Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum Double String Unison (DSU) Tuning Technique

Thanks, Mark, for your question.

In answering it, it brings up a point that must be made; a clarification on all subjects educational. One must make the distinction between fact and opinion, truth and technique.

With Double String Unison, my claim of the benefits is an opinion. In answering your question, we must start from the point of truth and fact. I present these "truths" for all to read, unashamed, because I am confident no one will argue, because they are truth and fact. From there, one makes their own opinion of what technique is more useful for them.

Fact: being able to hear beats takes time. The better teacher you have; the less time it takes. There comes a time with all good technicians, when being able to hear and pinpoint specific beat speeds at certain partials, is much easier, and not a challenge in most normal tuning situations.

Fact: in order to beat match an A3 DSU to an A3A4 octave (in one pass), you need to be able to pinpoint the beating of both the A3A4 octave, and the A3 DSU, at A5, the 4:2 of the octave, and the 4th partial of the A3 DSU. (or the 6:3)

So, if you are hearing the mixture of different beat speeds when trying isolate A5, I don't think you would find this technique useful.

But not yet anyway. The story is the same for unsions and stability. DSU forces you to concentrate on isolating beating partials, on locking in those unisons, on producing rock solid stability, because if you can't do those things, the technique doesn't work.

Other tuning techniques allow the tuner to be less than perfect with those elements. Unisons can be left less than clean and less than rock solid, if they are the last to be tuned, and you don't need them to be clean and stable in order to tune other notes. Tune your unisons and go home. That's how I used to tune. DSU won't let me do that anymore.

As for trying to hear the specific beat speeds at a specific partial, you actually don't need to, (Yes, I am contradicting myself), because the first approach doesn't have to be right on. You can use shimming to get more accuracy after the first approach. I used the analogy of golf previously.
With a 4:2 or 6:3 beat isolation, you have a better chance at a "hole-in-one". I just tuned 9 octaves using DSU and beat matching and got a few holes-in-one. I will post the recording soon.

Also, try to match the "colour" of the octave with the "colour" of the DSU. This is a suggestion I use a lot. Isolating beat speeds can be frustrating, but listening for the "colour" of an interval is much more musical and as musicians, we tend to be able to do that more easily.

So, try to match the degree of the honky tonk.

Good luck. Hope that helps.

Yes, I use this in the treble, but not plucking. I will sometimes use the Papps to isolate single strings. Up there, differences in Hz are easier to hear, melodically. Not so much, I find, down below.

Cynthia, look for my video on stability, coming soon. No technician should have any trouble with stability of they understand these points.

Watch Isaac's video on the unison colour, the Smiley, he calls it. Do not gloss over the later description and demonstration when he is tapping on the hammer handle. I have never seen anyone do that, but I know exactly why he is doing it. Watch it again. Think about why he is doing that, and what is happening when he is finished, and why he thinks the note is now stable. It is remarkable. Issac, with your permission, I am thinking of taking your video and overlaying a graphic showing NSL tension against the Tension Band. I think it could be quite instructive for some techs having trouble with stability.

Hello, tapping is as a test blow seen from the tuning pin side.

I do that to show the tension ratio and prove also the pin is stable,

I guess it can be one when learning, to prove himself the job have been one well.

What is sure is that if the note stay low (or high) after such tapping, it can happen later when playing, so it can get some more attention.

The pin flagpole as if the piano was playing very hard, also.

I need to know how are acting the NSL and speaking length anyway.

There is a clear sensation about the tightness of the pin, also. the ratio can be tested very fast by wiggling, and the pin left more free than in the demo, but whenever I need to I make the pin really tight. The tone mostly benefit from that, it get cleaner

Yes if you wish to add comments on what happens on the NSL side and capo, the video is public anyway.

