Saturday, April 06, 2019

Maybe I'm doing it wrong

I wrote yesterday about a song which I had recorded, hinting at problems encountered when mixing it. Whilst this might be clear to me, I realised that not everyone knows about the different terms used in this process.

Mixing is probably the simplest: this means taking any number of audio tracks and creating a new track by blending - mixing - the tracks. This primarily involves setting relative volumes between the tracks but also means using effects to highlight portions or to create a better sound.

In the old days, studios used tape recorders which had 2, 4, 8, 16 or 24 separate tracks (there were also 32 and 48 track recorders, but I think that these were obtained by using two machines simultaneously). A sound (like vocals, bass guitar, etc) would be recorded on a track, which was separate and not affected by any of the other tracks. Of course, the tracks all had to be in time. 4 track recorders were in use in the mid-60s; 8 tracks came into use 68-9; 16 tracks in 69-70 and 24 tracks by the mid 70s. Having more tracks enabled each instrument to be recorded separately and several times, but it made mixing much harder, which is why computer aided mixing appeared in the late 70s. My recording setup 'cheats' slightly: first, I record a music track which is effectively the final mix of all the instruments. Then I mix this music track with vocals.

There are three primary effects which I use on the vocals: compression, equalisation and reverb. Compression evens out volume levels; loud sounds remain loud, but quiet sounds become louder. I normally use this in order to provide a more consistent volume when singing, but I also use it on the final, mixed, track.

Equalisation (eq) is a souped up tone control. This allows different frequencies to be decreased or enhanced in the sound. Normally I reduce all the bass frequencies in my vocals, as these might clash with the music, but they also thicken the sound too much. I frequently enhance higher frequencies to help the voice stand out. Commercial studios spend a great deal of time with equalisation, 'carving out' space between different instruments so each one becomes clearer.

Reverb is a contraction of 'reverberation' and is the easiest effect to understand. Compare yourself singing in a room with open windows as opposed to singing in the bathroom: the sound bounces off the walls in the bathroom, creating echoes, otherwise known as reverberation. A room with open windows and walls which don't face each other, along with carpets and furniture, will absorb the sound and no echoes will be created, resulting in a dull sound. So one adds reverb to a recording, in order to add depth. There is also a relationship between the speed of a song (or part) and the reverb: short notes require a short reverb, as a long reverb would cause the echoes to dominate and swamp the original sound. Slow and long notes can sound exquisite with a long reverb - think of choirs in a cathedral.

My usual effects chain on vocals is first to compress, then equalise and finally add reverb. What happens sometimes - and this is what happens yesterday - is that the frequencies enhanced with eq cause certain sounds, especially sibilants, to stand out; the reverb causes these frequencies to become even more enhanced, creating many unwanted peaks in the volume, which is exactly what I don't want.

After thinking about this a little while walking the dog last night, I hit upon the following solution: change the order of the effects. First add eq, then reverb and finally compression, which smooths out the volume. I did a quick mix after returning home with this new order: the result was much better than all the previous mixes. For some reason, compressing after reverb causes the reverb to be more noticeable; as a consequence, a more minimal reverb has to be used.

So (with thanks to Randy Newman for the obscure song after which this blog is titled) maybe I've been doing it wrong all these years. I haven't decided whether to rerecord some of the vocals (primarily the opening verse) but otherwise, the mix settings are very good.

No comments: