Measurement mic for Piano Tuning Software?

Much disputed! The engineers and piano tuner/designers of "TuneLab", "Verituner," "Sanderson Accutuner" etc etc--all high end piano-tuning programs tell you, basically, NOT to spend money on a measurement microphone.

Why? Because piano-tuning software is based, essentially, on measuring "partials" generated by piano strings; and those partials can be measured very accurately by the cheapest mic on the market. Spending even, say, 100 bucks on the cheapest of measurement mics (with ostensibly flat frequency response and low distortion) is just a complete waste of money. A 5 buck mic will serve just as well.

Seems counter-intuitive, but these folks seem to be in agreement; and they are the engineers behind the product.

What gives? Is this true? And if so, why?

 

Makes sense to me. thejackattack will be the man to answer definitively as a piano-technician.

As a guitar play / repairguy who has had enough people bring them tuners they've dropped on the floor one too many times - I can only tell you my observations. Every standalone tuner I've ever seen employs a ridiculously cheap microphone inside.

Tonal accuracy doesn't seem important to the tuner's primary function of frequency-counting. In other words, you pluck/strike an "A" string - the tuner counts the vibrations per second to a high degree of accuracy (to several decimal points) even from the el-cheapo mic. The tuner (or tuning software) knows that an A (A4) should be 440Hz and indicates sharp or flat based on that. Software based tuners I suppose have more than one temperament.

That's my take. Hopefully we'll hear from an expert.

 

Dave hit it pretty much on the head. Why we as recording engineers purchase a microphone is to capture the complete tone of the source in all it's facets. We also purchase branded microphones because of the consistency a given manufacturer has for it's designated model. Remy is fond of pointing out that the much praised Crown boundary mic's use a cheap capsule that used to be had from Radio Shack for a couple bucks. The difference was in quality control and testing as well as the design of the housing itself.

For piano tuning, we don't need a completely nuanced representation of the sound. We need to have a measurement of frequency of fundamental and overtone, and that measurement is independent upon strength of the partial/harmonic. In other words if I am measuring 3:1 aurally (or with machine) in the 5th octave (eg g5:d7), the relative strength of those myriad partials that create the tone are irrelevant. It is the quelling of any dissonance between the third partial of g5 to the root d7 that is the point.

Robert Scott could explain it much better if you email him from a scientific view but that is the practical nub in a nutshell.

 

As a corollary, by the same token I would never use the cheapest mic or even cheap measurement mic to provide some sort of visualization of a note for tone building purposes. Actually, I would never try to measure partials to aid in tone building. Voicing is so conditional on not just the instrument quality and felt condition/prior work but also upon the room within which the piano is being voiced.

 

Sometimes i feel the whole enterprise (electronic tuning of a complex instrument like the piano) is a fool's errand. But I persist. My ears, what there is of them and such as they are, tell me that Verituner has provided me with some downright pleasing tunings of my Hailun 218. Maybe it's just the unisons, which are well nigh perfect with VT (or any other decent edt); but also it sounds like the whole instrument is resonating after a "successful" tuning.

There's the rub. Whether it's room vibrations, the inherent variability of partial measuring, or the microphone and its location relative to the string being measured or the microphone's "quality", the problem is that I'm getting a lot of variance in my IH targets, too much for me, in fact.

So I'm trying to isolate the weak links. I'm pretty sure it's not mic quality, although mic "type" (card, omni, usb, etc) might have some impact.

Robert Scott would probably be pretty bored with this discussion by now, since he's commented ad infinitum on it! He did pique my curiously, however, when he suggested that any mic "good enough to record your kid's high school band" would suffice. That left the door open for me (unfortunately).

 

Of course I use a $3000 mic to record my wife's HS orchestra with my son playing......

If I might be allowed a digression, if the room in which you are tuning your Hailun has quite a few hard surfaces that cause significant reflections, that can cause problems even for an aural tuner. When I learned tuning it was as an aural tuner. It was a natural extension of my horn pedagogy especially in regards just intonation, historical practice etc. After I reached a minimum level of competency as an aural tuner, my mentor recommended I select an eletronic tuner to augment and speed my growth. I did choose Tunelab but Verituner is an extremely good tool. Basically I tune aurally and double check with the electronic device. By changing the octave styles to purposeful ratios I began to learn what the different styles sound like in the upper and lower end of the piano-they are all extremely similar in the temperament octave.

