What's the right way to compress a song?


I made a very long pause in making music. For some months now I trying to do constantly music. but I am very often lost in technical aspects of the song, especially compression. So much I sometimes do not focus on the important parts anymore like melody, structure etc.

Maybe 15 years ago I made a lot of music, mostly using cubase and later logic. In this time I never thought so much about compression, to be honest very often just did some master compression using mda dynamics with no attack and long release, and the sound was better than today. E.g. for the drums, I just compressed the sum of the beat and did no fancy sidechaining/env follower for the compression or getting room for the kick.

So what do you think is a good approach to compress a electronical dance song? Do you compress or even limit nearly all instruments individually, or do you completely go without compression? What about using a maximizer?

Currently I do:

  • maximize only basses
  • compress the beat and the bass in sum
  • use an env follower on the kick (for lowering bass or complete volumes of all middle or bass frequency instruments)
  • limit instruments that maybe overshoot by fx like flanger etc
  • compress the sum of all instruments with beat slightly
  • high passing most instruments to get room for bass and kick

What do you think could I make better here? I have seen a lot of maximizers on the master track. Doesn’t it kill fine nuances and depth?

I also read about some approach using multiband dynamics on the sum of the beat. Seems to be much more precise.


Seriously, if you’re turning out great results with just eq and levels, don’t fix what isn’t broken. Especially if its keeping you from writing music.

There’s some really good advice in this interview:


"You’ve got a background in mixing/engineering as well. For many, mixing properly can be a difficult task. Do you have any one particular tip that you think really benefits a mix in Live 9?

As someone who works with other people’s mixes all the time as a tutor and in an engineering and mastering context the advice that I am giving out time and time again is: STOP USING COMPRESSION! It’s hilarious when you’re watching YouTube tutorials and you hear someone say ‘OK, now create a sub bass line then add a compressor’ without even thinking about it, as though you can’t create a sound without using compression. With the Petrichor project I try to avoid compressors completely unless I really need one to sort out spikes in a recording or something. Removing dynamic range from your track removes emotion in my opinion, so if you’re trying to make music with any subtlety to it then ditch the compressors first and work on your mix without them. If you can nail your mix without compression, a mastering engineer is going to make it sound incredible!"

Oh and here:


"Your tracks really slam, and I’m guessing the compression this thing provides is part of that.

There’s actually very little compression [in my music]. Everything is raw. [Points behind his mixing desk] That’s why the compressors are all back there. I’m getting a little bit of warmth from the Avalon on the out. My mentors are people like Todd Terry, who recorded in mono. I record in mono, just two mono tracks. When you go to a club, how many have the panning to left and the right? You listen to records and can tell [producers have] spent ages on this stereo sweep, and I think, “You are never gonna hear that.” That’s why I don’t waste any time with those effects."