Re: Goofy q for the luthiers

alcarruth@xxxxxxx wrote:
RichL wrote:
"Consider too that in your extreme example above, where the walls are
parallel, only certain frequencies create standing waves by means of
constructive interference. So the frequency response would be pretty
uneven. You'd have "live" spots and "dead" spots as you swept the
frequency. Clearly such a situation would be undesirable in a
musical
instrument! "

In fact you can't avoid a certain amount of 'peakiness' in the
spectrum, and would not want to if you could. The 'flat response' that
is the ideal for speakers would be a terrible musical instrument, IMO,
precisely beacuse every note would sound the same. It would also be
quite inefficient, and we can't afford that on the guitar.

In the low range the guitar is a 'bass reflex cabinet', albeit not a
very good one by hi-fi standards. If that was all that was going on
(the walls were rigid, and the air inside well damped, as they do in
speakers) then you'd have two output peaks in the low and mid range,
around the open G string and an octave lower. Above that open G string
pitch the output would fall off smoothly at something like 6dB per
octave, iirc. Every note above the open G would have the same spectrum
if you played it in the same spot on the string. This would be b-o-r-i-
n-g.

The reason is that your senses are set up to pick up changes, on the
theory that things that don't change, like rocks, won't eat you, nor
can you eat them. You need a certain amount of variation to stay
interested, but not so much that it becomes chaotic, since that's just
the same old mud. So what we try to do is to get enough resonances
spaced out in the right way so that at least some of the partials of
any note you play will be reinforced. If you get it right all of the
notes end up sounding at about the same perceived level, and they are
clearly the same instrument, but they are different enough to be
interesting. We often do this by increasing losses.

Many, if not most, of the resonances above that open G pitch are
'losers': they are less effective at putting out sound than the top
and air in the 'bass reflex' range. Many of them end up being dips in
the spectrum. But without them you don't have any 'tone color'.

A number of research reports I've seen, dating back to the 70s, show
that there's a 'proper' number of resonant peaks per octave, and a
'proper' ratio of peak to dip height, and that too much departure from
these values tends to yeild an instrument that is either 'harsh' or
'dull', depending. One researcher, looking at classical guitars,
found, for example, that making the 'main top' resonance too
efficient, so that it produced a tall, narrow spectral peak, tended to
give a 'harsh' sound, and I can attest that that is so. Controlling
the height and width of that peak is part of the 'voicing' process,
whether the maker thinks of it that way or not.

And that's what so cool, and frustrating, about all of this. As has
been pointed out, the 'old boys' cooked up these designs, and got them
to work really well, without any of this high-falutin' theory. All
they did (ALL they did: HA!) was build lots of instruments and pay
attention. Of course, it realy helped to be working in a living
tradition, where you could make changes in the design, and to have
apprenticed with a Master who could draw on a huge data bass of lore.
Yuo also needed to be talented. Some of us are trying to substitute
modern theory and research methods for parts of that recipe that we
lack, but it's still early in the effort, and we won't know for a
while yet whether we can actually crack the code.

Alan Carruth / Luthier


I am reminded that C. F. Martin apprenticed in Germany before he immigrated to the U. S.

-Raf

--
Misifus-
Rafael Seibert
Photos: http://www.flickr.com/photos/rafiii
home: http://www.rafandsioux.com
.

Relevant Pages

  • Re: Goofy q for the luthiers
    ... is the ideal for speakers would be a terrible musical instrument, IMO, ... Every note above the open G would have the same spectrum ... Many, if not most, of the resonances above that open G pitch are ... 'proper' ratio of peak to dip height, and that too much departure from ...
    (rec.music.makers.guitar.acoustic)
  • Re: Goofy q for the luthiers
    ... is the ideal for speakers would be a terrible musical instrument, IMO, ... Every note above the open G would have the same spectrum ... Many, if not most, of the resonances above that open G pitch are ... 'proper' ratio of peak to dip height, and that too much departure from ...
    (rec.music.makers.guitar.acoustic)
  • Re: Goofy q for the luthiers
    ... is the ideal for speakers would be a terrible musical instrument, IMO, ... Every note above the open G would have the same spectrum ... Many, if not most, of the resonances above that open G pitch are ... 'proper' ratio of peak to dip height, and that too much departure from ...
    (rec.music.makers.guitar.acoustic)
  • Re: Measuring blued paper with a DataColor
    ... order to be able to compare the Lab measurements. ... You would do white point adaptation on all the measurements, ... In our paper blues case, we are trying to fix an instrument error, ... instrument makes the reference light only in visible spectrum ...
    (sci.engr.color)
  • Re: Meaning of Brightness for TFT monitors
    ... Why does Spectrocam clip the spectrum at amplitude ... Eye-One Pro calibration, whereas GMB software indicates ... The spectrophotometer is a 16-band color measurement instrument that reports ...
    (sci.engr.color)
Loading