Opus player for Android or iOS? Options

[richard123]

Do any Android or iOS music players support Opus?

[2012]

Do any Android or iOS music players support Opus?

Yes.
vlc-android (being a libvlc-based player) supports ".opus" files now. You have to grab APKs from here until they release a new version.

A lot of cycles are used on my relatively low-end ARMv6-based phone (around 22-24% mono, 39% stereo). This could be caused by using the floating-point decoder implementation. Anything ARMv7-based should work better.

This post has been edited by 2012: Sep 13 2012, 14:20

[saratoga]

Do any Android or iOS music players support Opus?

Rockbox should have it fairly soon, and the decoder implementation is integer and reasonably efficient (much optimization is still needed however).

[IgorC]

vlc-android (being a libvlc-based player) supports ".opus" files now. You have to grab APKs from here until they release a new version.

I've installed VLC (android, Cortex-A9 version) to Galaxy II with Android 4.0.3.
Works!!!

[Murphy17]

Hi, today neutron music player get updated via the google store, and it read perfectly opus files encoded with opus tool 0.15 at 128 kbps.

[eahm]

Off topic

Opus streaming radio trial of 7 channels, each at 3 bitrates, that has been running since Sept 2012 has gone unmentioned

Absolute Radio's Listen Labs Opus Streaming Trial

Really http://www.hydrogenaudio.org/forums/index....st&p=808989

Why they hare three streams: http://www.hydrogenaudio.org/forums/index....st&p=809763

This post has been edited by eahm: Nov 9 2012, 08:02

[Oliver]

It should be in the next release of PowerAmp. Now I just need to decide what what bit rate to encode and re-encode all my music.

http://forum.powerampapp.com/index.php?/to...l__opus__st__20

[jkauff]

HD Player for iOS now supports Opus.

[Kohlrabi]

Rockbox should have it fairly soon, and the decoder implementation is integer and reasonably efficient (much optimization is still needed however).

Any info on how well it performs? From what I heard about the Opus decoder for Windows (or generally x86/x64), it is much slower than decoders for other lossy formats. On a portable device battery usage is a really important issue for an audio format.

This post has been edited by Kohlrabi: Dec 3 2012, 17:55

[saratoga]

Rockbox should have it fairly soon, and the decoder implementation is integer and reasonably efficient (much optimization is still needed however).

Any info on how well it performs? From what I heard about the Opus decoder for Windows (or generally x86/x64), it is much slower than decoders for other lossy formats. On a portable device battery usage is a really important issue for an audio format.

Its pretty well optimized now for ARM. The main thing remaining is to go through an write the FFT in assembly. Unfortunately, Opus does not use the same FFT lengths as other lossy formats, so it will need to be done from scratch. Once thats done it should be comparable to any other lossy format.

[DonP]

Rockbox should have it fairly soon, and the decoder implementation is integer and reasonably efficient (much optimization is still needed however).

Any info on how well it performs? From what I heard about the Opus decoder for Windows (or generally x86/x64), it is much slower than decoders for other lossy formats. On a portable device battery usage is a really important issue for an audio format.

This is a bit of a repeat of what I posted in a thread about audiobooks, but more on topic here. At this point (still not considered a stable release) Opus decodes fine on my E200 sansa at 12 kb/s (speech). It's about at the edge but works at 64 kb/s, and hangs up the controls at 128 kb/s. I'm told there is still a lot of improvement possible. At this point I will switch to Opus for speech but for music I'll wait.

Haven't tried it on my android tablet, but the power usage analysis shows the vast majority of power going to the display pretty much whatever I run even with wifi and GPS running, and the display goes to sleep after a few minutes if I'm just playing audio. On that device I've been sticking with Google music as I can have everything on the cloud and specify albums to be cached locally (for when I'm out of wifi range) and all works well... except that program is basically centered around MP3. I may get another app that handles Opus just to try it, but I'm limited to 16 GB total local storage so I really like the cloud aspect of google music.

[skamp]

I've extrapolated battery life on my iPod Classic with Opus @ 128 kbps, at about 30 hours, vs. 32 with lossyFLAC.

[Oliver]

Hi, today neutron music player get updated via the google store, and it read perfectly opus files encoded with opus tool 0.15 at 128 kbps.

That is great news. Do you know if it is fully supported? It is not listed among the supported codecs.

[Speckmade]

It is not listed among the supported codecs.

The Release Notes for v.1.63 list it as new feature:
http://neutronmp.com/forum/viewtopic.php?t=99

[Kohlrabi]

From what I heard about the Opus decoder for Windows (or generally x86/x64), it is much slower than decoders for other lossy formats.

I ran some tests with latest foobar2000 1.2 beta 1, which now uses ffmpeg decoders for some popular lossy formats. I used presets for both lossy and lossless formats which yielded comparable bitrates (~62kbps/~700kbps). LAME didn't quite reach the desired bitrate even with V9. Here's the result:

CODE

System:
CPU: Intel® Core™ i5-3570K CPU @ 3.40GHz, features: MMX SSE SSE2 SSE3 SSE4.1 SSE4.2
App: foobar2000 v1.2 beta 1
Settings:
High priority: no
Buffer entire file into memory: no
Passes: 10
Threads: 2
Postprocessing: none
Stats by codec:
AAC: 3960.371x realtime
ALAC: 673.897x realtime
FLAC: 1190.631x realtime
MP3: 3468.749x realtime
Musepack: 1712.903x realtime
Opus: 409.543x realtime
Speex: 513.438x realtime
TAK: 875.835x realtime
Vorbis: 2788.389x realtime
WavPack: 537.514x realtime
Total:
Decoded length: 1d 1:09:52.000
Opening time: 0:00.079
Decoding time: 3:22.789
Speed (x realtime): 893.462

So, Opus is not even faster than Speex, with which it also competes, and at least on x86 it is considerably slower than other common lossy formats. You can also see that the new ffmpeg decoders are really blazing fast, Musepack is now third to last in lossy codecs (if you include Speex), too. Let's hope Opus decoders will see some development in the future, but seeing how Xiph steadily improved all their other formats (Vorbis,Theora), I am quite optimistic.

This post has been edited by Kohlrabi: Dec 18 2012, 23:45

[saratoga]

You're basically just looking at how much x86 asm each codec has in ffmepg.

[IgorC]

Kohlrabi,

The math doesn't work by itself. You need to analyze a whole system.
The math can't explain this!

For a modern mobile devices (smartphones etc.) there won't be a gain in battery life in practice if You choose codec A that requires 6-7 MHz for realtime playback comparing to codec B (60-70 MHz) because the lowest frequency of today smartphone's proccessors is 200 MHz (Cortex A series). Something that your math can't see.
On top of that, CPU consumes only a fraction of power because other parts like display consume comparable amount (if not more).

This post has been edited by IgorC: Dec 19 2012, 17:07

[vinnie97]

The excellent BSPlayer now supports Opus.

It's a good thing, too, because it seems like Rockbox development has stalled.

This post has been edited by vinnie97: Mar 1 2013, 20:28