Opus seems to be the best lossy codec around, so it would be nice to have support for it.
2 Likes
+1 from me.
Yes, please. This would be most welcome
Hi. This post seems quite old, so I’m bringing up the original question.
Opus is a much better lossy audio format than MP3, which is already considered a rather outdated codec.
Would Renoise be compatible with the Opus format? Could we bring this?
I’m already converting all my portable files to Opus and no longer use MP3. I also want to use this format for uploading content to social media and the web, as it has become a standard.
I use this to convert:
Can we discuss this? What are your thoughts?
At the very least, I’d expect to be able to drag and drop Opus files into the sampler.
There is currently a lot of Opus content on the internet.
when we talk about support do we mean “drag .opus file to renoise, loads as waveform” “export as .opus” - or do we mean, “.xrns format with .opus files instead of wav or flac”?
Any situation that involves using Opus files to load into a project.
For saving to a project, lossless formats (WAV/FLAC) are still suitable.
It should allow the selection of multiple Opus files and their loading into the sampler, either through the native browser or from an operating system folder.
So you can do the same as with mp3 files.
As technology stands, people should stop using mp3 and use opus, until a better format comes out. What cannot happen is that, because of an old, overly standardized format, it prevents other, better formats from being compatible.
Many platforms have already made the switch years ago (abandoning mp3 and using opus).
The idea is: let’s continue using mp3, but let’s also allow the use of opus.
I assume Renoise decodes the MP3 and converts it to PCM (uncompressed WAV) at the project’s sample rate, probably using Lame or derives it to the operating system.
Well, it should do the same with the Opus file.
For this to happen, it might be necessary to integrate a decoder (libopus + libopusfile), or use a backend like ffmeg or libav (libavcodec + libavformat). This would provide greater compatibility with other formats, including AAC, Vorbis, etc. However, the focus is primarily on making Opus compatible.
With libopus + libopusfile a BSD-like license is required, free for commercial use. Libopus uses a permissive BSD license, so it can be used in proprietary software. It would be a good multi-platform solution…
1 Like
Links, please?
Which ones?
I’ve actually never even heard of Opus before yours and @fladd’s post, so I’m genuinely curious…
Why is Opus better? What measurable evidence can you provide that clearly shows we’ve all been doing it wrong? (I do genuinely want to learn!)
Are we talking orders of magnitude here… or just slight minor improvements to lossy encoding?
I suppose the fact that @fladd’s original thread is from 2017 and basically got zero attention doesn’t really bode well for the adoption of this interesting (?) new (ish?) audio format…
I mostly just stick to .FLAC / .WAV myself, because disk space is cheap and internet is fast, but hey… it’s nice to learn about different things!
Considering the Amiga era, the sound quality in Opus is more than sufficient.
The beauty of tracker data structures lies in their high efficiency—enabling broader expression with minimal data.
If Renoise Song/Instrument audio files could optionally support Opus, I feel this file format could be considered an ideal form from the perspective of the tracker’s tradition of high data efficiency.
It’s practically the de facto standard for streaming services.
It also offers high information efficiency in files. I apologize if I’m mistaken, but its low-latency characteristics might be particularly well-suited for tracker software.
1 Like
But what does streaming or low-latency have anything to do with producing music using Renoise at home or in the studio?
Are we talking day-to-day usage here, i.e. making music at home or in the studio, or are we talking about the final delivery of a song to the wider world?
When pushing music out into the world, then yes… lossy compression formats make some sense… MP3, OGG, whatever… but FLAC will always be the better choice here if you actually care about how your sounds will sound to the end listener 
(And FLAC cuts the filesize to ~50% which is plenty!)
Funny that you mention this.
I’ll be 47 next April.
I’ve been using trackers exclusively since I was about 8 years old.. almost 40 years I’ve been fucking with the hex! 
I understand the pain of 4 channel .MODs in Protracker, Soundtracker, Noisetrekker, etc.