Regards

Hello Mark,

I've spent the last week devoting your tuning theory to practice. Tuning all double strings from the temperament upwards and downwards. Here's what I've found:

Although I have an experienced ear, I had a hard time sifting through beatrates. In the case of two strings tuned as a unison, very close in pitch to the target note, beats arise that conflict with the whole tone target.

Example: Tuning a 4th/5th pair and referencing the 8ve with 10th and 12th checks in the region of A5, the second string can cause beat interference that matches the string being moved, making the pair seem more in-tune at incorrect places, or masking the in-tune point. Several times I had to detune the other string further away to get an accurate reading on the one being moved.

In some cases, I got the pitch right on the first try by using experience to cut through the noise and really zone in on one partial. This was the exception, rather than the rule, however.

I found that the stability to be gained by tuning two strings together was really a loss for accuracy. I could shim to move the point, but I wouldn't achieve the whole tone target that way - only a close compromise that was further away than any kind of Weinreich effect. It may be that this is a method useful if you are a partial-only tuner. If the only target is beatrate matching, then possibly guessing and shimming is a good way to go.

I've pursued methods like this - and open string methods for the temperament - for almost 8 years now, long before you posted this information. I keep coming to the same conclusions about it, which I wish wasn't so, because besides for loss of accuracy in placement, stability should increase.

The application of methods like these seems to boil down to two different basic ideologies in tuning: deterministic and variable-dependent.

There are those who pre-determine an interval and move the note to its target point, and then there are those who use networks of notes to define the tuning. Unfortunately, what is gained in one method results in a loss somewhere else.

When many variables are used, muting and single string tuning is essential, otherwise the competing data and the time to retune increases exponentially. When pitches are predetermined, more strings can be tuned at once. Which method results in a better product? Losses occur with both.

The variable-dependent method is subject to drift, but the deterministic method seldom gets the pitch exactly right in the first place. I would prefer to make shimming corrections to variable-dependent tuning.

Thanks Tunewerk,

Are you listening to an interval while detuning the DSU? There should be absolutely no beat interference.

The method is no different than tuning A4 to the fork using F2. Trial and error, within a small, very small window of acceptability.

I hear what you are saying, and struggled with posting this method because the benefit is only realized, for me anyway, when all aspects of my method are used together.

So, I use a "network of notes" to determine the best placement of pitch. And I use P4 and P5 windows to make the "pre-determined" placement of the final pitch very precise. In this way I am able to choose which kind of tuning temperament I would like in the treble, for example. I can tune pure 12ths, or pure 22nds, or, pure 19ths. And they each give a different tonal quality to the tuning.

I recently recorded myself tuning a console with pure 19ths, which produces the most stretch acceptable, in my opinion. Very wide 2:1 octaves and bright fourths in the treble, but pure 19ths which ring better than tunings with less stretch. An interesting byproduct of pure 19ths is the wide 12ths which I didn't expect. I've read about pure 12ths, and tempered 12ths, but never wide 12ths! I'm starting to really like the flavour of this type of stretch. Also, the windows allow me to be more consistent with SBI sizes in the treble, rather than just guessing and stretching by ear. DSU allows me to be more precise within the windows when using shimming. I will post the recording soon, if there is interest.

Plucking is good for beginners to train the ear.

It is not really possible to fine tune that way because of the low energy entered in the strings that way.

But most tuners begin by chipping and plucking the strings, be it to get a hold on lever manipulation and listening to a lot of partials without the ears being agressed by the power of tone.

Then using earplugs would help when beginning to tune while playing the notes.

Then the ear and the playing hand can get accustomed to listen and play more quietly.

That was the points of my remarks

Yes, please post your record. Thanks

Yes that may be interesting.

I personally avoid more than pure 12th (as more than pure 5ths, they sound not nice to my ears) . That said on very small pianos it may be unavoidable, (with "standard methods)".

So, if you are hearing through all the noise of two strings, one in movement and one out of tune, with all those test intervals, I'm impressed.

This is certainly not a beginner technique and many people who are advanced cannot do this.

