Comment from a Audio Masterclass visitor…
I enjoy your newsletter very much. I have a question regarding the 'air', or extreme high frequencies I hear so often in the vocals on modern pop songs. I'm trying to determine if this is a result of the mic/pre/converter combination, or if it can be brought out during the mix with proper compression and EQ.
I've got a very nice set of vocal mics (Peluso), and a middle of the road compressor (Joe Meek TwinQ) running S/PDIF into an M-Audio Delta 410.
Seems like some of the air is there during recording, but I cannot get it to come out during the mix. I've tried sweeping 10k-18k with my digital EQ, but only seem to bring out harshness, not the smooth high-frequencies heard on commercial CD's.
Is there just more to this than I have capability/equipment for?
What is the range of frequencies I am hearing on commercial CDs? And what is the secret to capturing it, and making it smooth, not harsh?
Todd Johnson, USA
David Mellor replies…
The frequency response of a CD can extend at absolute maximum up to 22.05 kHz, which is precisely half of the sampling rate of 44.1 kHz.
If it were attempted to record a frequency higher than 22.05 kHz, this would result in an 'alias' tone, which is musically unrelated to the input signal, and very unpleasant to the ear.
In practice, the frequency range is limited further to 20 kHz, so that high frequencies may be filtered gently – a steep filter can introduce unpleasant audio artefacts of its own.
Many adults cannot hear frequencies above 15 to 16 kHz. At least they can't reliably hear them, but in tests they will often show that they can somehow 'perceive' these frequencies without being sure that they are hearing them.
It seems fair to say that it isn't frequencies in this range that are creating 'air', and I think that spectrum analysis would demonstrate this. It would be interesting to hear an example of 'air' as you perceive it.
My view on 'air' is that it exists in the frequency band of around 8 – 12 kHz. But it isn't simply a matter of level.
If you boost the level of frequencies in this band, you will, as you say, achieve harshness rather than a desirable 'sheen' – there's another word that is often used as a synonym for 'air'.
The reason for this is that by boosting the high frequencies with EQ, you are boosting only what is already there.
What is necessary is the addition of extra high frequencies that were not there previously.
This is done by the controlled application of distortion. The easiest way to generate controllable, desirable second-order harmonic distortion is to use a tube microphone or a tube preamplifier.
It is possible to use tube processors at a later stage in the signal chain, but it is oddly more difficult to achieve the same result, and I would say that the earlier in the chain the tube comes, then the better the sound is. So that makes having a tube microphone the best option.
Hmm, I can feel a test of tube microphones coming on…