Nord User Forum

[robinlechols]

Has anyone tried to or changed the internal tuning of the Nord Piano 88? I'm interested in tuning it to A444 and playing around for a while.

Just curious...

[pablomastodon]

Hi Robin,

The Fine Tune function, found in the System Menu, will do that for you, sort of. You can vary the tuning of your NP +/- 50 cents. The scale is not graduated in Hz.

The way I figure that is:

On an A=440Hz scale, the higher Bb=466Hz. A cent is 1/100th of a semi-tone; +50 cents should raise things 1/4 tone. The frequency difference is 26Hz. This should then equate to 0.26Hz/cent. For simplicity let's round this off to 0.25. You wish to go to 444Hz, or a total of 4Hz higher. 4Hz divided by 0.25Hz/cent equals 16 cents. This is a rough approximation but will surely put you in the close neighborhood of A=444Hz.

Bless,

Pablo

Perhaps some math whiz here can either confirm or correct my calculation for you. It is late and my daily quota of higher level brain function has already been exhausted today....

Bless,

Pablo

[maurus]

The following page offers a range of useful calculators for note frequencies, tuning, etc.:
http://www.sengpielaudio.com/calculator-centsratio.htm

And yes, +16 cent is indeed the closest you can get to 444Hz.
Also try the opposite, lowering pitch a bit to 437Hz or so...

[robinlechols]

Hi Robin,

The Fine Tune function, found in the System Menu, will do that for you, sort of. You can vary the tuning of your NP +/- 50 cents. The scale is not graduated in Hz.

The way I figure that is:

On an A=440Hz scale, the higher Bb=466Hz. A cent is 1/100th of a semi-tone; +50 cents should raise things 1/4 tone. The frequency difference is 26Hz. This should then equate to 0.26Hz/cent. For simplicity let's round this off to 0.25. You wish to go to 444Hz, or a total of 4Hz higher. 4Hz divided by 0.25Hz/cent equals 16 cents. This is a rough approximation but will surely put you in the close neighborhood of A=444Hz.

Bless,

Pablo

Perhaps some math whiz here can either confirm or correct my calculation for you. It is late and my quota of daily higher level brain functions has already been exhausted today....

Bless,

Pablo

Thank you for your math and wit. I burst out laughing at the quota comment.

[Mr_-G-]

Actually if you use a more precise value for Bb (466.164), the closest setting to 444Hz is 15 cents.
Not that it might noticeable, but we want the best tunning on the best instruments!

Edit: this is incorrect see maurus' explanation below.

[maurus]

Not according to the calculator I linked to above (which I think is quite accurate):
440Hz + 15ct = 443.82887286777583...Hz
440Hz + 16ct = 444.08531253322485...Hz

[Mr_-G-]

Hm... Not sure where the difference comes from. Here is my calculation using a spreadsheet:
Bb frequency should be 440*(2^(1/12)) = 466.163762...
so one cent of the semitone between A and Bb should be:
(440 * (2^(1/12)) - 440) / 100 = 0.2616376152 Hz = 1ct

15ct=3.92456423... (that is -0.0754357723 Hz from 4)
16ct=4.18620184... (that is 0.1862018429 Hz from 4)

So according to the above, 15ct is closer to 444 thant 16ct.

Edit: this is incorrect see maurus' explanation below.

[Mr_-G-]

I think the problem with your calculation is that you divide the semitone in 100 linear (additive) steps, whereas the cent division is in fact defined as a ratio of frequencies (just as the semitones themselves are ratios of frequencies, not differences). So, one cent is in fact the following ratio between frequencies (the number is taken from the site I linked to):

1 Cent = 2^(1/1200) = 1.0005777895065548592967925757932...

Given a note of a certain frequency f, to increase pitch by 1 cent means to *multiply* f by the number above. Of course the linear division in a semitone (which you calculated) is a pretty close approximation to the proportional division, so differences are not very large. And who knows what the DSP's in our instruments actually calculate...

But now we are again quickly reaching the daily quota etc.

[Mr_-G-]

Oh, I see what you mean. Yes, that site's formula states: c = 1200 × log2 (f2 / f1) which gives the correct result (16 cents).
Thank you for the correction!

[pablomastodon]

Somehow I knew that this thread would generate much comment. If only MJBrands were still here, we'd be on page three by now.

Thanks for the kind words, Robin. I do like to keep it entertaining when possible...

Bless,

Pablo