If you want to learn to hear better I recommend the following process. Program your Verituner for a fresh tuning (or recall a good tuning). Tune C3-G5 as absolutely accurate as you can. Go back and redo them as many times as necessary to get them to hold. Now, beginning with G#3, Tune the P5 C#3-G#3. Get it as absolutely perfect as possible. Now play D#3-G#3. Lower G#3 slowly until most of the dissonance is gone. Re check and make sure that C#3-G#3 is not beating too much. If so restart the procedure. Once you have a good balance between the two intervals check it against the Verituner. It should virtually match. Move on to the D3-A3/E3-A3 intervals and repeat the process. Work in this way as high up the keyboard as you can hear the partials. At least work through F5. Once you have worked up to middle C then you will include the octave in your aural measurements. This is a much easier process to demonstrate than to type. At any rate this process will train your ear along. Do not cheat an look at the Verituner until after you have decided aurally the note is correct. In this fashion your learning time will be halved provided you don't cheat. The lower registers will just be a bear until you learn what NOT to listen to.

 

Oh, and unisons should ALWAYS be tuned aurally whether one is a hybrid tuner or strictly electronic tuner.

 

Endnote: I love the tunings I get from VT, even more I fear than the tunings I get from supposedly top notch Toronto tuners. Perhaps I've accommodated to bad etd tunings, I'm not sure!

I tune the unisons by ear AND check with VT (ie I'm not a tuner, but I do tune my own piano, 'cause I like it in as perfect tune as I can get it, ALL the time.) The higher one goes of course, the less relevant the VT numbers are on individual strings of the trichord. They might all be 1 cent sharp, for example, and yet, under certain circs, all three as a whole can very quickly go quite flat as the note decays, and the note under these circumstances usually has a very weak sustain.

Interesting stuff, the way in which vibrating strings interact!

 

Yes indeed. One thing you should know up front. The tri-chords are not always perfectly in tune with each other. In order to cancel phase anomalies of the central string perhaps do to bad termination points or just imperfections in the wire (or for recalcitrant notes, one of the outer strings) the strings often will not be exactly the same to a sensitive meter.

Point 2: As you pull the other strings to pitch, you are often adjusting the already tuned strings as well. An experienced tuner learns to feel how much "overpull" one has to do on a given piano in order for them to fall into place. This will be dependent on many factors including how tight the pins are as well as how well the bridge pins hold tight. If the frame of the piano is itself flexing that is also a factor. In short, only experience can tell one how much to pull. Always approach the pitch from above however-just not from a great deal above.

There are many reasons you might not like the local tuners versions of equal temperament. I personal prefer just intonation as I am mainly a horn player musically speaking. Just intonation is not practical on a modern piano however. Also, not all tuners aural or mechanical are equal. If one has not the chance of being mentored by a knowledgeable and sensitive master one has to reinvent the wheel. I only know a few quality tuners in Canada and they are mostly in the BC area. There are others of course but I don't know them.

 

I did a wreckless thing: I purchased a down-and-dirty cheapo measurement mic, a dayton umm-6 usb measurement mic--just to put certain "issues" to rest.

Anyhow... now I have this unexpected issue: how "sensitive" are these measurement mics? This thing is hard to clip, and short of talking right into it, barely stirs up win7's mic level meter!

Is that typical of measurement mics? (It does the same thing on all my computers.)

JG

 

Do you have the auto mic level clicked on? If not turn up the level in the windows mixer.

Sent from my DROIDX using Tapatalk 2

 

Oh... it's up all right, level at 100 "properties/level/ etc.

I was thinking that maybe these meas. mics are pre-calibrated for LOUD speakers. Although a large grand piano is pretty loud when you've got the mic right up against the strings.

JG

 

They are calibrated to have as even a response to all freqs as possible by man's hand. Some are more sensitive than others but most I've seen should be too hot even on a laptop mic oreamp.

