From: Patent Tactics, George Brock-Nannestad
when I made my speed calibration record in 1982 I was quite concerned with
precision and stability of the oscillator. The precision was ensured by
having a counter (with a crystal oven), and the stability by letting the HP
Wien bridge oscillator run for 3 hours in the recording room before starting
work. The RC (typically Wien bridge) oscillators were much less accurate than
the contemporary BFOs (beat frequency oscillator). The latter were used for
sweeping the a larger audio range, for instance when using a level recorder.
I have information from the General Radio Corporation and Brüel & Kjaer in
1965: the GR Type 1304-B BFO had an accuracy of ±1% +0.5 Hz from dial after
zeroing. Stability:drift from a cold start is less then 7 c/s in the first
hour and is essentially completed within two hours.
The comparable Brüel & Kjaer Type 1017 the accuracy is 1% ±0.1 Hz
The GR Type 1210-C is typical with ±3% accuracy, but a warmup drift of less
than 1%, again stability is obtained after 1 to 2 hours.
The 11-fixed-frequency GR Type 1311-A Audio Oscillator was easier to tame,
because it used calibrated fixed components, and accuracy was ±1%, and the
All of this is analog, of course.
In-board calibrators for recording equipment could be much less reliable.
> One comment on that -- it could be anything from a HP unit circa 1950
> onward. For instance if you're
> hearing this between 1K and 10K, that's because most oscillators had a
> switch for 1x, 10x, 100x of a
> frequency that was dialed in on the rotary control. Function generators were
> typically push-button
> and 1x, 10x, 100x was standard. Recording consoles and tape decks like the
> Technics RS-1520 or the
> Otari MX5050 (at least the 4track version I have) had switches for 1K, 10K
> tones (rotary on a
> Technics, pushbutton on an Otari). I've seen recording consoles with
> pushbuttons for 100, 1K and 10K
> and other tones. Older-school studios just used an external oscillator
> plugged into a slate input.
> Basically, the way I've always operated was never to set speed/pitch from
> slate tones. If you're
> lucky they'll give you an accurate 0-VU level and that's about it. Again, if
> you're lucky, you can
> use a scope and set azimuth from them (that's assuming they were recorded
> just before the program
> content was recorded, on the same deck, and the deck was azimuth-aligned
> differently between tones
> and program content -- not always the case).
> -- Tom Fine
> ----- Original Message -----
> From: "Jim Sam" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Thursday, January 27, 2011 1:58 PM
> Subject: Re: [ARSCLIST] Test tones circa 1978
> I should add that I'm interested in this due to tape that was recorded
> at 7.5 ips, 250 nWb/m. The tone sequence is not standard, with tones
> changing abruptly as if a button was pressed, rather than a dial
> On Thu, Jan 27, 2011 at 10:40 AM, Jim Sam <[log in to unmask]> wrote:
> > All,
> > Does anyone know how much test tone generators varied circa 1978?
> > I know not to expect digital accuracy. For example, I would not
> > expect a 10.00 kHz tone, but how much leeway was there? +/- 0.25kHz,
> > +/-0.5 kHz, +/-1.0 kHz, etc.?
> > Thanks,
> > Jim