It is interesting how you get different aspects pointed out from different=
listeners.
What I heard was enharmonic distortion (but not a lot) and I confirmed this=
by listening at quarter speed and 1/16 speed, with audiograms and with and=
without frequency filtering. Enharmonic distortion is at frequencies not
related to the original signal and is much more audible at low levels than=
harmonic distortion. On a single note, several percent of harmonic
distortion is often inaudible. What we complain about is enharmonic
distortion when less than 1% can be unpleasant.
The peak distortion near the end is slight and in practice inaudible as it=
appears to be less than 1dB overload. When this happened, (before or after=
digitisation) I don't know, but with many recorders, levels like this
trigger an automatic volume control giving a "hole" after the peak which i=
s
a much more audible defect than a simple peak distortion.
The answer, as ever, is to record low. I would advise aiming at peaks comin=
g
in at -10dB or -12dB below 100%. We are long past the era when you have to=
cram everything on as high as possible to overcome tape hiss, but the habit=
sometimes lingers on. Plenty of headroom is essential with nature recording=
where peaks cannot be predicted. Out of doors, natural noise will almost
always swamp out any digital noise you will hear after lifting the playback=
level to peak to -3dB or -6dB.
Note about Blumlein recording (and ORTF etc). Any rig using spaced mics wil=
l
introduce phase shifts between left and right. This sometimes enhances the=
stereo effect on headphones, but with open speakers, the image you hear is=
very dependant on the room acoustics. The is why most studio mixes use leve=
l
stereo using pan pots rather than phase stereo to give a more predictable
stereo image.
David Brinicombe
|