Path: archiver1.google.com!news1.google.com!newsfeed.stanford.edu!news-spur1.maxwell.syr.edu!news.maxwell.syr.edu!diablo.theplanet.net!news.theplanet.net!not-for-mail
From: "henri" <[email protected]>
Newsgroups: rec.music.classical.guitar
Subject: Re: Bach Lute Suite #4
Date: Sun, 23 Feb 2003 14:05:52 -0000
Lines: 65
Message-ID: <[email protected]>
References: <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]> <[email protected]>
NNTP-Posting-Host: 81.76.196.219
X-Trace: newsg3.svr.pol.co.uk 1046009375 16441 81.76.196.219 (23 Feb 2003 14:09:35 GMT)
NNTP-Posting-Date: 23 Feb 2003 14:09:35 GMT
X-Complaints-To: [email protected]
X-Newsreader: Microsoft Outlook Express 4.72.3110.5
X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3



William D Clinger wrote in message ...
>Matanya Ophee wrote:
>> When a tape moves across a _fixed and stationary_ head, the pitch it
>> produces is directly related to tape speed. But when the head is not
>> fixed, but rather _rotating_, the pitch and the tempo can be altered
>> at will by modulating the head's rate of rotation _and_ the tape
>> speed.
>
>LOL!  No, Matanya, rotating the tape head, as was done for example
>in some video recorders, does not change the Hz*sec invariant: the
>tape has some fixed number of peaks on it, and exactly that number
>of peaks is going to pass beneath the head no matter how you move
>the head and/or tape.
>
>If you want to change the tempo without changing the pitch, you
>have to resort to fairly sophisticated signal processing.  That's
>why this was not generally possible until the advent of digital
>audio.
>
>> Here is for example, a machine that was widely available in the
>> 80s:
>>
>> Http://www.orphee.com/rmcg/ad.jpg
>>
>> As you can see, this come from the December 1988 issue of Guitar
>> Player magazine.
>
>ROTFL!  This ad clearly says "two speeds".  You can play back
>at the original tempo or pitch, or you can play back "at half
>speed...exactly one octave lower".  Half speed implies twice the
>time.  Thus the Hz*sec invariant is preserved by this machine.
>
>> Similar _professional_ level machines were in use at
>> the time I worked for Nagra, 1962-63  and every decent recording
>> studio had one.
>
>No doubt.  These machines could not have performed the miracle you
>claim.


Also, altering pitch is not just a question of altering the fundamental
frequency ( the major regular peaks in the analogue waveform). The harmonics
of that frequency are embedded in the signal and will show as a ripple of
peaks in the frequency spectrum. To preserve the acoustic integrity you need
to remove the effects of the original harmonics, and compute and apply new
harmonics to the spectrum as simple multiples of the new pitch, and even
this will only work for alterations within a restricted pitch range,
otherwise the characteristic timbre of the instrument gets distorted

I worked for a major acoustic research lab from 1970s - 2000. Pitch
extraction of voice contours was a main area of speech research. We had
professional quality recording studios and several specialised analogue
electrical techniques for extracting pitch contours. However I am certainly
not aware of any commercially available devices which would have allowed
fundamental pitch alteration in analogue recordings without corresponding
distorting alteration in speed/tempo of the whole recording. Had this
facility been available, it would certainly have been exploited in research.
Such techniques were only really exploited when specialised digital signal
processing software became available allowing the digital/mathematical
separation of pitch and harmonics from the spectral envelope and resythesis
at a different pitch or with a different pitch contour.

Henri Lascalles