "piracy" (if not de-jure, certainly de-facto)?

Last edited by kurkosdr; 16th June 2024 at 15:59.
kurkosdr is offline   Reply With Quote
Old 16th June 2024, 18:29   #1013  |  Link
rwill
Registered User
 
Join Date: Dec 2013
Location: Berlin, Germany
Posts: 379
Quote:
Originally Posted by benwaggoner View Post
I'm not sure what you think rights holders should do instead? Leave gaping security holes just so people on low market share operating systems with older hardware can play optical disks instead of just buying a player?
Ah, but what would these people do with a player. Most of these people won't buy discs even if they have a player. A player is useless without discs.
__________________
My github...
rwill is offline   Reply With Quote
Old 16th June 2024, 18:36   #1014  |  Link
rwill
Registered User
 
Join Date: Dec 2013
Location: Berlin, Germany
Posts: 379
Quote:
Originally Posted by kurkosdr View Post
Recording movies playing on a flat-screen TV is what the MPAA proposed in the past as a way for teachers to exercise their fair-use rights:
https://arstechnica.com/tech-policy/...-not-rip-dvds/
but now it's apparently a "piracy exploit" (and is been targeted by schemes like Cinavia as a "loophole" to be closed). See how DRM in its current form slowly redefines fair-use rights as "piracy" (if not de-jure, certainly de-facto)?
I do not think the problem is with a teacher showing his pupils some snippets of a movie in a classroom.

I think the problem is with certain people making whole movies available as CamRips available on the internet for download.
__________________
My github...
rwill is offline   Reply With Quote
Old 16th June 2024, 20:10   #1015  |  Link
kurkosdr
Registered User
 
Join Date: Aug 2009
Posts: 356
Quote:
Originally Posted by rwill View Post
I do not think the problem is with a teacher showing his pupils some snippets of a movie in a classroom.
And yet, the teacher's fair use rights are affected. Not only the teacher has to film the video from a flat-screen TV in real-time (as per MPAA instructions), but if the resulting DVD with the snippets gets played on a standalone Blu-ray player, it can trigger a Cinavia warning (and mute playback).

Instead, people who know some things about DRM know not to use Blu-ray players for copied content that may contain Cinavia (aka content copied from DVDs or Blu-rays) and aren't affected.

Quote:
Originally Posted by rwill View Post
I think the problem is with certain people making whole movies available as CamRips available on the internet for download.
And yet, not a single DRM system has managed to stop those people. Also, ever notice how DRM systems for movies tend to persist even if the DRM has been cracked to bits? Unlike the videogame industry where Denuvo is usually removed after cracked for a given title? It's as if fair-use, format-shifting, and content ownership are the real targets here. Not that compromising users' fair-use, format-shifting, and content ownership is acceptable even if some magic DRM that stops "pirates" existed, but it shows the true priorities of DRM: You can rip an audio CD to your tablet using Windows Media Player or iTunes, but not a DVD because DVD contains DRM (and the anti-circumvention provisions of the DMCA come into play). Also, MicroSD cards and MemoryStick cards support DRMed content but no studio has provided official support for ripping DVDs to MicroSD or MemoryStick using those DRM schemes. I wonder why... maybe selling the same content over and over has something to do with it.

Anyway, this topic has been well-covered in doom9 (both forum.doom9.org and doom9.org). But I can't let it slide when people try to bring up the "content security" excuse.

Last edited by kurkosdr; 16th June 2024 at 20:17.
kurkosdr is offline   Reply With Quote
Old 16th June 2024, 21:42   #1016  |  Link
modus-ms325c
Registered User
 
Join Date: Dec 2023
Posts: 13
Quote:
Originally Posted by kurkosdr View Post
Unlike the videogame industry where Denuvo is usually removed after cracked for a given title?
except not really?
FFXV WINDOWS EDITION has a Steam ver with denuvo *still* built-in, same with MGSV. both hadn't had a new crack in *years*.
FIFA16, JD2017, Handball17, M&MH7 and even EA SWBF1 got their actual real crack literally YEARS after they were released and they (NOT the cracks themselves!) still have denuvo.
modus-ms325c is offline   Reply With Quote
Old 17th June 2024, 00:56   #1017  |  Link
kurkosdr
Registered User
 
