Compressing voice recordings

freakazoid · **Joined:** Wed Jul 18, 2007 5:45 pm **Posts:** 598

lautrepay wrote:

I suppose that other players based on the BASS audio library (XMPlay, Trout, Nemp, MusicBee) will start to offer support for this new codec soon.

XMPlay already has an OPUS plugin in their forum:
http://www.un4seen.com/forum/?topic=138 ... 9#msg96799

m^(2) wrote:

FLAC is rubbish. Unmaintained for years, weak and slow, except for decompression.
It gained some popularity because of missteps of competition, but I can't understand why is it growing.

I never liked FLAC either, but people started using that more. I prefer WavePack, but that never really caught on.

Midas · **Posted:** Thu Sep 13, 2012 2:19 am

m^(2) wrote:

FLAC is rubbish. Unmaintained for years, weak and slow, except for decompression.
It gained some popularity because of missteps of competition, but I can't understand why is it growing.

Non-lossy OSS? Compress once, use forever...

http://wiki.hydrogenaudio.org/index.php?title=Free_Lossless_Audio_Codec wrote:

Pros:

Portable to many systems
Source open and freely licenced
Hardware support (PhatBox, Kenwood MusicKeg, Rio Karma, etc. See below)
Streaming support
Extremely fast decoding
Supports multichannel and high resolution streams
Supports ReplayGain
Supports cue-sheet (with some limitations)
Gaining wide use as successor to Shorten

Cons:

Compresses less efficiently than other popular modern compressors (Monkey's Audio, OptimFROG)
Higher compression modes slow, for little gain over the default setting.

m^(2) · **Posted:** Thu Sep 13, 2012 8:37 am

Midas wrote:

m^(2) wrote:

FLAC is rubbish. Unmaintained for years, weak and slow, except for decompression.
It gained some popularity because of missteps of competition, but I can't understand why is it growing.

Non-lossy OSS? Compress once, use forever...

Just like WavPack, ALAC, TTA, Shorten, ALS, SLS and arguably MAC.

Midas wrote:

m^(2) wrote:

http://wiki.hydrogenaudio.org/index.php?title=Free_Lossless_Audio_Codec wrote:

Pros:

Portable to many systems
Source open and freely licenced
Hardware support (PhatBox, Kenwood MusicKeg, Rio Karma, etc. See below)
Streaming support
Extremely fast decoding
Supports multichannel and high resolution streams
Supports ReplayGain
Supports cue-sheet (with some limitations)
Gaining wide use as successor to Shorten

Cons:

Compresses less efficiently than other popular modern compressors (Monkey's Audio, OptimFROG)
Higher compression modes slow, for little gain over the default setting.

WavPack beats it in all regards except for power usage and hardware support in portable use. Which neither is good for anyway.
And unlike FLAC its developer supports it, so when something breaks (I've had data loss caused by a bug in FLAC, mangled order of channels), you have a chance to get it back.

Midas · **Posted:** Fri Sep 14, 2012 8:44 am

I'm no audio expert, but my portable players (H3x0 & Fuze) are compatible with FLAC -- as well as with some others you mention -- and it was what was available and prescribed for many years for long-term audio preservation... BTW, it being OSS, one can hardly consider the dev under obligation to support it. But I'll be looking into Wavpack after your input, that's for sure...

http://www.rockbox.org/wiki/SoundCodecs

http://www.rockbox.org/wiki/CodecPerformanceComparison

http://wiki.hydrogenaudio.org/index.php?title=WavPack

m^(2) · **Posted:** Fri Sep 14, 2012 9:00 am

Obviously, nobody is obliged to support it, but lack of a person willing to do it is a potential problem when you don't need it and can be a show stopper when you do.

And sorry, I'm being somewhat aggressive on this topic, before discovering other codecs I was a FLAC fan too, but now it really hurts me to see how much more worthwhile projects seem to stall while FLAC is growing.
Frankly, I'm not a fan of WavPack because it's too weak for my liking, but among options that do what I want them too (cross-platform portable with good software support, open source, replay gain, extended tags, cue sheets, multichannel, high resolution), this is the the strongest one (and the only that has always worked for me).

TenaciousD · **Joined:** Thu Nov 15, 2012 11:38 pm **Posts:** 31

For my music I just reduce the bitrate to around 64 or even 32 Kb/s and the sound quality is not that bad at all.

freakazoid · **Joined:** Wed Jul 18, 2007 5:45 pm **Posts:** 598

TenaciousD wrote:

For my music I just reduce the bitrate to around 64 or even 32 Kb/s and the sound quality is not that bad at all.

I used to do this (convert to 32kbps) when I had very little HDD space 15 years ago!

If I listen to some of these files again, they sound like crap.

Midas · **Posted:** Mon Nov 19, 2012 2:26 am

freakazoid wrote:

TenaciousD wrote:

For my music I just reduce the bitrate to around 64 or even 32 Kb/s and the sound quality is not that bad at all.

I used to do this (convert to 32kbps) when I had very little HDD space 15 years ago!

If I listen to some of these files again, they sound like crap.

For MP3s, the established knowledge supports that the minimum acceptable bitrate should be 128kbps -- so called 'radio quality' level; but to each, his own... :mrgreen:

webfork · **Posted:** Sat Apr 20, 2013 12:08 pm

Update: my current process

Tested out LameXP's Opus compression and so far very impressed. Ran a basic comparison between the Ogg voice-specific algorythm (SPX) and Opus at roughly the same bitrate and the Opus compression was much clearer.

Part of this may come from the fact that the SPX algorthym is a few years old now and my encoder (speexdrop) hasn't been updated in a long time. Whatever the case, the toolset for Opus and the support for the codec via Mozilla makes playing the resulting files wonderfully simple.

The general conclusion is that I'll be adding another program to my process so:

V2V - encoding voice memos to mono WAV format (also does normalization)
Wavosaur - delete any silent areas and do some more fine-grained edition/normalization (if necessary)
LameXP - encoding to Opus on 16 kbps VBR (optimize for speech in the advanced options). The lowest setting (8 kbps) wasn't bad either. None of the tin-cup or raspy-sounding audio artifacts I'm used to with a low-quality encoding.

Anyway, thanks to m^(2) for the Opus suggestion and LameXP's dev for making conversion extremely smooth.

Compressing voice recordings

Who is online