NetNews Usenet Archive 1993 #3/NN_1993_3.iso/spool/bit/listserv/allmusic/15314

Text File | 1993-01-21 | 5.0 KB | 95 lines

Comments: Gated by [e-mail]
Path: sparky!uunet!paladin.american.edu!auvm!LSUVM.BITNET!HEROY
Message-ID: <[e-mail]>
Newsgroups: bit.listserv.allmusic
Date:         Thu, 21 Jan 1993 17:01:37 CST
Sender:       Discussions on all forms of Music <[e-mail]>
From:         Paul Heroy <[e-mail]>
Subject:      Technonerdophilia
Lines: 84

On Thu, 21 Jan 1993 14:16:05 EST Ilia said:
>
>I am pretty sure that our perception of the spectrum
>depends upon the overall volume of the sound, and thus the volume
>controls in hi-fi systems usually compensate certain frequencies when
>the volume is changed (I think).

That's true - for decades the reference for this has been a series of
hearing-response/frequency-level plots called the Fletcher-Munson curves
(after the researchers who performed the experiments, of course). Generally,
the human ear is much more sensitive in the middle region (approx. 200-800
Hz), which (surprise!) corresponds roughly to the range of the average human
voice. Frequencies above and below this are usually boosted by the 'loudness'
buttons found on many stereo receivers, which are a compromise since most of
them don't take into account volume level, much less speaker variances (which
would be pretty difficult and impractical anyway). More recent research has
varied the shapes of the curves some, mostly by lowering the treble boost
needed, and this is reflected in the character of 'loudness' buttons on most
recent preamps/receivers - they mostly boost the bass, not the treble.

On Thu, 21 Jan 1993 18:46:42 GMT Jeff Preston said:
>  For the most part, I think you're right on this count, but I don't
>have that much faith in the engineers. You'd be surprised how "low" the
>peaks are on some of my CDs. But in general, I don't have to adjust very
>far in either direction switching between CDs. Certainly nowhere near as
>much as you would on vinyl.

Yep. Er, BTW - are my posts getting out? I seem to remember saying just this...
(I think there've been some network problems this week, actually.)

Anyway, because digital recording has a very well defined, sharp transition
from a 'safe' level to a 'saturated' level, you've gotta be a little more
careful to leave headroom, because digital distortion is sudden and very
unpleasant, whereas analog distorts more slowly and less unpleasantly.

>>if the processing is all done digitally, I suppose the loudest
>>bit can be found,
>
>  Yeah, that "loudest bit" thing makes perfect sense. I suppose that's what
>Marsha's CDP-297 is looking for when she pushes the "peak search" button

I've noticed that feature on a few CD players, and it's really handy to have
if you do a lot of taping. I remember how, with vinyl, you could just look at
the record grooves at the right angle and find the loudest part of the record
by the way the grooves looked. Gosh, vinyl actually had an advantage??? ;-)

>>of course the levels jump around - it is not a fixed level of output,
>>it is a fixed *maximum* level of output.
>
>  Has to be thataway... there has to be a "largest number" you can write
>in that format. 16-bit?  Who's the compunerd here?

I don't remember the exact math, but every bit basically adds 6 dB to the
range, giving a nominal 96 dB range for 16 bits. How then do most CD players
claim a dynamic range of over 100 dB??? (Actually this is legitimate and
depends on the way you measure noise, or something like that.)

>  The one factor you leave out here, TTT, would be the frequency response
>of the tape deck *and* of the tape itself. A max bit at 50Hz versus a max
>bit at 1000Hz versus a max bit at 17,500Hz -- which one will likely peg
>the meters more?  If you can say "all the same," you gots a nice tape
>deck!  Still, the tape would know the difference.

Ya... you definitely don't want to push an analog deck as hard when you're
recording, say, jazz with lots of cymbals, brass instruments with lots of
HF overtones etc., as opposed to solo acoustic guitar and voice. With digital
you don't have to worry about the frequency content, but overall levels are
the killer, as I mentioned above.

One other thing I'll mention wrt this is that some decks are a lot better
about telling you what's going on than others. When you're reading those
little meters, what you *really* want to know is what's going on the tape
vs. what's coming off the tape. If you've got a nice 3-head deck that allows
playback of the tape as it's recording, your meters should monitor the signal
that you have selected. In other words, if you're listening to the source
that you're recording, your meters should be reflecting that signal, and if
you're listening to the signal from the tape, the meters should reflect that
instead. If there's much of a discrepance when switching, something is wrong.
The problem is that some decks screw up and don't switch the meters from the
source signal, so it's tougher to tell what's going on. Of course, since this
only applies to 3 head decks (like my Nak :-), you *should* be able to *hear*
that there's a problem too!

Paul