I want to know how to amplify all tracks but have the analysis for the amplification amount to be based on each track individually.
For example…
(1) If I simply choose Amplify, without selecting an individual track, then Audacity will amplify all tracks based on the the track with the highest peaks. So suppose I have one track that is kind of high (snare drum) but the vocals are low. Then Amplify will say “amplify up 10” and bases that recommendation off the high track, but that is not enough for the vocal track. Etc.
(2). If I select an individual track, just one track, and choose Amplify, then it will amplify based on just that track. If I do this in turn with each track, then I get what I want. High tracks go up just a little. Low tracks go up a lot. Etc. But that is a manual process.
Can I get Audacity to do (2) without having to manually select each track?
Yes you can
That’s exactly what the “Normalize” effect does. Rather than amplifying by a specified amount, it amplifies ("normalizes) to a specified level. Audacity Manual
Where did you read that?
It depends on what you are doing.
If you want to amplify multiple tracks by the same amount, then use “Amplify”.
If you want to amplify multiple tracks to the same level, then use “Normalize”.
Some other audio programs (incorrectly) apply dynamic range compression to the audio when you “normalize”. The Normalize effect in Audacity does what “normalize” is supposed to do and simply “amplifies” to the specified level. Please see the manual for full details of the Normalize effect: Audacity Manual
They’re different tools. Everybody looks at the word “Normalize” and thinks it’s going to make their show “Normal.” It’s just a variation on Amplify with a funny name.
Normalize has the ability to change the volume of, for example, Left without affecting Right. Or affecting them differently which was your request. If your show is intentionally out of balance for valid theatrical reasons, Normalize will destroy the show.
Neither of these two tools is a gift from the angels. They work on peak sounds. If the loudest thing in your show is a dog barking or a gunshot, that will take over the show and the guy quietly reading Shakespeare is irrelevant.
Neither tool affects the show note by note. You need the compressors to get that. They have the effect of twisting a volume control at the beginning of the performance and then leaving it there for the whole show.
You also get killed when you’re transferring vinyl records. One cat hair pop and you’re dead. That pop is now the dominant feature of the show.
(1) If I simply choose Amplify, without selecting an individual track, then Audacity will amplify all tracks based on the the track with the highest peaks. So suppose I have one track that is kind of high (snare drum) but the vocals are low. Then Amplify will say “amplify up 10” and bases that recommendation off the high track, but that is not enough for the vocal track. Etc.
Are you saying you have a multi-track recording with separate tracks for vocals, drums, etc.?
If so, “mixing” should be done by ear… The tracks in a multi-track recoding are not supposed to have equal loudness.
Note the peak levels have poor correlation with perceived loudness.
If you are trying to match the loudness of several songs (or your entire music library) look into ReplayGain (and/or the associated WAVgain & MP3Gain).
If you have a regular stereo recording and the snare it too loud, normalizing or otherwise adjusting the volume linearly will affect the snare & vocals by the same amount (just like changing the volume control before the song starts). You can use dynamic compression to make loud parts quieter or quiet parts louder. (But, sometimes the overall results of reducing (or distroying) the dynamic contrast are undesirable.)
I read Amplify is preffered over Normalize-- so is it?
They can both have the same result, depending on the settings.
“Normalize” means setting the peaks at (or near) 0dB. The Amplify effect does that by default, adjusting the left & right channesl by an equal amount and using the channel with the highest peak as the reference. Or, you can amplify by a defined positive or negative dB level.
The conceptual difference is that Amplify adjusts by a certain amount, and Normalize adjusts to a certain amount.
Are you saying you have a multi-track recording with separate tracks for vocals, drums, etc.?
Exactly. You should go through the performance (on good speakers) and set the individual volumes of the instruments and voices — manually — for a pleasant total mix. That mix is exactly the thing that will get messed up if you run Normalize in split channel — independent volume setting.
