Support for Opus codec

Any plans? My .opus files are not recognized when scanning.

[mherger]

I'm sorry, no plans. Maybe you could ask Andy, author of the Audio::Scan module we're using:

https://github.com/andygrundman/Audio-Scan

[andygrundman]

I'll comment here to keep the discussion in one place. So, the good news is that adding Opus support to Audio::Scan would not be much work, probably about on the level of the work that was done for DSD support recently. There would be 4 bits of work needed in Audio::Scan:

1. Parse the info header. https://wiki.xiph.org/OggOpus#Packet_Organization
2. Parse the comments (tags). No extra work as the comments are Vorbis comments.
3. Seek offset calculation, which may or may not work just like in Ogg files.
4. Comprehensive test suite. Probably start with https://people.xiph.org/~greg/opus_testvectors/

After all this stuff, it's relatively easy to add to LMS as a new transcoded file format.

All that being said, can you explain your use case a bit? Are these your own encoded Opus files or are they from somewhere else? If they are your own, why do you prefer Opus over FLAC? I think Opus is great for things like low-latency voice chat, but I don't see a reason to use it for music. I'm not trying to discourage you, just pointing out that the chances of this getting worked on are pretty slim (no pun intended).

Thanks Andy, my case is pretty simple:
I encoded the files myself (on Windows) using the downloads from xiph.org.
I'm simply trying to get the best possible sound quality for a given bitrate (=Opus codec?), because I have a couple of (Rockbox-based) players with limits on the SDHC-card size - and they play Opus just fine.
I just bought a couple of cheap Raspberry Pi's where I use PiCorePlayer - I really like the combination, so
why not use the Opus files on the pi's, instead of having all my music duplicated using another codec?
And last, but not least: I'm 65, so lossless encoding is only used for archiving purposes :-)

[mherger]

FWIW: somebody seems to have patched Audio::Scan to support Opus: https://github.com/jrmuizel/AudioScan-opus/.

Somebody else further developed LMS support: http://forums.slimdevices.com/showthread.php?t=107011

Thanks a lot, I'll follow the discussion(s)

[tschlotfeldt]

Hi, just for the question "Why to support Opus?" Because it is the standard of the future. Opus will also be the standard of the broadcasting industry as all new hardware in this business will support Opus.

See this posting on the mozilla blog: https://hacks.mozilla.org/2012/09/its-opus-it-rocks-and-now-its-an-audio-codec-standard/

[andygrundman]

Reposting from the Audio-Scan issue, for the benefit of those watching this one:

After reviewing the existing Opus patch, I can't accept it. The samplerate is hardcoded to 48000 when the format definitely supports more than one samplerate. If anyone would like to continue work on Opus support, please fork this repo and do so via a pull request. I also can't stress enough that a proper set of tests is essential, with test files that cover the range of possible Opus encoding options.

my_hv_store( info, "samplerate", newSViv(48000) );
samplerate = 48000;

[andygrundman]

Hi guys, I changed my mind about Opus and merged Jeff's great patch into Audio-Scan 1.01. Hope this helps!

https://github.com/andygrundman/Audio-Scan
https://metacpan.org/release/Audio-Scan

I'm reopening this, now that Andy has been kind enough to take the initial, important step.

Any change of plans to support opus in LMS?

[mherger]

You should be able to build this for your own use pretty easily. Did you try to do so? Head over to Andy‘s repository to get the latest and greatest.

Thanks Michael, but I'm not into development - I'm a github user because I want to be able to report issues and make feature requests, I only use Linux on my tiny PCPs (normal desktop is Windows).

I realize that you do this in your spare time, that there are tons of real problems for you to deal with, and that implementing just another lossy codec doesn't exactly get top priority, so please close the issue with whatever code you find suitable.