I’m very familiar with the Amiga and Atari ST days, using floppy disks with around 800KB max storage, 11kHz 8-bit samples, etc.
I understand “minimal data”, haha. 
But I also understand the need to keep integrity… pure data… raw data… uncompressed unfucked data… bit-for-bit perfection.
MP3, OGG, ETC compressed mangled samples ain’t the way to go… imho!
Cheers,
K.
Yeah, that was a concern. Using compressed files might not be ideal for tracker software. But wouldn’t the same apply to FLAC?
I don’t know how it’s processed internally, so sorry if I’m wrong.
EDIT:Come to think of it, it’s definitely not bit-perfect.
FLAC is essentially just a ZIP archive for WAV files.
It’s 100% accurate to the original sound.
Renoise also supports GZIP for samples in your song/instrument, although the compression ratio is slightly less efficient compared to FLAC.
Either way, they both offer perfect 1:1 untouched audio.
You can take a .FLAC vs the original .WAV and perform a null test, and it will be identical.
Yeah, you’re right. I found the idea of data-efficient project files appealing, but considering the potential for various unexpected issues, it might not be such a good idea.
Processing related to samples, such as -Sxx and slicing, may start to malfunction.
It’s great if you only intend to share the workings of a project to someone for learning purposes…
Check out the included demo song “@Hunz - Soon Soon” for a superb example of this… where all of his lovely lyrics would occupy hundreds of megabytes of storage, if not compressed to .OGG format
But for working on my own music, saving, reloading, tweaking, editing, saving, reloading, editing, etc… I would never use lossy formats for this… The sounds would just turn to mud after several recompressed generations.
It’s all about what your intension is.
wow, with all due respect, but how can you work in the music software domain and not know Opus? That is…wild! 
Opus is the successor to OGG Vorbis (https://wiki.xiph.org/OpusFAQ#Does_Opus_make_all_those_other_lossy_codecs_obsolete?) and Renoise supports this, too, right?
Indeed. See als above.
Indeed, it seems that Ogg is being used in that project.
$ pwd
/tmp/DemoSong - Hunz - Soon Soon
$ find -type f
./SampleData/Instrument17 (Vox Bridge 02)/Sample00 (Recorded Sample 25).ogg
./SampleData/Instrument16 (Vox Chorus 03)/Sample00 (Recorded Sample 24).ogg
./SampleData/Instrument15 (Vox Chorus 02)/Sample00 (Recorded Sample 22).ogg
./SampleData/Instrument14 (Vox Verse 03)/Sample00 (Recorded Sample 20).ogg
./SampleData/Instrument13 (Vox Verse 02)/Sample00 (Recorded Sample 19).ogg
./SampleData/Instrument12 (Vox Chorus 01)/Sample00 (Recorded Sample 18).ogg
./SampleData/Instrument11 (Vox Bridge 01)/Sample00 (Recorded Sample 17).ogg
./SampleData/Instrument10 (Vox Bg 04)/Sample00 (Recorded Sample 13).ogg
./SampleData/Instrument09 (Vox Bg 03)/Sample00 (Recorded Sample 12).ogg
./SampleData/Instrument08 (Vox Verse 01)/Sample00 (Recorded Sample 11).ogg
./SampleData/Instrument07 (Vox Chorus 00)/Sample00 (Recorded Sample 08).ogg
./SampleData/Instrument06 (Vox Bg 01)/Sample00 (Recorded Sample 06).ogg
./SampleData/Instrument05 (Synth 02)/Sample00 (New Sample 05).flac
./SampleData/Instrument04 (Synth 01)/Sample00 (New Sample 05).flac
./SampleData/Instrument03 (Bass 01)/Sample00 (New Sample 04).flac
./SampleData/Instrument02 (Snare 01)/Sample00 (New Sample 03).flac
./SampleData/Instrument01 (Kick 02)/Sample00 (New Sample 02).flac
./SampleData/Instrument00 (Kick 01)/Sample00 (New Sample 01).flac
./Icon.png
./Song.xml
$
If that’s the case, wouldn’t adding support for Opus alongside Ogg be a worthwhile upgrade?
