 |
Welcome to the Piano World Piano Forums Over 3 million posts about pianos, digital pianos, and all types of keyboard instruments. Over 100,000 members from around the world.
Join the World's Largest Community of Piano Lovers
(it's free)
It's Fun to Play the Piano ... Please Pass It On!
|
|
41 members (Dariusz D, axomas, CraiginNZ, Damien PG, Andrew E., Ciferz, butchkoch, daz100, Catlady, BlackKnight, 6 invisible),
809
guests, and
231
robots. |
Key:
Admin,
Global Mod,
Mod
|
|
|
Joined: Apr 2020
Posts: 24
Full Member
|
Full Member
Joined: Apr 2020
Posts: 24 |
I forgot to mute the strings in the initial recording. Too late now. It sounds bad, can't go back. I've tried learning to tune by ear, but I think there is too much insider secrets. Everything online isn't really good material or they want to charge you a bunch of money to take a class. I wish I hadn't done anything now.
|
|
|
|
Joined: Aug 2008
Posts: 1,310
1000 Post Club Member
|
1000 Post Club Member
Joined: Aug 2008
Posts: 1,310 |
I mean you already went through the trouble of tuning it once, what's a second tuning? :P I bought a better tuning lever and it makes the job easier so that I dread it much less. Still, 232 strings to tune is a lot, so I'm trying to finish my algorithm refinements before I go do another tuning. I struggled for an entire week trying to get the thing to compile and only just got it to work.
I tried touching up the tuning a little a few weeks ago but a bunch of unisons have already gone out, so clearly my technique with the lever is deficient and this will probably be an ongoing thing, but I figure it'll be faster in the future with a tuning file I like saved.
For me, this software is an absolute necessity, because I'm too tone deaf to tune by ear, and it seems like too much voodoo magic when I see how they determine the temperament. The EPT theory is pretty elegant and is a sensible extension of how aural tuning is done, so I am pretty happy to go with it.
Last edited by trigalg693; 06/18/20 01:52 AM.
|
|
|
|
Joined: Oct 2019
Posts: 110
Full Member
|
Full Member
Joined: Oct 2019
Posts: 110 |
Too late now. It sounds bad, can't go back. Why? You can do the measuring again. But think of this: 1. Entropy-PT ist very ugly. 2. A (good) Software ist not enough. The first tunings sound bad - always. You need patience an many tries. If you only want to tune your piano 1 or 2 times per year - let it been done from a technician.
excuse my bad english, I'm not native. Corrections are always welcome!
|
|
|
|
Joined: Oct 2019
Posts: 110
Full Member
|
Full Member
Joined: Oct 2019
Posts: 110 |
New try in correct English:
If you only want your piano to be tuned 1 or 2 times per year - it's better done by a technician.
excuse my bad english, I'm not native. Corrections are always welcome!
|
|
|
|
Joined: Aug 2008
Posts: 1,310
1000 Post Club Member
|
1000 Post Club Member
Joined: Aug 2008
Posts: 1,310 |
By the way ThePenist, you can reset your tuning to the recorded one if you saved the tuning file, the pitches are saved and can be viewed with the "Graph" button.
I'm collecting some data right now, been running the "stock" EPT algorithm over and over with my tuning file. My crappy rough tuning got the entropy to 6.489, EPT usually converges on around 6.399 at the infinite setting, and if I mess around with it I can get lucky and it'll go to 6.396 ish. So that's the number to beat.
|
|
|
|
Joined: Aug 2008
Posts: 1,310
1000 Post Club Member
|
1000 Post Club Member
Joined: Aug 2008
Posts: 1,310 |
Okay after doing a bunch of testing with different modifications to the algorithm, I am pretty convinced that the EPT objective function is not good. I wrote a blog post about it for the slightly technically inclined: https://www.quora.com/q/gjwwwxgohpqdfldt/Entropy-Piano-Tuner-UpgradesFor a rough tuning, its inharmonicity estimate for the starting tuning curve is not bad (basically makes octaves in tune), and I like the software aids for doing the actual physical tuning, but without hearing the piano I'm already not happy with the tuning it's producing. With my non-zero temperature single-note-change-only method it's getting a lower entropy but tons of bass strings being sharp is just ludicrous. I want to fix this without resorting to further restricting the allowed ranges of adjustment, I'll have an update tomorrow if I can make it work with an alternative frequency weighting.
Last edited by trigalg693; 06/23/20 07:28 AM.
|
|
|
|
Joined: Jan 2013
Posts: 624
500 Post Club Member
|
500 Post Club Member
Joined: Jan 2013
Posts: 624 |
As usual I’m late to the conversation. I was an active tuner/technician from 1984 to about 2001. Varying from full time (10-15 per week) to part time (2-3/week). And I have kept slightly in shape by occasional tuning friends’ and relatives’ instruments since leaving the profession. I would not consider myself an expert tuner.
I was trained to tune by ear but in the early days I had trouble setting a great temperament. But I was fine with octaves, stretch and unisons. So after a couple of years I bought a Sanderson accu-tuner. It always seemed too cumbersome and time wasting to me so I began using it just to help with my temperaments. Eventually I got pretty good at aural temperaments and pretty much abandoned the SAT. (Still have it somewhere).
But there are two cases where a machine can be critical.
First is the home where you arrive and the residents decide it’s time to eat and you’re battling against the clanking dishes. Or homeowner decides it’s time to run the vacuum cleaner or, worst of all, the dishwasher. These are conditions where I found it impossible to hear anything. So out came the SAT and I could rely on the strobes to know where the note was.
The other was for pianos with lots of false beats or top octaves that just were beyond my capability to hear. The machine could. These were usually spinets with cruddy or rusting strings and I really didn’t want to break one.
In these cases I might have been able to provide a superior aural tuning completely by ear but my philosophy is to provide the best service possible given the constraints of time, budget, owner’s expectations, and condition of the instruments.
I still do not like using machines but there are certainly times where the machine is the best choice.
-Bill
-Bill L. - former tuner-technician
|
|
|
Haessler
by PianistEsq - 07/06/22 12:05 AM
|
|
|
|
|
|
Forums43
Topics213,823
Posts3,205,776
Members105,732
|
Most Online15,252 Mar 21st, 2010
|
|
|
|
|
|