Safe Sound Audio P1 User Manual - page 27
P1 Audio Processor
User Manual
Page
27
But can’t I just record in such a way that this maximum audio level is never reached?
In theory, yes you can; if you are very careful and record well below the maximum
permitted level of your recording system; but as so commonly happens, you set up
the levels carefully during rehearsal but when the actual recording takes place, the
vocalist has turned from shy retiring folk singer to the rock legend from hell! And the
once in a lifetime performance is ruined by digital clipping which is almost impossible
to repair even with the might of software based audio processing tools.
So it’s much better to be safe than sorry, and after all we are Safe Sound Audio!
The second issue is audio noise and distortion. Let’s assume you leave 12dB of
spare headroom for unexpected vocal excesses. So you are setting the MAXIMUM
recording level at -12dB during rehearsals. Let’s also assume you are using a typical
affordable 24bit recording system which will present a usable dynamic range of
around 100dB RMS (around106dB A-weighted).
So now our safe usable dynamic range has been reduced from 100dB (-12dB
headroom) to 88dB. Well it’s not too bad and it is in the same ballpark as the
dynamic range of commercial CDs.
Less well known is that the wonderful distortion figures quoted by digital recording
systems, typically 0.003% or better, are only achieved when every single bit of the
front end A/D converter is being exercised and this only happens when you pump in
the very highest audio level which the converter will accept before digital clipping.
This level is called 0dBFS.
So, allowing our 12dB safety margin, a quiet vocal phrase may only be peaking 20dB
below this maximum safe level (down at -32dBFS). In the analogue world this will
makes no practical difference to the achievable distortion, but in the digital world,
lower levels into the A/D means less bits representing the audio means more
distortion. Typically a -32dBFS signal will achieve a distortion performance of around
0.01%. Not quite the dazzling figure quoted in the sound card spec, is it? Once we
get down to –60dBFS distortion degrades to around 0.3%.
So there are audio performance advantages in getting a decent level of audio into 24
bit digital recording systems. If your working 16 bit then it’s even more critical.
How the P1 limiter manages dynamic range
Despite the claims of many soundcard and A/D chip manufacturers, the usable
dynamic range of most affordable 24 bit soundcards is around 100dB. We define this
as;
‘the difference in dB’s, measured RMS unweighted, between the maximum signal
which the soundcard can handle, called 0dBFS, and the minimum tone level which is
audible (not necessarily the same level as the soundcard’s noise floor).’
_________________________________________________________________________________