Dear friends,

Please help me with defining partials for these specific instruments: SPINET,CONSOLE,VERTICAL,UPRIGHT, SMALLGRAND,STUDIO GRAND and CONCERT GRAND.

I´m using Tunelab Pro 2.0 on my laptop computer with settings: 6:3/4:1 A0-E2 =6
F2-E3 =4
F3-GIS4 =2
A4-C8 =1

From 2000. to 2005. I´ve been using Peterson 490 ST and I was not complitely satisfied with it. Mostly, I´ve been using Small Grand stretch table which was the best for the most instruments. A0-C8 =1

In the literature I found variety of solutions and need advice.

Verituner A0-B3 =4 CYBER TUNER A0-Gis2 =6 TLA CTS-5 A0-Gis1 =4
C4-B4 =2 A2-Gis4 =3 A1-Gis2 =2
C5-C8 =1 A4-Gis5 =2 A2-C8 =1
A5-C8 =1

There is more interesting solutions in the literature . A0-Gis1 =6 A0-Gis1 =12
A1-Gis4 =4 A1-Gis 2 =6
A4-Gis5 =2 A2-Gis4 =3
A5-C8 =1 A4-Gis5 =2
A5-C8 =1

I didn't find any data about Yamaha PT 100 mk2 i ACCU TUNER III

Thanks...

Ivan Mihajlovic
Stanka Hondla 10
43000 Bjelovar
CROATIA
e-mail ivan.mihajlovic@bj.t-com.hr

With TuneLab, you can change the partial number without changing the resulting tuning.

So if you want beatless 6:3 octaves in the bass, and a particular note doesn't have a strong 6th partial, you can switch to the 8th or 10th, and TuneLab will recalculate the target frequency on the fly for you.

In fact, TuneLab will automatically find the loudest partial in the bass. In the options, find "Automatic partial selection up to", and change it from the default "A0" (which turns it off) to "E2". It's a huge help, especially on short pianos where the 6th partial is missing on the bottom few notes.

If you're asking about which octave styles to use, start with the choice that makes the least deviation in the tuning curve from beatless. Pick the combination that makes the lower tuning curve graph the flattest.

Are you measuring inharmonicity on each piano before tuning?

--Cy--

From Cy
"With TuneLab, you can change the partial number without changing the resulting tuning."

Whoa there, take a look at this graph again from a NICE piano:



This graphs the actual measured partial information of each note. Notice how the 4th partial is the smoothest throughout most of this portion of the piano? Now imagine if you then chose another partial to force into a smooth curve. Wouldn't the tuning change?

It is sometimes easy to forget the assumptions made by the tuning software - this one being that the partials in the scale of a piano follow a predictable linear pattern.

So Ivan, I don't have an easy answer to your question. Usually, it's best to have the machine work from one of the strongest partials, like Cy wrote. You will then get the clearest display. Some people like to use partials that have direct comparison relationships to octaves higher and lower. Different choices can have the effect of emphasising certain test intervals. (smoother 3rds and 6ths vs. smoother 4ths and 5ths, for example.)

Ron, I'm talking about changing the partial you listen to on a single note, not recalculating the curve for a different partial.

If you want a beatless 6:3 octave, and it can't hear the 6th partial, when you switch to the 8th, TuneLab adjusts the target frequency in your display to compensate for the fact that the 8th partial has a higher offset from pure than the 6th, so that you don't have to redo the tuning curve. This is a unique feature of TuneLab, as far as I know.

TuneLab already interpolates from your inharmonicity measurements, so no further error is introduced by switching partials.

--Cy--

Ok, yes I get that.... so tell me how tunelab "knows" where to place the 8th partial? That is at the core of the assumptions. If you measure a few notes, can it then through the use of an inharmonicity constant REALLY predict the location of partials not only on those measured notes, but the notes inbetween? You can if the graph looks like this:


But not if it looks like the other graph. It is fortunate that the differences are small, and amazing that the assumptions work as well as they do!

Yup, it's an approximation, but it works pretty well.

Consider the SAT, where you only measure two partials on three notes to construct the entire tuning curve!

--Cy--

Hi!

First, thanks for your help.

1. Where in TuneLab 2.0 Pro is option "Automatic partial selection up to"?

2. What stretch chart you are usualy using for upright pianos (115 cm)?

Thanks,

Ivan

I don't use a stretch chart. I make a new tuning curve for each new piano. Here's how I do it:

1. Select File | New | Tuning File.

2. Measure A1, A2, A3, A4, A5, A6. Mute to a single string. Take three readings per note; discard any that are way off. Keep the microphone within a foot of the string in the treble. Don't move the microphone during a reading.

3. Edit the tuning curve. Start with 6:3 octaves in the bass, 4:2 octaves in the treble. Click on the arrows so that the lower graph ("deviation curve") is as close to horizontal as possible at the outer edges (A0 and C8). Different octave styles may give better results. Save the file if you wish.

4. Tune from A0 to C8. Mute the left string of bichords, tune the right, then move the mute to the next bichord and tune the left (or do it aurally). Use two mutes for trichords. Mute the outer strings and tune the center. Move the right mute to mute the right string of the next trichord, and tune right to center aurally. Move the left mute to mute the left string of the next trichord, and tune left to center aurally.

5. Check all the unisons one more time, to listen to all three strings.

--Cy--