Thank you for your advice, That Guy! It appears to have worked. I found an "uninstall" function in my phone settings panel, and uninstalled EPT trial/demo version.

Then I powered down, removed the phone battery, replaced, powered up, and went to the EPT website and clicked the link to Google Play Store. I installed from there.

After some tense moments of no visible response, the phone finally started downloading something. At first, it showed on my phone as the Trial/demo/Free version. I went to the pull down menu in the upper left and clicked the upgrade button again to see what level of upgrade I was now being offered, and it still offered me the same upgrade to Plus that I had already paid for. But, a minute later, I got one of those translucent messages that said "Plus activated".

Now EPT shows I have the Plus version.

Thanks again, for your very helpful reply.

Yeah! Glad it worked!!

Ralphiano,
I'm not sure what went wrong with the purchase. Sometimes the Google Play Store is weird. Sometimes it takes several minutes or longer for the payment to "clear" and pass risk checks, particularly when you don't have a credit card linked to your Google account. (I checked the details of your purchase from the receipt info you emailed me and there was a 5 minute delay, but I don't know if that's enough to explain what you experienced.) There could also have been a delay or error with other checks the app does before unlocking. Often the problem is resolved by waiting and trying to unlock again, although uninstalling and reinstalling sounds like a good idea as well. Honestly I'm not an expert in this area, and most of what I think I know is second hand from other users' feedback.

I want to take the opportunity to express my gratitude for the promptness of your response to my problem. I sent my request via your website's "Contact Us" button late on the evening on Sept 11, and had your email response in my inbox early the following morning. I really appreciate that, and, it makes me feel good about using your product. I hope this EPT endeavor goes very, very well for you.

It was thought .. in connection with the addition of the adjustment of the weight of nonharmonicity ..

I can probably guess how the calculation of the tool building works in this tuner. An attempt is made to produce as small a discrepancy as possible between certain types of overtones, taking into account their amplitude. This occurs over the entire range of the instrument. If the instrument does not have very high-quality tenor strings, this results in an increase in unevenness in the tuning. If these irregularities are distributed over the entire range, then the temper zone also loses its purity. But with auditory tuning, the temperament zone is very clean and it does not depend on the unevenness in the tenor or bass. (with auditory tuning, the central octave sets the basis of the system).
When changing the adjustment of the weight of disharmony, I noticed that the adjusting curve in the zone of temperament is also slightly modified. Accordingly, we can get more purity as a whole, but a little lose in accuracy the intervals of the zone of temperament. Is it important? Still, the zone of temperament is in the zone of maximum sensitivity of hearing. Also in this zone the strings are fairly even even in the most substandard instruments. Classical setting implies the ideal in this zone. And I thought that maybe the tuner could calculate the system creating more changes around the edges of the band and try to achieve perfection in the center(Creating a classical form of beats in the intervals in the zone of temperament)(The central zone does not react to the features of the edges.).
Those. if I choose the adjustment of the weight of the inharmoniousness in 0.95, it would not create differences from the setting of 0.25 in the center. But the bass at the same time changed. Maybe this setting also had less effect on the treble.
Of course, I absolutely do not know what's going on under the hood of the tuner .. but such thoughts ..

I definitely like the sounding of the intervals at the transition from double twisted to triple unworn strings when tuning using the 0.95 inharmonicity compensation value.

Alex,

In what way did you find it different?

Ian

Just a question...

to measure the inarmonicty in EPT, when you say just play all notes of the piano througt. .Do you mute the unisons, so that sounds only one string per key.?
I think its must be obvious, but maybe not.
Can EPT measure inharmonicity with the 2 or 3 strings of the unison singing at the same time??

Thanks

Pablo

Yes, I always mute off the 2 outside strings and in the bass section mute off one of the doubles. I'm sure EPT measures some kind of inharmonicity with all the strings sounding but I would think you'd get a more accurate reading with only one. Also, keep in mind that you don't have to play every note, I just choose to do that so I have a full "picture" of where the piano is currently at.

This is normally done with wavelets, which is easily google-able. An FFT integrates time from negative infinity to positive infinity, whereas the wavelets break up time in an ascending series of chunks. So, a long(er) time frame for low frequencies, because you need more time to get enough information to determine the frequency, and then progressively shorter time frames for higher frequencies.

