[music-dsp] Pitch

[Iain Sandoe]

Hi laszlo,

On Fri, Aug 17, 2001, TÛth L·szlÛ 
> On Fri, 17 Aug 2001, Iain Sandoe wrote:
>
>> I would say it *is* possible but that the degree of success and the time
>> taken to achieve this are not well-specified or boundable respectively.
>> A friend and I used to 'score' progressive rock ...
>
> My point was that scoring a complex music is NOT equivalent to separate
> the signal sources. The result of the first operation is a score sheet.
> The result of the second one are the originally mixed signals themselves.

OK.  I will concede that the result is different - I would say, though, that
the ability to produce the first result suggests (strongly) that, somewhere
within the processing, the individual sources must be sufficiently well
separated to be identifiable.

I'd never claim that the result is perfect - it's sometimes extremely
difficult to hear what's going on.

> And I am looking for not what they might have been but what they exactly
> were!

In that case, I don't suppose that you will be successful ;-)

'exact' is a very difficult concept in engineers ;-) ... most of the time we
settle for "sufficiently useful/good estimates".

ciao,
Iain.

dupswapdrop -- the music-dsp mailing list and website: subscription info,
FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/