Re: Is this true regarding digital recording?

In article <1J-dnVAHAsJ_3unZnZ2dnUVZ_vadnZ2d@xxxxxxxxxxx>,
gizzledgeezer@xxxxxxxxxxx says...


"The instant you digitize a signal, you destroy the phase-angle
relationship between the high frequencies and the lows.

Stop. This is no more or less true about digital than it is about analog.

Most recording processes upset phase relationships. In digital recorders,
the errors are introduced by the anti-aliasing filters (not the sampling and
quantization). These errors can be corrected, either before or after
recording. I have a Sony PCM-F1 with phase-compensated Apogee filters.

Analog tape recorders also have this problem. It's caused by
non-minimum-phase wavelength-related spacing-loss effects during playback.
(There's no room to explain it in detail here.) I assume the best open-reel
decks compensate for it, but it's not as easily done as with digital
recorders.

It's also caused by the simple fact that the response of an analog tape
recorder is essentially highpass and no practical analog highpass filters in
the 20 Hz range have constant group delay because it requires many
milliseconds of wideband, frequency-dependent delay to equalize such filters.
For example, the group delay of a single-order highpass 20 Hz analog highpass
filter is about 8 ms at DC, and this would require adding at least 8 ms to
the group delay at high frequencies. Practical analog filters can't do that.

What about highpass Bessel filters? Little-appreciated fact: the desirable
group delay properties of a lowpass Bessel filter are not carried over to the
highpass transformation (using the classic s -> 1/s lowpass to highpass
transformation). The shape of the group delay curve does not change much, but
the part of the group delay curve that was relatively flat in the lowpass
Bessel filter moves into the stopband of the highpass case, while the very
non-constant group delay in the stopband of the LP Bessel filter is in the
passband of the HP Bessel filter.

[Although irrelevant to analog tape recording, you make symmetrical, constant
group delay FIR digital highpass filters (realized directly or via fast
convoluton), but they all have VERY large delays if you need reasonable
selectivity and a 20 Hz cutoff. These filters will, of course, be non-
minimum-phase.]

.

Relevant Pages

  • Re: iPod loses again...
    ... So, AGAIN, what sound quality battle did it lose? ... That's the Zen Nano. ... Analog is reality. ... professional analog tape recorders to record symphony orchestras and jazz ensembles back in the mid 1970's (I was the recording engineer for NPR's "Jazz Alive" series in those days, for instance) I know what 15 ips/half-track tape can do and I certainly know the specs. ...
    (comp.sys.mac.advocacy)
  • Re: The Old Vinyl vs. CD Debate
    ... I never said vinyl was "better" than digital. ... One personal reason I think vinyl may sound superior to some people is ... audible result which is naturally defined by the original recording ... It's funny how some flaws in analog such distortion, tape grain, ...
    (rec.music.classical.recordings)
  • Re: MS Vista with service pack 1
    ... everything that I used to sleep through as recording to memory is much ... Daughter records lots of TCM movies and such using Media Center in Win ... The TV card has three coax inputs, analog tv, digital tv and fm radio. ...
    (soc.retirement)
  • Re: The Limits of the LP
    ... analog tape, ... What do you consider the advatages of analog recording to be? ... this time after time in th form of tape hiss and the ability to hear ... Tape hiss is indeed one of the sorry artifacts of the analog medium. ...
    (rec.audio.opinion)
  • Re: Krueger mostly responsible for own issues
    ... fact that soooo many studios around the US have Pro Tools with 888's ... Why should I demand better? ... to demand the makers of digital recording products do better. ... >>and the fact that I record analog has something to do with that. ...
    (rec.audio.opinion)
Loading