The point of this is to have better time information about when frequency appear. If I sample 1 second of a song/piece with FFT, I can tell you what frequencies are present and what amplitude, but can't tell you, for example, that there is a staccato piccolo playing only in the first half second - I just see high amplitude, high frequency components. With wavelets I may look at the bass with a 1 second window, but higher notes with a 1/128 second window (the implication is I use windows of 1, 1/2, 1/4, 1/8, 1/16, etc). This gives me a high resolution look into the timing of the high frequencies.

Is that useful for a tuning app? I dunno. You are adding time information, not new frequency information, and at the cost of more processing.

So I am very new to this - I bought a hammer last night, along with the app, and tuned my piano. It went okay, as in it sounds a lot better now.

However, I have no idea to what precision I should be aiming. Before tuning I had around -2 to +8 cents off across the keyboard with a few notes with pretty mismatched unisons. Now it is all well under 1. But what is your threshold for 'good'. Each time I hit the note (single string) not only does the needle dance around, but it also isn't repeatable. As in one time I may get it bouncing from .1 to .4, the next time it is -.2 to .3, and so on. Also, my pin turning technique is nonexistent, so once I got it down to around .3 to .4 or I was pretty happy. But as a pro what are you aiming for? Is it reasonable for a pro to see 0.0 for every string? I want to be as accurate as possible, but I don't want to sit there and endlessly adjust pins trying to get 0.0 if it is an impossible standard to meet. More wear and tear, more chances of an accident. I just aimed for it to bounce no more than around .3 or .4, always by dropping down from slightly above (1cent or so), and to trend just slightly positive vs some strings positive, some negative.

Roger, I doubt that any pro would be able to manage 0.0 for every string. I'm sure that everybody's standards are different, but if I can get everything within a <1 cent window I'm usually pretty happy. My target window is probably from about -0.2 to +0.3 cents. (If I have to choose between leaving it slightly below pitch or slightly above, I choose slightly above for consistency.) Also I recommend not paying as much attention to the numbers and the dial and focusing more on the spinning strobes. They will give you faster and more consistent feedback. (Also a reason I tend to err on one side of the pitch...I think you get better precision by keeping the wheel between stopped and turning slowly clockwise than by keeping it between slowly clockwise and slowly counter-clockwise.)

Re: Wavelets, that's actually something I looked into, but I stuck with the FFT because it was fast and simple. I'm actually taking the FFT more often, but I'm only drawing every half second or so to save on...well...drawing time. I agree it could be done much more elegantly, but I've had more pressing priorities like getting the dial and numbers to give faster feedback, fixing bugs, and adding features like custom temperaments. In any case that's a beautiful spectrum you posted above and it makes me want to redo mine :-)

Thanks Anthony. My ears don't seem to tell me there is a tuning problem at .2 or .3, but then I have tinnitus! I'll just make my ears happy in this range and call it good.

Pleasant update!

Thank you! I'm glad you like it. It's still just for Beta testers (good thing too, since there are a few bugs...) I pushed it out a bit prematurely mainly because of the improvement to the dial/number speed, but some of the other new stuff is still a work in progress.

I just finished a piano with the latest beta version and I really appreciated the improved response of the needle with the strobe rings. Very nice job on a small console!

I'm new to tuning, but found EPT very intuitive. I found it useful to get started fairly quickly, then as reassurance as I started to trust my ears.

I have not yet had the opportunity to evaluate the latest beta version.
You say there are some more mistakes. How do they appear, do they affect the result? How can they be avoided or compensated?

That's great! Thanks for creating the app by the way, I just bought it two days ago.

Current known bugs (in order of severity) are:
1. Starting/loading a new tuning file when the current file is in the "locked" state results in the new tuning file being "locked" as well. The danger is that while the app measures inharmonicity as normal, it won't recalculate the tuning curve. The workaround is to press the "lock tuning curve" button twice.
2. I'm looking into two reports of bugs affecting certain tablets (one where the automatic note detection isn't working, another where the strobe rings don't work in Landscape mode)
3. Creating a custom non-equal temperament causes the app to crash on certain devices.
I hope to have fixes for these go live by the end of the week.

Some other things I'm currently working on are:
Some new graphics for the pitch raise dialogue to help new users understand what is meant by terms like "lowest unwound/plainwire string"

And better note detection for the top and bottom octaves (don't have an ETA for that yet, but hopefully soon...I've got some code working in Matlab)