However, simply importing Ogg files and saving the song file doesn’t seem to store the samples as Ogg. Are they being manually replaced?
If that’s fundamentally possible, it should also be possible to create lossy compressed versions of song files using external tools.
If we enable importing Opus as described below, would it then be possible in principle to store Opus within the song file?
Of course, I see no issue with you having a policy of prioritizing quality in your project.
I pretty much take a back seat these days.
I don’t pay much attention to the “trends” in the wider world
So Opus was a new one to me, even though it’s actually an old one… I simply had no need to even know that it even existed.
Please don’t take any of my comments as an “official” Renoise team response.
I realise that I’m an admin here, but I’m also just (mostly) a stupid dumbass user, like everyone else haha.
So please take anything I say with a healthy grain of salt!
Indeed. When preparing the demo songs for final release, we internally swapped large .WAV samples for .OGG versions in the .XRNS (aka .ZIP file) and Renoise is then just smart enough to pick up the correct sample.
Renoise doesn’t save .OGG natively, because like I said before that would be a lossy/destructive process, and you risk turning all your sounds into muddy crap very quickly.
But you CAN manually open the .XRNS in something like 7-zip, WinZip, WinRar, etc., and then replace the embedded .WAV files with an .OGG version for final release.
I’d love to be able to import opus, although usually the opus-stuff I get are samples someone sends me with whatsapp. I save the data and convert it to wav with ffmpeg.
But concerning opus in xrns-files? I don’t see the advantage of it. Probably in Demos if you do not want to exceed the filesize (64kb demos) but afaik, there are no demos with a renoise-player out there.
Okay, there are two issues here:
- Being able to use Opus samples directly to load them into the sampler. If this existed, Renoise would do the same as with MP3 files: convert them to PCM (lossless digital audio). This is the point I was referring to.
- Being able to save an xrnx song project whose container saves in Opus format instead of WAV or FLAC. I hadn’t considered this, but it might be great to be able to do so in certain uses.
Regarding the second point, I see a clear advantage in being able to share xrnx songs within the community. For years, we’ve lost a forum where we all shared xrnx files of our productions. Server capacity is also limited. Imagine a large number of xrnx song files that people simply want to share, the advantage of which is that the file takes up very little space compared to FLAC/WAV. It’s a way to optimize resources. For example, I have a piano sample with 5 layers based on velocity. How can I share xrnx songs if the file is too large? This would be an extreme case. But the idea of ​​high-quality lossy audio combined with low file sizes is something Renoise could leverage.
That said, obviously, when working on a project, lossless audio data should be used, as Renoise currently does, using WAV or FLAC, and this is necessary to preserve the project’s quality.
For example, for competitions where people can share more complex songs with numerous samples, Opus could be advantageous, both for loading samples and using those sounds that are only available in Opus, and for saving a project with Opus (lossy audio) simply for sharing, not for further developing that song project.
Why don’t we have a dedicated forum for sharing xrnx anymore?
Is it because many composers use instrument/effect plugins, or is there more to it?
As I usually do orchestral stuff, I HAVE to use (commercial) plugins like musio, kontakt or sine player which many people do not own. Besides that I tend to use plugins like Waves Mastering Suite or Ozone.
Of course I could use stuff like vital or helm which are free and in that case I could upload xrns-files, too. (in fact, I did → Made something again. MP3+XRNS-File for learning purposes )
But I deleted the xrns afterwards.
The other big issue with uploading xrns might be that people use commercial samples. As far as I know it’s not a big deal releasing music with commercial samples but the companies do not want you to spread those samples. (Which is what you do in xrns-files)
And as renoise is located in germany, the “Urheberrecht” is nothing you want to fÜck with.