Update: Be very careful where you put your metering plugins when mixing. It turns out I had Insight one step before the final mix. As a result, my mono mixes are now about -2 to -2.5db softer than the stereo sounds, not the same as the article below says. This is acceptable, though not ideal. The sections about compression and maximizing are still good things to do in your tracks, particularly if, as in my case, the instruments are overall not loud but do have extremely high peaks from time to time. Any advice on how to get mono and stereo mixes to exactly the same level would be appreciated. Above all, this post is to emphasize the need to make sure your mono mix sounds as good as your stereo mix, even if it isn’t always as loud.
Whether you are new to audio mixing or a veteran you’ve hopefully heard people say that you should always check your mix in mono. How come we need to do this? I discovered that a Zoom meeting I was providing music for was only setup for mono audio and even if the meeting is in stereo, all recordings of Zoom meetings are in mono. There are also many venues that only play music in mono including restaurants and clubs. Expect also that some of your listeners will be hearing their music on phones or tablets that only have one speaker or poor at best stereo playback.
Listening to your mix in mono can also point out weaknesses in your overall mix. In some cases a mono mix will result in instruments getting lost or buried in the mix, sometimes due to phase cancellation. Variations in phase, like when you take two sine waves and shift one 180 degrees so you don’t hear anything, as you go from 0 to 180 you will still hear a tone, but it will get softer and softer. Examination of your master track with various metering plugins (there are many free ones) can show if you are suffering from phase cancellation. Some areas to look for are phase variances between snare mixes above and below the snare; overdubbing of the same part with different settings or effects; stereo wideners and synthesizer type instruments that have a lot of processing applied.
My problem was that the mono version (as I found out from the recording) was 3db to 4db softer than the stereo version and softer than voices on the recording. As some people will point out, it is not unusual at all for the mono to be a bit softer than the stereo, but 3db is noticeable and more importantly the mono mix wasn’t as good as the stereo. The recording I was working on was a one instrument stereo track, a virtual software instrument piano, with various processing on it in the instrument software.
The solution I came up with was to get rid of any processing I didn’t really need in the instrument (like a stereo widener), put a compressor (Native Instrument’s LA-2A emulation) on the track aiming for about 4db in reduction, then a maximizer (I used Ozone 9) to aim for -15LUFS which seems to work best for me in Zoom. Each recording, whether with one instrument or many, will require adjustment to the compressor and maximizer settings as no one setting works for all situations. Right away this made a positive difference.
I frequently “freeze” my midi based tracks to audio so I can work only in audio and also so I’m not tempted to spend forever tweaking the midi parts. Next I found that cutting off some of the really low frequencies (below 60Hz) with a gradual (around 12 db/octave) slope as the first thing in the audio processing helps. Exact frequencies for the HPF and the “Q” will depend on the instrument. Excluding these lower frequencies helps the compressor to focus on musical elements rather than low frequency rumble. Don’t use a HPF at 60HZ on an organ, kick or electric bass, instead aim for below 40 or 50Hz, if at all. As to the compressor I used for one of my recordings, since the NI LA-2A emulation has no attack/release or ratio settings and its own HPF, I used the Peak Reduction to set the compression and didn’t use an EQ HPF prior to the compressor. While I like the LA-2A emulation, I also used the built-in ReaComp compressor in Reaper (my DAW of choice) with a 4:1 ratio and 300ms release with everything else being pretty close to the default. I found that it worked pretty good too, but I do have to say the LA-2A emulation requires less thinking and tweaking.
(See opening paragraph for corrections to this paragraph). For the final check before rendering to the stereo file that will be used for the Zoom meeting I listened to everything in mono and stereo along with an LUFS metering plugin (I think the free Insight plugin from iZotope). Exact LUFS levels happened between stereo and mono in some recordings and those mono mixes that didn’t were only about 0.2db softer than the stereo. I hope this helps or is at least interesting. I would double check everything I said before taking it as gospel. If you have any corrections, suggestions or other methods for making sure mixes are good in both mono and stereo, please comment below.