Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum Using Easy Piano Tuner

Thank you, Anthony! I understood that.

This is the graph of a Bechstein upright I tuned the other day for music exams. It's about ten years old, and it always seems to me that the tonal concept is similar to the century-old Bechstein pianos that abound here. It's a very nice piano. My tablet microphone seems to struggle at the top, in this case especially with A7 and A#7. I've wondered about using an external mic, which I think has been touched on before.[img]http://forum.pianoworld.com//ubbthreads.php/galleries/2858448.html#Post2858448[/img]

Fantastically smooth!

It's a very nice piano to tune.

I'm trying this, to see if I can get the photo to show here...

David,
I have not found a way to disable the internal mic of my Galaxy tablet.
Ian

I find the inharmonicity graph more valuable. When you notice big jumps, that's an opportunity to remeasure to make sure the jump on the screen reflects the piano, and isn't just a measurement anomaly.

Often it can lead to an "aha", that's why those notes sound like that...

Ron Koval

I'm not an expert on mics. The highest quality mic I own is a $50 "Blue Snowball ICE" USB mic which, from what I've read, is kind of a starter mic for YouTubers and Skype calls. It's a clear step up from the internal mics on phones/tablets etc, but nothing special.
Maybe this week I'll play around with that and see if I can get it to work with my Chromebook (which can run PianoMeter).

That said, I know it is possible, at least on my Pixel 1, to have PianoMeter use the audio from the 3.5mm TRRS [tip-ring-ring-sleeve] jack, because I've done it. Granted I was using audio out of my computer and using RCA cables and alligator clips to send the signal to the correct sleeve on the jack for the mic input. Plugging the regular 3.5mm TRS jack from your mic into your phone won't work because the phone jack isn't wired to receive that. It's built for wired headsets that have earphones and a mic. So at the minimum you'd need some sort of adapter/splitter like this plugged into your phone (assuming you have an external mic with a TRS output).

I've also had success using the splitter pictured below and that worked, but it was a bit buggy because it occasionally triggers the Google Assistant on my phone, which I think would normally be done by pressing the little button that some headsets have. I suspect it has something to do with the open connections on the other side of the splitter. So in summary, yes, I think an external mic should work, but only if it can be connected properly to your device.

I bought this mic from Best Buy and it seems to work fine: AmpRidge
It plugs right into the headphone jack. I'm using it on a Samsung Galaxy Tab A model SM-T350. From what Anthony says, it shouldn't work but it sure seems like I picks things up better with the mic.

Well I've certainly been found to be wrong before :-) I don't know if there's anything special about the little connector cable for that AmpRidge mic, but the product page says it's compatible with "smartphones" so I guess it's no surprise that it works. I'm a bit out of my depth when it comes to the smartphone jack that is able to take cables with 3 or 4 contacts and be able to recognize the difference between a mic, a headset with a mic, and regular headphones.

No smartphone jack can an does actually notice that difference. A regular 3.5mm jack has ground, stereo output and monaural input from Microphones integrated in a headset. Ampridge only has monaural microphones.

There may be at one point a compatibility with mic preamps outputting USB-C but until then you're pretty much stuck with what's in a smartphone or an external monaural microphone.

Actually I am really happy with what the OnePlus 5 records with its internal stereo microphones, specifically with piano. This is unprocessed audio along with the video from a OP5:

https://www.youtube.com/watch?v=-9htzzbAHKE

This is also the smartphone I use to tune my grand with PianoMeter and I am happy with the result.

I just noticed something in an article in the MTNA Journal (I'm a pianist and teacher) that might help folks looking for a microphone to use with a smartphone and Easy Piano Tuner. "External microphones, such as the Rode VideoMic Me or the Movo VXR10, can attach to the smarphone headphone jack... significantly increase sound clarity...".

From the specs and price points, the Movo looks to be the better option, more extended frequency response, less expensive.

Might be worth a look.

David, I dare you to tune that piano entirely by ear and then use Piano Meter to record and demonstrate the results.

Sorry, I don't accept the challenge! I'm not an aural tuner, by and large.

Out of curiosity, I measured the frequency response of the built-in microphone. I used the REW program and pre-calibrated monitor speakers GENELEC+GLM calibration. This measurement does not purport to be accurate, but is merely an introductory example.



You may notice some frequency irregularity. Different phone models may have different dips and peaks depending on the design features of the sensor (microphone), body, microphone hole, software. How much such unevenness can affect the accuracy of the setting?

Some more point about the dark theme - when using a dark theme, I can not see the tuning curve and the points around it.

That was intentional. In addition to being easy on the eyes, I want Dark Mode to also be a power saving mode. It saves significant power on devices with AMOLED screens and it takes less CPU time to draw solid black instead of the patterned wood background. I chose to have dark mode turn off the graphs because 1: the graphs also take up a significant portion of the drawing/CPU time, and 2: they aren't essential for the bulk of the tuning. (I figured users could turn the graph on when they want to check the tuning/inharmonicity curves.)

