This is because bass frequencies take up a lot of headroom in the mix, and can very easily make a mix “muddy”. Most recorded sounds have some bass information - even hi-hats and claps. If you add a high-pass filter to pretty much every channel, you can raise the EQ frequency threshold until only the information you want remains. Quite often, rogue bass information won’t be audible, but it DOES effect the final mix (a good way to check is to use a spectrum analyser).
You’ll naturally mix anyway as you compose and arrange, and once or twice my dedicated mix down hasn’t sounded as good as when I was arranging! In these cases, you can always open the original session and work out what was better.
Try grouping (or bussing) similar tracks together and processing them as one channel, e.g. group all the drums into one channel, then apply compression and EQ to that whole group to get a cohesive “gelled” sound.
This is a pro tip for those who need to provide two mix-lengths of a track (e.g. a 5 minute version for club DJs, and a 3 minute version for radio). If you have two separate sessions, tweaking the main mix in one means you then have to open the radio edit project and manually repeat the same changes. Keeping the radio edit in the same project as the club mix (just further along the timeline) means that any mix tweaks are automatically applied to both versions.
This is done for three reasons:
As I wrote at the beginning of this post, not all of these tips need to be applied to every mix. If it sounds good, don’t worry about adding parallel processing or tuning your drums. When it sounds right, it sounds right.
Create space in your mix by side-chaining certain sounds with others. For instance, use the kick to slightly compress the bass, or use the vocals to sightly compress the synth’s higher frequencies.
This was a game changer for me. I remember I had just finished a hot new track, and was DJing later that night. I dropped it at peak time, and the vocals were completely missing! This was before I knew about phase cancellation. Some stereo sounds will cancel themselves out in mono, so always mix (or at least check your mix) in mono before working on the stereo field. You can do this easily in Ableton Live by adding a “Utility” plugin on the master chain, and assigning a keyboard shortcut to toggle the stereo width between 0% and 100%. All DAW’s will have a similar capability.
So, how do you get a phat sound without slapping effects on the master channel? Ah, the million-dollar question. In a nutshell, by using the techniques in this list, as well as advanced techniques such as parallel processing. Parallel processing is basically mixing an affected signal into a dry signal to maintain transients, and compression works really well in this way. You get the in-your-face compressed signal, but the transients of the dry signal cut through and maintain the dynamics.
There is so much about this online, and - as a heavy practitioner of both - I can confirm that mixing WITHOUT mastering effects on the master channel is definitely the way to go! For years, I started by adding compressors and limiters to get my kicks sounding as phat as Deadmau5’s, and - whilst I achieved it - by the time I brought everything else into the mix it was a horrible, squashed, distorted mess with no dynamics. A mastering engineer will want to receive your mix peaking at -6dB (and if you’re doing the mastering, you’ll want the same) so sending them a completely squashed 0dB brick will not allow them to do their job properly.