Assembling remote ensemble mp3 files

Our small wind ensemble is in lock down - I am trying to continue by using Audacity to combine 5 individual mp3 recordings.
These recordings have slightly different lengths and amplitudes - in the extreme they may even have different tempi.
As a returning Audacity user I’m a tad lacking in experise. I would welcome any general pointers in solving these issues.

Everybody should be using the same [u]click track[/u] or the same starting/backing track.

You can still get some timing drift and/or pitch errors when mixing & matching “cheap” soundcards, but hopefully that won’t happen.


individual mp3 recordings

Ideally, you should avoid lossy compression because when you open the file in Audacity (or any normal audio editor) it gets decompressed. Then if you re-export as MP3 (or other lossy format) you are going-through another generation of lossy compression, and the “damage” does accumulate. (AAC is much more immune to accumulated damage so that may be a better option if you need the smaller-compressed files.)

If you are not familiar with mixing, mixing is done by summation so you can easily get clipping (distortion). You can lower the levels before mixing/exporting and/or you can export to floating-point WAV (which can go over 0dB without clipping), then re-import and Normalize or Amplify to bring the levels down.

I agree with Doug, so these comments are in addition to what he wrote:

A “click track” does not need to literally be “clicks”, though clicks are good as a count in (bands often start with the drummer clicking his sticks together while verbally counting a couple of bars (“1, 2, 3, 4, 1, 2, 3, 4” → start playing)
If the music has a rhythm part (such as a strummed guitar) or an accompaniment part (perhaps a piano part), then that can work well in place of a click track.

Having a clear count-in is extremely helpful.


“Ideally” yes, though uncompressed files (such as WAV) can be too big for some email services. If you absolutely must use a compressed format, use a high “kbps” setting so as to minimize sound quality degrading.

Consider using “cloud storage” such as dropbox, Google drive or One drive. These will usually allow large files to be shared, so sending and receiving WAV files is not a problem (assuming you all have broadband).

If any of your group are recording on their phones, ask them to set the quality setting to high / highest quality, and send you the raw, unedited files.


Using a click track / guide track should prevent that from occurring.

Use the “Time Shift Tool” to drag tracks left / right so that they line up (Time Shift Tool - Audacity Manual)

in the extreme they may even have different tempi.

How did you arrange sync? You can certainly set up for Audacity overdubbing where you play to a backing, sync or rhythm track, but short of that (which can be a little magic), you can send a sound file around that they play on their phone into headphones while Audacity records the actual performance. The computer only has to do one single thing really well rather than try to record and play, perfectly, at the same time.

There’s tricks for beginning sync. You send a sync start sequence at the beginning of the master track. They hold their headphones up to their microphone—just for that one little piece.

What you hear when you get their file is your own voice saying, "One And Two And [loud clap], russle russle and then they start the performance based on your music leadin.

Are you trying to do this live on the internet? That can be fraught with problems. My home internet watches movies, OK, but my uplink speed is terrible and doesn’t support any real-time work. There is a local live roundtable split-screen from Los Angeles news people and one of them has my problem. Five of them have “normal” bubbling here and there where his sounds like three and a half miles of bad road with stuttering, halting video.

He has my uplink.

Koz

Thankyou for your replies.
I have made some progress using a backing mp3 tracks produced from a Lilypond Midi output. One for each player with his part missing.
So far all I have to do is
1 to clip the start a bit and change tempo to align the length of the individual tracks
2 to adjust the gains.

I will take on board your suggestions and maybe branch out into wav files.

Thanks again.