This is something I would appreciate some user feedback on: Do you prefer to have the graphs to still show up in dark/power-saving mode, even if that means extra drain on the battery?


I've edited your image to highlight the frequency ranges that PianoMeter uses.

From approximately 107 Hz to 1,880 Hz it is looking at the relative strength of 2 to 5 different harmonics. From 1,880 Hz to approximately 4290 Hz we are only looking at the first (fundamental) harmonic, so the mic response doesn't matter in that region. We don't measure any frequencies below 107 Hz or above 4300 Hz, so the big mic dropoff at lower frequencies doesn't have any effect on the tuning. The dips and peaks in the multi-partial zone will have some effect on the calculation of the tuning curve because we take into account the strength of the harmonics to decide which intervals get more "weight". However, we also do some pre-smoothing of the harmonic amplitudes before calculating the tuning that (I think) should help mitigate this effect.

There could also be a small effect on the position of the needle while tuning since the cents offset is calculated from multiple harmonics and then averaged together in a "weighted average" based on the harmonic strength. But this effect won't be large since the frequency targets are closely matched to the inharmonicity of the string (so it doesn't matter very much which harmonic is being measured).

I don't know of a way to easily correct for the irregularities in the responses of different microphones, although to be fair, similar artifacts exist in the frequency response of our own ears :-)

Thanks for the explanation of the frequency chart. In the dark mode, it seemed that the missing graphics is a mistake. But if this is done to save energy, then I do not mind that it would be so. Let it be maximum efficiency.

Compared to my Neumann KM184 (matched pair) this frequency response is very erratic

Neumann KM184
Ian

I gradually notice some features related to tuning with the tuner. I find the difference from the auditory tuning in a different spatial sounding of the instrument. Good ear tuning creates a very pleasant enveloping spatial sound picture. Even listening from afar, the piano can (and should) sound voluminous. After tuning with the tuner, the space is as if folded into a plane. Some sounds in the duration of the sound seem to be eaten by others. Performance on such an instrument becomes more difficult. The instrument is becoming less melodic. I played with the parameter "coefficient of inharmony". As the value increases, the tool becomes, I would say, flatter. With a decrease, spatiality improves. More evenness appears in the chromatic sequences, but the purity of the intervals decreases. Perhaps high values ​​of the coefficient result in a too clean setting. The overtones, if possible, are close to each other and this leads to the "eating up" of the energy of sounding strings. Just like this happens in very well tuned unisons - they become glassy, ​​fast, dry and with a short attack. May be..
Perhaps the auditory tuning implies, sometimes intuitively, in addition to a certain ratio in overtones, also the "melodiousness" of the interval and its "spatiality". Physically, this can be described as the interaction of coupled oscillators. When the frequencies coincide, they fall into a strong resonance. Absorb each other's energy. With some frequency mismatch, they try to drag each other's phases. Elements of chaos appear.
Could it be possible to control spatiality through correlation and weighting of overtones?
Is the non-uniformity of the frequency response of the microphone related to the change in the weight of certain overtones in the calculation of the tuning and as a result of the appearance of a larger resonance of certain frequencies? The appearance of a buzz of some frequencies in the sound of chords?
Do you have confirmation of my words?
What are your thoughts on this?

If I understand correctly, you are saying that intervals, like unisons, should be pure, but not too pure? I think this makes sense to me, in the sense that exciting a wider range of frequencies can give a perception of bigger volume. I think I remember reading something about this in a physics of music class where they said that this is one reason we use Vibrato in music. I don't know if that's the correct explanation, but it makes more sense to me than the destrutive interference hypothesis.


These results make sense with what the coefficient of inharmonicity does: a low value gives a more even tuning by giving more weight to the smooth "best fit" inharmonicity curve, and a high value should give more pure individual intervals by giving more weight to the actual measured inharmonicity values for each note, accounting for more of the irregularities in the piano's scale. But I think the inharmonicity coefficient is not the correct tool for the job.


The frequency response of the microphone can affect the weight given to overtones, possibly causing some intervals/partials to be slightly more in-tune than others. I suspect the effect would be small due to the range of frequencies we're measuring and the multi-partial multi-interval approach to tuning each note, but I can't say for sure.


Maybe, and I'm currently working on an update that will allow users to freely modify the weights given to specific intervals. For example, you could significantly change the "stretch" in the bass by modifying the weight of the 5:10 octave. Increasing the weight given to fifths or fourths would increase or decrease the stretch of the entire tuning. But even with control over all the intervals I don't know if it would give you easy control over the cleanness of the tuning for one reason: PianoMeter will still be trying to tune all the intervals as pure as possible. Changing the weights will make some more pure than others, but the target is still minimizing the impurity.

Something that might do what you're looking for would be for me to add an addional control that adds or subtracts additonal "stretch" from the tuning after the pure tuning curve is calculated. So for example, you could choose to add 0.5 cents of additional stretch per octave in the Treble, after PianoMeter has calculated the best fit "pure" tuning...not enough stretch to make pure intervals sound bad, but maybe enough to broaden the range of frequencies excited.