[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
▼ Settings Mobile Home
/g/ - Technology
Return Catalog Bottom Refresh
Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • You may highlight syntax and preserve whitespace by using [code] tags.
08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]

Janitor applications are now open. Apply here!

[Advertise on 4chan]
[Return] [Catalog] [Bottom]
Anonymous
0.99 SSIM H.264 vs AV1 No.108968571
File: chart.png (20 KB, 1259x500)
20 KB PNG
0.99 SSIM H.264 vs AV1  No.108968571
I took the 73 Mbps 4K NLUUG Tears of Steel encode, scaled it to 720p, and used CRF 10 veryslow x264 to simulate 720p BD.
SSIM cheating was disabled in AV1 with tune 0 and variance boost, 720p was chosen to handicap AV1's 128x128 superblocks.
Even when nearly maxing out x264 compression , preset 4 AV1 = ~30% better compression and encodes ~40% faster as well.
I confidently declare that H.264 is now official boomer trash like flickering CRT displays.

ffmpeg -i tearsofsteel_4k.mov ^
-vf "scale=1280:720:force_original_aspect_ratio=decrease,pad=1280:720:(ow-iw)/2:(oh-ih)/2:black" ^
-c:v libx264 -preset veryslow -crf 10 -pix_fmt yuv420p input.mp4


base_av1=ffmpeg -i input.mp4 -c:v libsvtav1 -crf 28 -preset 4 -pix_fmt yuv420p10le -svtav1-params
psy_av1=keyint=5s:lookahead=120:tune=0:enable-variance-boost=1:variance-boost-strength=3:variance-octile=4

frame=17616 fps= 18 Lsize=  212825KiB
bitrate=2375.4kbits/s speed=0.73x elapsed=0:16:45.14


base_h264=ffmpeg -i input.mp4 -c:v libx264 -crf 18 -preset veryslow -pix_fmt yuv420p

frame=17616 fps= 13 Lsize=  304524KiB
bitrate=3399.0kbits/s speed=0.535x elapsed=0:22:51.14


https://ftp.nluug.nl/pub/graphics/blender/demo/movies/ToS/tearsofsteel_4k.mov
>>
Anonymous
 No.108968584
 No.108968584
>>108968571
AV1
When am I getting unbanned btw ?
>>
Anonymous
 No.108968585
 No.108968585
>blender garbage
May as well have tested against modern garbage anime.
Do it with analog, grainy footage, and observe how av1 is defective by design.
>>
Anonymous
 No.108968601
 No.108968601
>>108968585
>"Noooo, don't use video compression for real world use-cases!"
pathetic boomer cope. Don't you have some CRT to deyaw or some shit?
>>
Anonymous
 No.108968609
 No.108968609
>>108968571
No matter how good a codec is, it will be considered shit here if it's not supported on 4chan
>>
Anonymous
 No.108968644
 No.108968644
>>108968571
Are you going to shill for av1 until av2 becomes usable? Why do you care about it? It’s such a non topic for anybody who isn’t running a streaming company
>>
Anonymous
 No.108968710
 No.108968710
0.99 SSIM is special because it's correlated with a CRF ~20 x264 encode, both objectively and subjectively rated as very high quality.
When AV1 was initially released it was only good for low to medium quality video encodes and not recommended for personal encodes.
To make things worse, AV1 uses tricks to cheat in SSIM but Tune 0 and variance boost unnoficially disable SSIM cheating.
>>
Anonymous
 No.108968717
 No.108968717
>>108968644
Streaming companies do not target 0.99 SSIM thought.
>>
Anonymous
 No.108968950
 No.108968950
>>108968710
>but Tune 0 and variance boost unnoficially disable SSIM cheating
proofs?
>>
Anonymous
 No.108968970
 No.108968970
>>108968950
https://github.com/psy-ex/svt-av1-psy/blob/master/Docs/Appendix-Variance-Boost.md

>The use of subjective mode (--tune=0) often results in an image with greater
sharpness and is intended to produce a result that appears to humans
to be of high quality (as opposed to doing well on basic objective measures, such as
PSNR).

https://gitlab.com/AOMediaCodec/SVT-AV1/-/blob/master/Docs/CommonQuestions.md
>>
Anonymous
 No.108971210
 No.108971210
File: azxW2NN_460swp_proc.png (255 KB, 460x640)
255 KB PNG
>>108968644
>Why do you care about it?
Can you not read, H.264 now:

A) bloats video filesize
B) takes longer to encode than AV1
>>
Anonymous
 No.108971220
 No.108971220
>720p
huh?
>>
Anonymous
 No.108971501
 No.108971501
File: av1-savings-vs-h264.png (177 KB, 1209x997)
177 KB PNG
>>108971220
720p is a stress test for AV1 because there is less opportunity to use large 128x128 superblocks and advanced spatial prediction, whereas 4K gives it more room to leverage those tools. H.264's 16×16 macroblocks remain reasonably efficient at 720p but become somewhat inneficient at 1080p and dogshit inneficient at 4K.

4K = 8MP / 256 = up to 32,400 16x16 H.264 macroblocks
vs
4K = 8MP / 16,384 = up to 506 128x128 AV1 superblocks
>>
Anonymous
 No.108972402
 No.108972402
>>108971243
>>108968571
why does daiz make the same thread multiple times per day, and post anonymously and change trip code periodically?

