On Your Mark 1080p DTS AAC - AV1 VIDEOTEST

Category:
Date:
2022-11-07 20:21 UTC
Submitter:
Seeders:
0
Information:
No information.
Leechers:
0
File size:
381.3 MiB
Completed:
60
Info hash:
b708ba573ea1a534c203e2f9926a7402f1cbf1f9
Its nothing new. First and foremost, it is an AV1 10bit Videocodec Test. I wanted to give everybody the chance to judge for themself if this new codec lives up to the promis of super high quality with very low space use. MEDIA INFO: Format : AV1 Format/Info : AOMedia Video 1 Format-Profil : Main@L4.0 Codec-ID : V_AV1 Dauer : 6 min 46s Bitrate : 2 750 kb/s (in reality it was encoded in 2048kbps) Breite : 1 920 Pixel Höhe : 1 080 Pixel Bildseitenverhältnis : 16:9 Bildwiederholungsrate : 23,976 (24000/1001) FPS Color space : YUV Chroma subsampling : 4:2:0 Bit depth : 10 bits Color range : Limited Color primaries : BT.709 Transfer characteristics : BT.709 Matrix coefficients : BT.709 Audio #1 Commercial name : DTS-HD Master Audio Codec-ID : A_DTS Bitrate : 4 719 kb/s / 1 509 kb/s Kanäle : 6 Kanäle Bit depth : 24 bits Audio #2 Format/Info : Advanced Audio Codec Low Complexity Codec-ID : A_AAC-2 Dauer : 6 min 46s Bitrate : 311 kb/s Kanäle : 2 Kanäle

File list

  • On.Your.Mark.1995.Japanese.1080p.Bluray.AV1-10bit.DTS.AAC.mkv (381.3 MiB)
If you're gonna dare promising super high quality, we're gonna ask for encode settings sir :) Pretty funny you posted this BTW since I have an AV1 encode of it already for personal use.
> AV1 > Bloats the audio Worst streetshitting I've seen in awhile

eikira (uploader)

User
@NekoTrix, well i let ffmpeg do what it thought to be right, i only specified pretty much the preset... -c:v libaom-av1 -preset slower -vb 2048k onyourmark.mp4 and not i am promising super high quality, the codec creators did, the ao alliance did. @Mizz141 well yes, its very out of ratio for video and audio size. should i have used only aac 5.1 384kbps for it?
I'm not sure where you even saw that claim, AV1 was made with the goal to achieve better appeal at lower bitrate. VOD services are always seeking to reduce costs. Thinking ffmpeg can do *what it think is right* is pretty delusional. Un-tuned aom is simply bad. And target bitrate mode is shit, especially used this way. Also willing to use AAC when you have the superior Opus available while already going the trouble of using AV1 is also pretty bold. If you wanna improve substencially your AV1 encoding knowledge, I advise you to join the [AV1 discord server](https://discord.gg/YEsZsgckMr). If you ask nicely, many kind people will help you out figure things out.
2 eikira - ANY visual information have kinda 10 times more rare info in spectrum than audio info. So ANY talking about miracle visual codecs - is a just bunch of lies with laboratory synthetics of some extremelly simple cartoon frame with overcompression artifacts. Also - there is a problem how much of bitrate is for each video pixel - or will be bitrate starvation with kinda uselessness of large amounts of pixels - as info will be divided to support just large amount of them instead usefullness. Also - there is some theoretical lab compression limit - if I not confusing - it is about 50% without useful info loss and it is pure unreachable theory. In dry result - I've seen some old video that looked better trouth VHS tuners and encoded in Xvid than x264. And same sometimes with x264 better than x265 - especially at low resolutions - as x265 love to spam pixels and soap them for economy - at least I have such feelings at stop-frames with magnitude. AV1 can probably use some visual trick of human perception to trim "unneaded info" from video - like when older codecs like to soap fast-moving frames - still my personal opinion that 1080p lower than some limit of bitrate will be in overcompressed artifacts. At least if it was real 1080p where each pixel used, as framerate also not kinda slideshow.
2 eikira Which lead Us all to fact - that most of modern animations kinda drawn on PC and realisticly their level of detalization kinda at level of 720p probably in 19/20 cases. When it was hand-drawn - problem was in technical capture and limit of TV/DVD of 480p or 576p. In other words - before quality of anime was better but impossible to digitalize correcly. Now it maked quite laidback in CG 3d redactors and "flatted" most oftenly. So we have another problem - quality of average anime become kinda lower than technical possibilities to save it digitally in 1080p let say. And nature itself of CG-based graphics is mathematical formulas - that make them even seen more compressible in modern codecs. Altrouth they never been in real 1080p and same colours not have any analog texturing or hand style of artist - they just feel the same and sterile. What I mean by sterile - they have very simplistic visual info - or probably DETAILS of animation - which may and often do easier to compress them from their soft-bloated size closer to their real sizes that IMO around 600p and most of their pixels is just waste of space & bitrate.
2 eikira Sorry for multipost in comments. What I mean to say - if You try to compress very old animation, especially with unstable floating frames (before wide computer usage from around middle of 1990's) - it OR will bloat the videosize by pure mathematic calculations OR will kill DETAILS of old animation (and replace it by friendly-compressive soap - that make seem video way-way worse than it is, as avalanche of artifacts and cycle of visual errors will be clearly worse than overcompressing digital-derived animations). 1995y - even with good source - in 1080p/23.9fps on 2048 bitrate - is a screaming disaster on level of Youtube fake1080p "Quackity" sorry "Quality" ))) overcompression & throw out some parts of visual info is unavoidable - too much pixels, too much frames...
This post reminded me to watch the official VCD. Back when the song was actively released, the promotional Studio Ghibli music video was also released on VCD. I was lucky enough to find one being sold in a video store. They had so many other videos that this VCD basically managed to hide and avoid being bought by someone else.