Recording Articles

Recording String Quartets


I have reviewed a number of CDs recently where my overall comment has been that the recording was ‘Too wide and  too close.” Perfectly good musicians playing well and  creating a masterful rendition of a piece only to have it ruined by the recording quality.  When I use the expression ‘recording quality’ I am not  referring to  the  technology of the microphone,  pre-amps,  recorder and so, all of which work  brilliantly these days but there a re some basics that seem to me to be not right.  I will also hazard that they are not right because all training in recording technology  is focused on rock and pop technique (I hope I am wrong about this).  What we hear in these recordings of classical repertoire is firstly that the microphones are too close and that there are too many of them. This is very much a rock technique where, to get detail, and to reject noise, the engineer moves in to  capture the sound. But not being fluent with classical sound and form, does not really know how a ‘cello should sound in a recording so getting close in seems right to them – (about two feet away from the bridge). Similarly other instruments in a quartet will have close microphones at about the same distance. The wise engineer will also have placed a stereo pair a couple of meters away.

It’s what happens next that brings about the poor sound. Having captured a sound that is too close (you can tell it’s too close if you hear the  musicians breathing too much, or finger noises or too much bow sound) coupled with an out of balance sound of the instrument. Let me dive down a little.


There is a distance for capturing the whole sound of an instrument without it feeling as if you were on top of the player. You need a little instrument and ambient noise to help articulation of the rhythm (hearing a tiny tap of keys on a flute helps rhythm become clear for example ). You need little instrument noise for realism – one of the criticisms of samples is that they have no body noises, no extraneous human artefacts. There is a distance that makes sense – it also balances the sound at the top of the instrument with the sound emanating from the bottom.  Theorists have suggested that the microphones minimum distance in these cases should be at the point of an equilateral triangle whose base is the  major dimension of the instrument. This is fine if there are no directional projections of sound – the piano for example requires a lot more finesse.

One of the reasons for getting in close is to be able to exclude sounds of other instruments (I am sure some engineers would like to put their artists in booths to achieve this). The players however, being friendly and in need of eye contact, like to sit quite close together which makes sound separation quite tricky – hence they put the microphones in close.

But why do they need separate channels for each instrument?  Because they feel they might need to do some ‘mixing’ and balancing later? Most quartets balance the sound themselves as they play – they have been trained to. There are some  top names I have heard who were recorded in this multi-miked way and it is a little insulting even to dream of touching the fader unless in consultation with the players.

Another reason for close miking might be to suppress the acoustic of the recording venue (too wild? too boxy?) with a view to adding artificial reverb later . We all do it, but usually just a touch in mastering to put the varnish layer on the  recording – not as a prime component of the sound.

All of this leads me to say – first make sure the venue you have chosen is good for the job. Generally studios (unless very good indeed) lack the sound to make the players blossom and  blend as they would in concert.  Secondly recognise that  the sound of a quartet does not  exist in close proximity to the players – the sound blends in the air somewhere about 3 meters away and the use of a good stereo pair should suffice. Now – if you need spot mikes on each instrument they are for the gentlest touches to bring out maybe a weak viola sound. But here is a  golden rule –  never move the faders in the middle of the piece – I can hear you doing it.

I’ll say that again – if you ‘push’ the cello part in a section to  help it along, I can hear the increase in volume and it makes the whole thing sound comical. Less well trained ears might not know what has happened bu they will experience the image wandering. Don’t forget that panning is  more correctly called amplitude panning and works by  dropping one volume and raising another so if you raise the volume of  violin 1 then the sound will  move over to the left and will be evident.


Here though is the biggest sin of all and the greatest evidence of a pop engineer  working with material he/she does not understand.  Just because the violin is on the left and the pan control suggest left by twisting it all the way to the left does not mean you get a stereo image.  Back to the beginning…. Stereo does not mean anything other than ‘solid’ (look it up).  Just because the quartet sits left to right and you have knobs that suggest left and right does not mean it is a good idea to put the Violin 1 far off to the left of(f) the stage and the ‘cello far out to the right – about 50 feet from his fellow.  This is to make the sound too wide and break down the ensemble.  By which I mean – sounds which go together – such as a chord between the players no longer resonate together but are spread across the sound field. One recording I heard recently, in which melodic lines are handed from violin to Viola then ‘cello and back (It was a Beethoven Quartet) had me feeling like a tennis referee, so twisted was the sound, and so broken down that the identity of a few of the chords was in doubt.  When a major triad is formed by two instruments on the left and the 7th degree of the scale by the ‘cello on the right you run the risk of hearing a split ensemble – it is too wide for the ear to put it back together.


Well – I have re-mastered a few recordings that were worth the time and trouble so that I narrowed the stereo field to one in which it feels like the quartet actually sat together. But better and more generally – We need to train more people in Classical music recording and not pretend that it is the same as rock. It isn’t and by a very wide margin. Ability to read music is a pre-requisite but few engineering courses require it.  a Knowledge and love of classical music is also required. You cannot move from rock to Bach with the same set of skills.

Am I ranting?  I hope not. Ranters don’t solve anything and I hope above I have suggested enough to give pointers on what is wrong with so many new recordings.  If you are a player about to record – get to know the engineer and their limits – if they don’t know what you are talking about then get another one or be prepared to have to work very hard alongside them telling them how to do it.