Hi again David. I'll just do a real quick version of this as I am preparing an article on the subject that will explain my present approach much better. I do use Tunelab yes... but I think you can accomplish basically the same thing with Sat... and Cypertunner for that matter... tho Tunelab makes it real easy to do this kind of thing. I rely on a kind of template default P-12ths priority. In Pocket Tunelab I edit the partials table so that all notes from A1 to E6 are set to the 3rd partial. From F6 up tunelab does not allow the 3rd partial...so I let those notes stay on the fundemental. Below A1 the table has the 6th partial... but I never use these directly. I sample A2, A3, A4, & A5 and D2, D3, D4 & D5 and generate the curve. The curve uses the 3:1 setting for both the bass and treble regions. Its important to remember this is just a template for the sampled instrument. From D3 to G#4 I do pretty much tune exactly what the machine says... but all notes outside this are <<suspect>>. That said... essentially I tune in chromatic order both the above named region along with octaves below and 12ths above to the note in that same D3 to G#4 range. Remembering that at all times I am looking at the third partial for this range this means that the octave below defaults to a perfect 6:3 octave and the 12th above defaults to a perfect 3:1 12th. This approach includes the pianos inherent inharmonicity in all respects because the curve is generated after sampling to begin with and secondly I am direct referencing all intervals. For example D3's 3rd is set to a frequency of 440. Tuning A4 and D2 to this same yields the above mentioned relationships. To refine and check other relationships as I go I use Tunelabes partial switcher. For example again on D3... I can switch to the second partial.. confirm a motionless phase display and make sure the 4:2 octave is not unacceptable... and at the same time I can take a look at the 5th BELOW D3 as its 3rd partial is coincident to D3's 2nd at the fundemental of D4. While I dont look so closely at this relationship at this particular area of the tuning... I do exactly that much lower which allows me to make sure that all 4:2 and 6:3 octaves along with closely related 3:1 12ths are all <<compatible>>. In the treble I default at the outset to pure 3:1 12ths direct referenced as described above. Tuneing A4 to the D3 display yields both notes relevant partials at exactly the same frequency. I like to switch the display up an octave and then up a second octave and see how the resultant A4 is going to work with notes yet to be tuned. Switching D4 on the display and playing A4 shows how A4 is going to work as a 5th with D4 (when that gets tuned). Switching the display D4 gives me a look at how it will work with the fourth above... and as I progress I can use this same to get a good idea of how notes being tuned function as octaves and double octaves as well. It all adds up to using a template based on P-12ths utilizing the 3rd partial all the way along with direct referencing of notes to accomplish a kind of visual multi-partials tuning. All the way I make sure that the 3:1, 4:2, 6:3 and in the higher regions the 2:1 are all compatible with each other visually... and I finish with various aural checks to confirm this compatibility. Remembering that all this is based on an old aural test for double octaves.. which essentially leans in the same direction... it should become clear with a bit of practice what this is about and why it works so well. That aural test simply takes a major third below any given note as the control note and compares both the major third, the 10th and the 17th. Throwing in the major 6th above the control note allows a test for the P12th by comparing that with the 17th. Morphing all these together for the best fit is either done aurally by listening to beat rates as best one can... or can be done by direct referencing the relevant frequency of the coincident partial in question. Either way... finishing with a run through of listening closely to the overall combined sound... a kind of Virgil Smith approach to confirm all is well.... makes for a very fine tuning. It may sound time consuming... but with a bit of practice this goes fast. I typically use 70 minutes tops for a complete concert tuning.. 90 for a minor pitch change. Cheers RicB
This PTG archive page provided courtesy of Moy Piano Service, LLC