online pitch source

[Jason Kanter]

This program, like all the others, generates sound based on the computer's
internal sound chip, which can be inaccurate. The only way to make it
accurate is to calibrate to a known external source such as NIST, using a
calibration system like that in Tunelab.

Question for Bob Scott: WHen Tunelab calibrates, does it make a registry
change that will then apply to other sound-generating applications? In other
words, can you use Tunelab to calibrate a laptop and then use a different
app (e.g. NCH Tone Generator) on that laptop and trust that it is now
accurate? 

-----Original Message-----
From: pianotech-bounces at ptg.org [mailto:pianotech-bounces at ptg.org] On Behalf
Of Geoff Sykes
Sent: Monday, May 15, 2006 12:12 PM
To: 'Pianotech List'
Subject: RE: online pitch source

I just remembered this nifty little program I have on my computer that might
be useful here. It is a tiny little downloadable program called NCH Tone
Generator. It has something like a 30 day free trial period, but it only
costs about $20 US. 

http://www.nch.com.au/tonegen/index.html

-- Geoff Sykes
-- Assoc. Los Angeles


-----Original Message-----
From: pianotech-bounces at ptg.org [mailto:pianotech-bounces at ptg.org] On Behalf
Of Philippe Errembault
Sent: Monday, May 15, 2006 4:31 AM
To: Cy Shuster; Pianotech List
Subject: Re: online pitch source


Hi all,

With a computer, you have two possibilities, both can come from a file, so
having the ressource on line will not help (except if you don't knowhow to
download it on you computer). Only "usefull" online ressource I could
imagine, would be a reference timebase like we use in the network time
protocol. The problem with such a ressource, is that on the short term the
delay introduced by the network will completely garbage it. This is only
usefull to have a precise time of the day.

Now, about the sound files themselves, In all cases, you will be depending
on clock frequencies generated on you computer. Those clocks are generated
by quartz which have a relatively good precision, but which will most
probably be more precisely tuned for a time related clock than for a music
related clock. now, I expect that electronic synthetisers will be relatively
precisely tuned, or you would have a problem playing with more than one. I
don't know, it might be the case...). Instead of using your computer, you
have another option which is playing this from an MP3 (or alike) walkman (In
this case, I the error will probaby be larger)

you have two possibilities : Either you use recordings ("wave kind" files)
or synthe "midi kind" files. The waves can be compressed (.mp3 files, .wma
files, .ogg files, etc.) or not compressed (mostly .wav files). In both
cases, waves are numeric samples supposed to be played at a certain speed
(like on a CD).  their precision will depend on the precision of the clock
inside your wave player. In the case of compressed files, you have another
problem with the quality of the file itself. according to the compression
rate, the compression will more or less reshape your wave, in a way which is
not related to your wave period. so the risk, is not only that you add
harmonics to your wave, but that you add a slight modulation, which would
add spurious frequencies around you center frequency. I have never tried to
measure this so I don't know in which measure it can cause a problem for
tuning.

You can record you own waves, but it would probably more precise to build it
by calculation (and optionnaly to chech it with you tuning fork). I built my
own set of wave files according to the inharmonicity I measured on my piano.
I put it on my MP3 player and tested my A440 against my fork, which shown a
1/2 Hz beat, which make an error around 2 cents. By the way, I found quite
difficult to tune from this source, but since I'm no experienced tuner, this
doesn't mean much.

Another possibility, with a computer, is to use it's synthe, by piloting it
with a midi file wich is HUGELY ;-)  smaller than a wave. note that if your
computer has a software synthe, you will get the same quality than with
waves (with still the advantage of the smaller file. In this case, it might
happen that you use another clock which could be more precise (this is higly
hypothetic).

The best option, would probably be to make a program which would measure the
speed at which samples are eaten by the computers sound device, compare it
with the system clock which is probably more precisely adjusted (a drift of
5 sec/day is equivalent to a 0.1 cents error), and automagically compensate
the soundboard error by adjusting the waveform . Such a program probably
needs to be written.

Philippe Errembault

----- Original Message -----
From: "Cy Shuster" <cy at shusterpiano.com>
To: "Pianotech List" <pianotech at ptg.org>
Sent: Monday, May 15, 2006 1:22 AM
Subject: Re: online pitch source


> Might not an .MP3 audio file of an A440 tone be independent of network
speed
> or computer hardware?
>
> --Cy--
>
>
>