Audio formats – which are best for a school library?

As I was writing the review on Audacity, it occurred to me that it might be helpful to have an entry on open audio formats.

Audio Files for Distribution

The vast majority of people use MP3s (endless players and software), AAC (Apple and Sony products; some phones) and sometimes WMA (Windows).

 

.WMA

 

These are all effectively open-ethos formats although they are not truly open source. On my equipment, they all sound similar. (There are differences which are discussed at length elsewhere.)

In a library or school setting, MP3 files are the safe choice. All your users, no matter what software or players, should be able to play them.

The trick is to encode (aka “rip”) the file properly and, in my experience, you need just two compression levels (aka bitrates): one for spoken audio and one for music.

AudioFormats_encoding screen

For audio books and podcasts, I rip to mono, 64 kbps. This creates a small file and, to my ears, sounds fine.

For music, I rip to stereo, 192 kpbs. This creates a larger but still reasonably sized file and sounds good on my equipment.

I decided on these settings by encoding at different rates and playing it on my different equipment, including my best headphones and studio speakers. I compared the file sizes and made a subjective choice.

You may consider doing the same experiment on your equipment.

Keep in mind, if people are downloading your files, choose the smallest file that sounds OK to you. For me, 48 kbps is barely acceptable and 32 kbps is bad. Audiophiles and computer geeks will drive you crazy with their debates about the best audio format/codec. Bottom line: if it sounds good, it is good.

Unfortunately I can’t recommend the most common truly open-source audio format: OGG. It is a good format but it’s not widely adopted.

Audio Files for Archiving

For archiving, you’ll want a non-lossy, well-supported, format that can easily be converted to smaller, lossy, formats.

The safest choice is WAV, which is proprietary but open-ethos. Virtually every audio program can use a WAV file. Importantly a WAV can be converted to almost any other format and will be supported for many years. But, WAV files are huge.

. AIFFWAV . FLAC

Another good choice, especially for Apple users, is AIFF. Like WAV, AIFF is uncompressed, non-lossy audio. AIFF files are also huge. I suspect that AIFF files won’t be orphaned for a long time, although maybe sooner than WAVs. I am unclear on the license but it is open-ethos.

For years I have been archiving my audio in the open-source FLAC format. This is compressed but non-lossy. It is roughly 40% smaller than a WAV file but sounds the same. There are plenty of open source tools to convert FLACs to MP3s or burn to an audio CD. FLAC seems to be gaining popularity and I expect will be supported for a long time (but maybe not as long as WAV files.) I use the default settings on my encoding or exporting software.

(If you aren’t a purist archiver, a highest-level (320 kpbs) encoded MP3 should also be OK. These files will sound nearly as good as WAV/AIFF/FLAC and should be easily converted to any future file format. It will be much smaller than WAV or even FLAC.)


I tried to keep this simple but if you find this confusing, here’s what I do:

For distribution:

  • spoken word: MP3, mono, 64 kbps.
  • music: MP3, stereo, 192 kbps

For archiving:

  • FLAC, default settings