Join Date: Aug 2009
Posts: 356
Quote:
Originally Posted by modus-ms325c View Post
except not really?
FFXV WINDOWS EDITION has a Steam ver with denuvo *still* built-in, same with MGSV. both hadn't had a new crack in *years*.
FIFA16, JD2017, Handball17, M&MH7 and even EA SWBF1 got their actual real crack literally YEARS after they were released and they (NOT the cracks themselves!) still have denuvo.
I said usually, not always. Meanwhile studios still release all the DVDs with CSS (despite CSS being cracked to bits by this point) just to prevent format-shifting (aka ripping) using common tools. Gotta sell the same content multiple times somehow.
kurkosdr is offline   Reply With Quote
Old 17th June 2024, 05:44   #1018  |  Link
rwill
Registered User
 
Join Date: Dec 2013
Location: Berlin, Germany
Posts: 379
@ShortKatz: Is this with an Intel or ARM MacBook Pro ?

@birdie: Regarding x265 being better than vvenc at low resolutions ->

TL/DR:

Code:
       vvenc: 40.1422 YUV PSNR
        x265: 38.3294 YUV PSNR
Mystery HEVC: 38.7632 YUV PSNR
Long Version:

I don't know... does not look like x265/HEVC being better than vvenc at low resolutions.

Encoding good old 352x288 foreman_cif.yuv at 200kbit/sec yields:

vvenc:
Code:
./vvencapp -i foreman_cif.yuv -s 352x288 -c yuv420 --framerate 25 --framescale 1 --preset slower -b 200k --pass 1 --rcstatsfile stats.stats --sdr sdr_709 -o foreman_cif.266 -ip 300 --qpa 0
./vvencapp -i foreman_cif.yuv -s 352x288 -c yuv420 --framerate 25 --framescale 1 --preset slower -b 200k --pass 2 --rcstatsfile stats.stats --sdr sdr_709 -o foreman_cif.266 -ip 300 --qpa 0
Code:
vvencapp [info]: Total Time: 1139.730 sec. Fps(avg): 0.263 encoded Frames 300

298009 Jun 17 06:38 foreman_cif.266
-> 40.1422 YUV PSNR


x265:
Code:
./3.6+1-dd1ef69b2/Win64/x265 --input foreman_cif.yuv --fps 25 --input-res 352x288 --input-depth 8 --profile main10 --tune psnr --bitrate 200 --pass 1 -o trash_x265.265 --preset placebo --psnr -I 300
./3.6+1-dd1ef69b2/Win64/x265 --input foreman_cif.yuv --fps 25 --input-res 352x288 --input-depth 8 --profile main10 --tune psnr --bitrate 200 --pass 2 -o trash_x265.265 --preset placebo --psnr -I 300
Code:
encoded 300 frames in 105.40s (2.85 fps), 196.70 kb/s, Avg QP:32.86, Global PSNR: 38.391
encoded 300 frames in 118.37s (2.53 fps), 200.31 kb/s, Avg QP:32.50, Global PSNR: 38.639
(2 passes, and x265 seems to calculate Global PSNR 'differently')

304084 Jun 17 06:38 trash_x265.265
-> 38.3294 YUV PSNR


Mystery HEVC Encoder:
Code:
??????????????
Code:
Encoding time: 113.30 sec
Encoding time: 120.61 sec

300111 Jun 17 06:38 trash_mystery.265
-> 38.7632 YUV PSNR
__________________
My github...
rwill is offline   Reply With Quote
Old 17th June 2024, 06:56   #1019  |  Link
birdie
Artem S. Tashkinov
 
birdie's Avatar
 
Join Date: Dec 2006
Posts: 363
I'm not a big believer in PSNR and I generally prefer to inspect the results using my old analogue eyes.

Also, I mentioned visually lossless encoding and PSNR at 40 doesn't look like it at the slightest. Shouldn't it be much much higher?

You're not testing how well these encoders behave at visually lossless encoding, you've tested how much they maimed the source video when being severely bitrate constrained. I couldn't care less about this usage scenario. In fact I don't understand who and why would ever need it.
birdie is offline   Reply With Quote
Old 17th June 2024, 10:36   #1020  |  Link
oibaf
Registered User
 
Join Date: Nov 2008
Posts: 71
What is the "Mystery HEVC Encoder"?
oibaf is offline   Reply With Quote
Reply

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 10:48.


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