Cyber ears

[Richard Brekne]

Ron Nossaman wrote:

> >I have thought much along those lines. A tuning machine that simply repeats
> in real
> >time the actual beats per second registered for any given interval. This
> would leave
> >much more control in the hands of the tuner. It would also be an invaluable
> learning
> >tool with regards for getting a feel for beats per second.
>
> Richard,
> Back in 1978(?) at a regional seminar here in Wichita, someone showed up
> with a couple of prototype machines. One had a digital readout, and was
> intended as a tuning aid, I think, and another was a beat enhancer with
> microphone and earphones to pick up, clarify, and amplify the beats during
> tuning. I didn't pay much attention to them at the time because I still had
> too many basic survival skills to learn and didn't have any brain cells to
> spare. I don't know what became of either project (anyone remember this?),
> but what you describe here should be quite doable in software. It would have
> to take measurements from the piano and compute a tuning, like existing
> programs, and figure interval beat rates from the measured partial
> structure, adjusted to the computed tuning. I'm just not sure how much use
> it would be, since the actual frequency of the notes would have to be
> figured to generate the beat rates, the visual display gives you an
> indication of where you are. Then again, a second visual indication of the
> actual measured beat of a test interval against the computed optimal beat
> rate could, indeed, be interesting.
>
> Then again, something that "merely" measures and records beats of ascending
> interval progressions and visually graphs them without judgement as to the
> frequencies involved could be a flexible and comparatively simple aid to the
> aural tuner trying to set temperaments of any particular flavor or blend
> across scale breaks.
>
> Interesting thought.
>  Ron

Hi Ron. Thanks for the interesting reply. A bit of a different angle then my
origional thought, but I like it.

What I had thought of was something that does nothing else then tell me within,
say 3 decimal points, how many beats per second for all coincidentals (within
reason, for example all coincindentals upto the 12th partial of the lowest note
in the interval) for any given test interval.

It should show the results of each measurement in table form displaying each
coincidental it encounters and the corresponding beats, and it should do it in
real time or as close to it as possible.

My thinking was as follows

 1: We deal in beats per second, and have all kinds of specifications as to how
many beats per second any given interval should have.

 2: We also have a good picture in our minds of how different intervals relate
to each other. (For example the old 7,8,9,10 beats persecond White made such a
point out, or the idea that major thirds should beat twice as fast as the same
interval taken an octave lower, etc. etc.) Also;  we know how to apply these
specs to a larger perspective which we call "tunning".

 3: We have a good knowledge of why and how to "tweek" these specs to adjust for
different scaleings, inharmonicity problems, etc.etc.etc.,


One of the major problems confronting all tuners is, when it comes right down to
it, hearing. Its actually amazing that we do as well as we do when you think of
it, The human ear is truly incredible. In spite of this, such a device could
enhance the ears capacity by providing visual confirmation and an accuracy level
we are not able to reproduce. (I challenge anybody to tune a major third at
6,932 beats persecond consistently) Now the point may not be to tune a third at
exactly 6,932 bps. Yet I get the feeling that being able to see exactly how many
bps (not just the first coincidental but say the first 4 or 5 for any interval)
could be valuble input.

To my knowledge this approach has not been tried. That being the case who knows
what a little good old fashion "impiri" could turn up in the form of new ideas
by giving it a try. In any case such a tool should be relatively simple to
build, inexpensive (ie affordable to any tuner), basically useless to anyone who
doesnt know a bit of piano tunning theory, and a great aid for getting a feel
for beats per second, as mentioned before.

I'd like to hear from you guys out there doing programming on the side as to the
do-ability of this, and anyone elses thoughts for that matter.

thanks again

Richard Brekne