I have noticed that if I use an electronic tuning app with inharmonicity compensation, such as Pianometer, I can tune one string of three to the desired pitch, and then the other two strings to the tuned string. The unison can then sound perfect, but the entire group now registers as very slightly flat in the app.
This happens almost always, but since I'm tuning basically by ear once I have the first string tuned, it comes out okay in the end. It just seems strange and I don't know why this happens. Any thoughts?

Sometimes, very rarely, I also come across something like this. First, it is necessary to check whether the strings are stuck to the shteg and the agraph. Does the tension of the non-working and working parts of the strings equalize when the tension changes? Do the loops of the string on the wirbel(pin) fit well together?
It happens that when you pull the string and it "sticks" to the pins on the rod, then after a while the tension of the string moves slightly to the non-working part and the formation creeps away a little. To avoid this, you can "massage" the strings around the pins on the rod. (By the way! Are there any cracks around the pins on the rod?)
Also, sometimes for many notes, the string wire goes to two positions in the choir at once, through one pin in a cast-iron frame. And when the formation changes, very rarely, but the string can creep a little from one position to another(This is more likely to be encountered when a new string is being installed.).
Also, in not very tight-fitting loops on the wirbel(pin), tension may be released a little and the system creeps away.
Such problems are most likely to occur with tools that have not been used for a long time and/ or have been standing for a long time without tuning.
I will assume that the wirbels hold the formation well and they do not scroll back after pulling the string. (I hope the tool is working properly in this regard).
If this occurs in a serviceable, good instrument, then I assume that the reason is as follows (1st option): you pull one string, stress is added to the structure and the structure is slightly deformed compensating for the stress. You add more tension with the help of the second string and the structure is slightly deformed even more, respectively, this may lead to some weakening of the first string. Perhaps, with auditory tuning, this is simply compensated automatically, but with the device it can suddenly become noticeable.
The 2nd option may be due to the fact that if you tuned the unison very cleanly, then the three strings may begin to vibrate as a single body. Accordingly, it will be 3 times more massive than the strings individually. And if the body is more massive, then at the same tension it will oscillate with a lower frequency. (this is just my guess and maybe I'm wrong, I don't know).
So I also come across this and just pull up more, so that the height of the unison corresponds to the target setting. Well, an auditory check using intervals is mandatory in such cases.
The author of the application recommends making several tool settings. Especially if a lot of time has passed since the previous tuning. Because the piano is such a system that if you change the tension of one string, then somewhere in another place something will shift somewhere...

upd.. No. 2 wrote incorrectly.. not the summation of masses.. eh.. rather, the synchronous vibration of three strings will move the final node of vibrations a little further, which will lead to a slight decrease in the frequency of vibrations, relative to when one string vibrates, and others are muted.

    Scherbakov_al
    Thank you for the very thorough explanation. I cannot say that I understand all of what you wrote, but my initial sense was along the lines of scenario #2. In any case, I do raise slightly the initial baseline pitch of the first string tuned, and by the time I have finished with the unison, it's exactly where I intended it to be. As an amateur tuner, I would previously tune my 1970 Yamaha G2 every two months or so, but just recently having replaced it with a 2007 C2 in very good shape (no visible cracks around the pins), I mostly just fix unisons in between having a professional technician do the work every few months.

    If you're measuring within under a cent, this is a real effect that happens when more than 1 string vibrate together. If all three strings are individually tuned exactly to 0.0 cents, then when you play them all together, the unison might be some fraction of a cent flat. Here's a link that talks about this "unison drift". It's not enough to matter in the grand scheme of things, but when I'm tuning, I choose to aim for between 0 and 1 cent as an acceptable range, instead of trying to get the note between -0.5 and 0.5. So on PianoMeter, the strobe rings are either fully stopped or turning very slowly clockwise when I call it "close enough" and move on to the next string.

      Maybe your unisons aren't as nice as you think they are.

        AWilley this is a real effect that happens when more than 1 string vibrate together. If all three strings are individually tuned exactly to 0.0 cents, then when you play them all together, the unison might be some fraction of a cent flat. Here's a link that talks about this "unison drift". It's not enough to matter in the grand scheme of things,

        ^^^^ This ^^^^

        It's a science thing, not a sign of poor tuning skills. 🙂 I think most tuners have to learn about it at some point, especially when tuning with an electronic device that can very accurately display the frequencies of the notes being tuned.

        pseudonym58 Beyond them having no discernible beats, I wouldn't be able to tell you.

        I have learned that tuning software like PianoMeter and TuneLab are pretty good at identifying partials 4 or sometimes 5 octaves above the fundamental - and that's a region where most tuners don't listen to beats anymore. This is when the described effect happens.

        The other explanation might be a single string within a unison that simple has a different composition of overtones, maybe because the wire has a different dimension or different contact areas at the bridge pins - and this is where software can drive you crazy, because you'll never be able to find a remedy for unison that per se is unclean.

          clavierhaus I have learned that tuning software like PianoMeter and TuneLab are pretty good at identifying partials 4 or sometimes 5 octaves above the fundamental - and that's a region where most tuners don't listen to beats anymore.

          Neither PianoMeter nor TuneLab tune partials more than about 3 octaves above the fundamental. PianoMeter uses partials 4 through 8 for the lowest six bass notes on the piano, and I think TuneLab uses the 6th partial for those notes by default. TuneLab can be configured to use up to the 9th partial, which is 3 octaves plus a major second above the fundamental. In my opinion, there are few reasons to pay attention to partials above #10. For tuning anyway. Voicing is a different matter.

            AWilley In my opinion, there are few reasons to pay attention to partials above #10. For tuning anyway. Voicing is a different matter.

            In my opinion that's where really beautiful tuning starts. And voicing should only be done on a piano with really perfect and beautiful tuning.

            It's good to know that I was wrong about the range of partials used by the two pieces of software that I use on a regular basis. My intuition served me well so far when it comes to tuning unisons, because I only tune these by ear. And I listen to those really high partials, especially major thirds.

            The software is my hearing aid for a temperament and stretch in pianos with different inharmonicity.