>>108968644
so far we've confirmed that he:
- thinks jewgle will pay him for shilling their shit (every dav1d, tel avif and webpiss shill post comes from him).
- wants everyone to consume slop because he wants to paywall the original quality.
- testing how easily brainwashed modern retards are (it's been proven that 4cuck will gobble up anything that gets astroturfed).
>>
Anonymous
 No.108972612
 No.108972612
>>108972402
Thanks for the bump?
>>
Anonymous
 No.108972670
 No.108972670
You've never gotten someone banned for admitting breaking 7 and you don't need a reason to get someone banned, so why have you been trying that for so many years, Daiz?
>>
Anonymous
 No.108972742
 No.108972742
>>108972670
thanks for the bump?
>>
Anonymous
 No.108972977
 No.108972977
I find it fascinating how much SVT-AV1 has improved in just 5 years. Picrel is from 2020 and preset 10 was able to compete with placebo x264 and often you'd get better compression with x264 placebo so nobody took AV1 seriously at the time. Now AV1 mogs x264 with the slower preset hard.

https://www.spiedigitallibrary.org/conference-proceedings-of-spie/11510/1151021/The-SVT-AV1-encoder--overview-features-and-speed-quality/10.1117/12.2569270.full?tab=ArticleLink
>>
Anonymous
 No.108974081
 No.108974081
>>108968571
i calculated how much it would cost to reencode several TB of my video and the electricity alone was 70% of the cost of an external 2.5" 4tb drive
idk why anyone would bother unless you have tons of bluray rips and want to convert them into something more reasonable
>>
Anonymous
 No.108974586
 No.108974586
>>108971210
AV1 would obviously be great on 4chan since it would mitigate the anemic file size limit, but people aren't going to stop using h264 anytime soon since it will play on any device made in the last 15 years
>>
Anonymous
 No.108975576
 No.108975576
File: av1_minimum_maybe.png (40 KB, 814x780)
40 KB PNG
>>108974586
>but people aren't going to stop using h264 anytime soon since it will play on any device made in the last 15 years
I've heard that to decode 1080p AV1 well the only rough requirements are that the CPU must contain at least 4 CPU cores and have AVX2.
The i5-4570 was released in Q2 2013 and we're currently in Q2 2026 which would mean that AV1 is backward compatible with CPUs from 13 years ago?
>Q2 2026 refers to the second quarter of the 2026 calendar year, covering the three-month period from April 1 to June 30, 2026

Anyway you got it completely wrong senpai, H.264 is backwards compatible with like 20 year old CPUs, I doubt a core2duo is going to struggle with baseline profile H.264.
Still if AV1 is backwards compatible with CPUs from 13 years ago as of 2026 then that means in 2028 it's going to be backwards compatible with CPUs from 15 years ago.
So a conservative estimate places 2030 as the year when like 95% of all computers on earth will be able to playback 1080p AV1 just fine without hw support.
>>
Anonymous
 No.108975613
 No.108975613
>>108975576
I said any device. You're only considering PCs
Can a 15 year old smart TV play AV1 movies from a USB stick?
>>
Anonymous
 No.108976044
 No.108976044
File: smart_tvs.png (117 KB, 1197x661)
117 KB PNG
>>108975613
Smart TVs make up a tiny fraction of media consoooming centers though. It's primarily phones and like 80% of them are on Android 12 or higher.
Not sure what their software decoding performance is like though I imagine most of them should at the very least be able to decode 720p AV1 just fine.

https://www.appbrain.com/stats/top-android-sdk-versions

>Can a 15 year old smart TV play AV1 movies from a USB stick?
90% won't last 15 years. The idiot who buys a smart TV without an Android TV box/mini PC isn't going to be doing anything else other than use it for streaming slop anyway.

Anyway I don't see the point in delaying yourself AV1 adoption because of normalfaggots. AV1 to anyone who isn't a complete tech illiterate moron is now viable on 13 year old computers.
If you want to be extra conservative wait until 2030 when the last few remaining i5-2500K shitboxes finally croak I guess.
>>
Anonymous
 No.108976069
 No.108976069
>>108968571
>>108971210
okay now measure how many joules (watts is fine) it takes to decode and playback both of them or else you're no better than leddit and HN >>108968585 >>108968609
>>108968601
name one real world use case for encoding speed being a net utility. you won't because actual high-resolution simulation renders are always going to end after hours anyways, and you'll just do nightlies if you actually are doing something useful instead of hobbyist garbage. not a complete waste of time 6/10 for effort
>>
Anonymous
 No.108976216
 No.108976216
File: 1777330741593.webm (3.84 MB, 480x640)
3.84 MB
3.84 MB WEBM
>>108976069
There was a thread about AV1 playback on a phone and it doesn't seem to drain phone battery too much. Not sure about the wattage but it seems pretty low.

https://desuarchive.org/g/thread/108705144
>>
Anonymous
 No.108976320
 No.108976320
>>108975613
You're replying to a bot following the instructions of a Finnish samefagging retard called Daiz who thinks every internet user is on a 30 year old Intel CPU and on the latest phone.
[Return] [Catalog] [Top]
Post a Reply
Return Catalog Top Refresh
[Advertise on 4chan]
Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.