Sent from my DROIDX using Tapatalk 2

 

Did some poking around on the net. Turns out that these mics are made to accept 130 db levels without clipping. So probably normal that it's not that sensitive re piano strings.

Also, for the benefit of anyone into piano-tuning software, I've had a chance to check the mic out and find it provides quite a bit more precise "inharmonicity" measurements, which is the at the core of all piano-tuning software. By "precise" I mean that repeated measurements yield IH values that are much closer to each other, on average. So, it's much easier to arrive at an accurate IH target for any given string on the piano.

I don't think the reason for this is that the mic is more "accurate". Instead, it's a result of the fact that the mic "hears" everything, but without getting remotely close to clipping. This means you can move the mic within fractions of an inch from the piano string, pound the string as much as you want to without clipping, and consequently get very complete partial information. Another benefit of being able to get close to the string is that the mic is picking up fewer room reflections, and more string vibrations. The further the mic is from the string being sounded, I've observed, the more unpredictable the IH measurements.

So there you have it. I think I can get my piano in pretty good shape tuning-wise prior to recording it--which was my main purpose here. (Other than having the thing in tune pretty much all the time, which is nice.)

 

Your thoughts are inaccurate. Most modern condenser microphones can accept 130dB prior to clipping. Generally, a piano will be able to overload any condenser with very little effort and sometimes to the point of needing to pad the mic input. Also, most microphones have particular quirks and such that create unevenness in the eq response not only in the capsule itself but in the modulation caused by the head basket itself. A "measurement" microphone is designed to give as even a response across all frequencies. They are used for a variety of reasons but one main purpose is to "tune" a room's response including taming reflections and agregation (especially bass build up). They can also be utilized to find a room's particular resonant frequency the knowledge of which helps when dialing in eq to squelch feedback for live sound (especially churches auditoreums etc).

Inharmonicity definitely will vary by position. The further one is from the string modulation the more the room and reflections affect the readings. Honestly, if you find a particular tuning file to be good then just save it and reload it when you next tune. You will only be able to really use this file if you are tuning the whole piano. If you are "touching up" then the best you can do is to pull the unison strings into tune as once the piano begins to shift it will do so unevenly across the 88 notes.

 

Thanks for the heads up! So maybe/probably my umm-6 mic has issues after all, and I'm inclined to return it now, given what you've said about clipping.

But I take it I'm not wrong in thinking that any mic that can be placed really close to the string-being-measured without clipping will get more "accurate?" IH measurements, "accurate" being a weazel word to denote "absent the effects of room reflections on IH measurements."

JG

 

To a point yes. I am usually measuring with the PDA sitting on top of the vertical. On pianos that are particularly wonky or in a noisy environment I put the PDA on the hammer rebound rail or even set it on the key sticks that aren't being tested. This gives me a reading that is much more than accurate enough to create the custom stretch to make a piano sound awesome. The rest is all in the hammer technique to make the string stable and the experience to feather in the notes. It is often said that it is about 1000 tunings before one is considered a good tuner. I think if one is highly sensitive and kinesthetic and has good mentoring it can be done quicker.

Now as to clipping. The input on your Verituner or PDA is just plain NOT going to be able to take the same levels as a quality audio preamp. Al Sanderson felt like a magnetic pickup was all that was required. I have one and it just sticks on the plate to pick up the vibrations. I don't use it hardly ever as the PDA mic is good enough. On occasion I have to move it slightly around to get a good reading but not much.

 

OK. I've replaced the Dayton usb measurement mic with UMIK usb measurement mic. Same low levels. But it works as advertised. For whatever reason, the UMIK is actually quite quiet compared to the Dayton. It makes decent recordings. The Dayton had enormous noise; so something was definitely wrong with that particular mic.

Conclusion...(which I have no way of confirming with extensive scientific testing!): inexpensive usb measurement mics will work perfectly well to tune instruments (like the piano), but recording levels are (at least for the Dayton and the UMIK) likely to be low on the average pc soundcard. The UMIK made decent recordings, although that's not what I'm using it for.

If you want more control over levels, I guess you get yourself an interface or preamp, and opt for a standard (non-usb) measurement mic.



JG