[Ironclad] Dungeon Meshi - 22 [1080p.AV1] (Weekly) :: Nyaa ISS

[Ironclad] Dungeon Meshi - 22 [1080p.AV1] (Weekly)

Category:
Date:
2024-05-30 17:05 UTC
Submitter:
Seeders:
23
Information:
Leechers:
1
File size:
670.6 MiB
Completed:
1943
Info hash:
bd29793ff52104d080b8a1a4869b07205ed2de08
|Dungeon Meshi| | | ------------- |-------------| |Source|[NanDesuKa](https://nyaa.iss.one/view/1826376)| |Audio|[opusenc libopus 1.5.1](https://github.com/Chocobo1/opus-tools_win32-build) "--bitrate 256" [JPN] [ENG] [SPA-LA] [SPA] [ARA] [FRE] [GER] [ITA] [POR-BR] [IND] [THA] [HIN]| |Subtitles|[ENG] [SPA-LA] [SPA] [ARA] [FRE] [GER] [ITA] [JPN] [POR-BR] [POR] [POL] [DUT] [NOB] [FIN] [TUR] [SWE] [GRE] [HEB] [RUM] [IND] [THA] [KOR] [DAN] [CHI] [CHI-TR] [VIE] [UKR] [HUN] [CZE] [HRV] [MAY] [FIL] [HIN]| |Target SSIMU2| Mean: 80| |Encoder|[svt-av1-psy v2.1.0rc2](https://github.com/gianni-rosato/svt-av1-psy) "--crf 25.0-55.0 --preset 4 --tune 3 --sharpness 2 --variance-boost-strength 4 --variance-octile 4 --frame-luma-bias 100 --keyint 0 --enable-dlf 2 --enable-cdef 0 --enable-restoration 0 --enable-tf 0"| |FGS| [grav1synth](https://github.com/rust-av/grav1synth) "diff"| |Filters|[Denoise](https://github.com/AmusementClub/vs-nlm-cuda) "d=2, a=[3,2,2], h=0.3, channels=YUV", [dering](https://github.com/Irrational-Encoding-Wizardry/vs-dehalo) "planes=[0,1,2]", [deband](https://github.com/shssoichiro/soifunc/blob/main/soifunc/deband.py), [dither](https://github.com/Jaded-Encoding-Thaumaturgy/vs-tools/blob/a70d077077ed711ed8e1c18a48f8d8e24486dd11/vstools/functions/utils.py#L241)| [Mediainfo](https://rentry.co/ap5ucu8g) Interested in AV1?: [Discord](https://discord.gg/83dRFDFDp7) Recommended players: [MPV](https://thewiki.moe/tutorials/mpv/), [MPC-HC](https://github.com/clsid2/mpc-hc)

File list

  • [Ironclad] Dungeon Meshi - 22 [1080p.AV1].mkv (670.6 MiB)

nph

User
What's the rationale for 256 opus? I haven't followed trends in the anime encoding space for quite a few months. With the new 1.5 ML stuff improving quality for the same bitrate even further doesn't it make sense to leave it at the usual 128 for TV-tier stereo?
[@nph](https://nyaa.iss.one/view/1826503#com-1) We follow the following equation for Opus bitrates: `128 * ({target_channels}/2.0)^0.75` [(source)](https://wiki.hydrogenaud.io/index.php?title=Bitrate#Equivalent_bitrate_estimates_for_multichannel_audio). With a base bitrate of 128 because that's the threshold for transparency for stereo audio (edgecases exist), that equates to: - 256kbps for 5.1 audio - 320kbps for 7.1 audio. Opus 1.5 didn't introduce improvements in the high fidelity range, it improved stuff for very low bitrates, speech and web delivery, as well as some performance improvements.