Hey everyone! So, we are recording a podcast, which ends up being two 40 minute or so mono tracks for our mics, which I am then splitting in multiple places to insert music samples on a separate stereo track - I split the vocal track so I can slide it to the exact time we want for a fade in/fade out.
When everything is done with editing, I go to add compression to the vocal tracks. I kept getting too-quiet results, so I kept upping the ratio [I’m currently at 8:1].
I then realized I get drastically different results if I select the whole 40-minute track then compress, versus selecting each section individually and compressing. I suppose this is due to correcting the offset after compression, and that doing the whole track at once will be limited to that one peak, whereas compressing each section localizes a bunch of different peaks.
What is the best practice to proceed with this? I feel that just compressing the whole track will make the whole track quieter than it needs to be because of one errant loud sound somewhere in the 40 minutes, but compressing each section separately is tedious and may end up amplifying things we don’t want amplified.
Compression is “complicated” and all of the settings depend on what you’re starting with and what you’re trying to do.
8:1 should be good enough but the amount of compression you actually get depends on the threshold and the attack. You only get compression above the threshold and after the attack time has passed.
If you have time you might want to adjust the loud & quiet parts with the [u]Envelope Tool[/u]. Pro mixing engineers adjust the levels “manually” before using compression or limiting. The “trick” with the Envelope tool is to fade up & down and avoid any sudden jumps in volume.
And try the Limiter. Limiting is a fast kind of dynamic range with an infinite compression ratio and instant attack, or in the case of Audacity’s limiter it has look-ahead which is negative attack. There are fewer settings to foul-up.
I feel that just compressing the whole track will make the whole track quieter than it needs to be because of one errant loud sound somewhere in the 40 minutes,
Compression and limiting “push down” the peaks which evens-out the volume but it can make it quieter. The normal process is to use make-up gain, or Amplify or Normalize to bring-up the loudness after the compression gives you the headroom.
There is a [u]Recommended Audiobook Mastering Process[/u] and that might be a “good start”. With a podcast you don’t have the same strict requirements so you can limit to 0dB. And if it’s not loud enough you can use a higher (less negative) RMS target, and you can also use LUFS instead of RMS. (LUFS is a better measure of “loudness” than RMS but if your podcast is mostly dialog it won’t make a big difference.)
There is a recent free (donationware) plugin which does something similar,
called “LUveler” but it has controls … https://bedroomproducersblog.com/2022/05/19/luveler/
I have used used LUveler in Audacity2 where it works in real-time,
but as it’s not VST3, it’s not real-time in Audcaity3