So…yeah. Looking at file size, it clearly beats older 264 or even 265. I don’t mind if my server is going to have to transcode for most clients, I think the size difference in size might be worth it. But not sure which groups I could focus to look for these AV1 releases, seem they’re quite scarce still?
I’m pretty active in the av1 community. Most of us who use av1 encode our own from the raw blurays or high quality remuxes. Besides the av1 content on public trackers, I think I saw a group called onlyfaffs and another one called WhiskyJack who were both putting out some av1 content, but imo, their filesizes are too large, so I avoid them if possible.
The other thing to worry about is that most people who use av1 also convert the audio to opus. Fully opensource codecs and all that. The issue is you don’t know what bitrate they are using for their conversion. So audio is a concern too when downloading av1 content.
Okay so I just went a private general tracker and looked up an av1 movie. It is a 2014 feature film encoded by the WhiskyJack group (the better trackers don’t allow AV1 content yet). For this film, the audio codec is Opus. I looked in the nfo and for the audio file, it says that they are using Opus 5.1 with a 32-bit bitdepth. That’s not ideal. For 6-channel audio, we recommend 256 kb/s. 192 is acceptable, but it’s going to be another 20 megabytes to bump it up to 256, so why not do it?
Also, it doesn’t tell you what the source is. So if the original audio was ac3 or e-ac3, it is not going to sound great.
Bit depth and bit rate are not the same thing.
Yikes. Thanks for catching that. I mistyped.
What’s wrong with FLAC?
Nothing. It’s just that you don’t see people use flac very often as an audio codec for movies or TV series. At least I haven’t seen it very often.
The only issue would be if you were trying to transcode Dolby Digital to flac. That is not a transcode you want to do.
Why is that?
Dolby Digital (e-ac & e-ac3) are lossy codecs. So transcoding a lossy codec to a lossless codec, is not a good idea.
You can read more about it here: https://interviewfor.red/en/transcodes.html
Lossy to lossless is fine it’s just a waste of space.
It’s extra wasted space because compression artifacts are hard to compress lossless. It’s a shit sandwich, so people advise against it.
I didnt realise* that they were lossy. Makes sense now.
A properly muxed mkv will display the resulting audio bitrate. And if you use opusenc, it will embed the encoder settings in the track.
I think variable bitrate is preferable. With a variable bitrate you don’t have a single, specific, telling bitrate show up. In the end you depend on the encoder doing decent work. Which group names can be useful for, to identify and revisit good ones.