I'll give you an example of why I'm not confident with this approach. Let's say I'm tuning A5 and listening first to the 5th below. I tune both strings of the note and they are about 1 bps flat of the target.

I can perform all checks to know exactly how flat the note is. But, when I begin to move one string up, listening to the 5th, what's happening is 3 relationships:

- D5 to A5 5th in the mobile string
- D5 to A5 5th in the stationary string
- A5 mobile to stationary string

If D5 to A5 is creating 1 bps flat with the stationary string at A5, then when the mobile string hits 1 bps sharp, the 3 string pair will seem to be in the most tune. It will be hard to filter out the quality of the in-tune point.

Those of us who tune all the time, know what a small change 1 bps at A5 is.. just a flexing of the pin can create several bps change. Also, the window of acceptability at A5 can be large for a 5th; several cents in either direction.

I'd be interested to see how you actually do this.

YES! I thought so.

Do NOT play D5 while detuning the DSU.

Ok. So you have determined that A5 is flat by about 1bps from pure. You also insightfully refer to the sensitivity of the pin.

So, you know that A5 must go up by the smallest amount.

Play A5 ONLY, and listen to the clean DSU you recently tuned.

Play the A5 DSU and pick a string and nudge it up, breath on it, massage it up, whatever. You just want to hear the tone change a tiny bit. Maybe its colour just brightens a bit, or you hear a slow whine. If the NSL tension is close to the tension band limit, it will move. If not, you have two choices: massage harder, or move the foot. Impact may also work here.

For me, it is at this point that I get the best feedback on where, exactly, the NSL tension is, within the tension band, and can be more sure of my stability, without test blows.

Then, bring the other string up, erasing the colour or whine you just created, and restoring the unison colour you like, blooming or dead on.

Now, you have just raised the pitch of A5, the smallest amount possible.

This is called shimming or cracking the unison. Recheck the fifth and reshim, if necessary.

I hope this helps. Thanks for trying it out and posting questions. Let me know if it makes sense now.

Something can be noticed :

You hear the primary beat at a certain level

It is then more easy to recreate it in the tuned unison , at that level (for instance you hear a 4:2 beat more evident, and create a 2nd partial beat in the unison, both beats have the same coloration.

Now about final justness, that one relates more to theoretical model of partial matching justness than on congruence of partials all together (plus fundamental) .
I accept partial matches as a reality of course, but their usefulness for tuning is limited in my opinion .

To learn, yes, to learn to be precise, too, then a small switch to intervals activity (or beat speed) may be useful in my opinion.

Best regards

Okay, this answers the question for me. I have a good deal of experience with shimming, so no need to explain there.

You listen to the tests, and then separately move the unison. Every time then, the movement is a guess. You must not be tuning by whole tone, but strictly by estimation of partial distances to where the note is within tolerance.

This is what has always frustrated me with this approach. Talking with you about it has brought me something - to hear someone so passionate and also experienced with a technique; to confirm that the issues I've experienced with it are the same.

The only disagreement I have with the claims of this technique is: error from the ideal point can still be larger than the Weinreich effect. Other than that, it should be a quick way to estimate pitches, putting more emphasis on unison stability, which may be a great practice for beginners.

Thanks for sharing your thoughts and experience.

Consider that there is no ideal. Only a window in which we are pleased.

I like to think I have some sensibility that allows me to perceive the margin, top and bottom, while those 2extremes seem to result from a different approach to the problem.
Circumstances are also allowing that margin to be more or less small.

On any tuning, in the end, an accurate listening will point some weakness in a part of the resulting harmony.

Mark, I know it is just an example, but a 1bps is yet a noticeable beat, particularly in the a5 range. why not creating a 1bps in A5, directly, it is not that slow to ask for very small changes, the pin can be moved yet somehow. In mean, 1bps is not moaning but clearly a relatively fast beat. with a 4-5 seconds sustain we have enough time.

Regards

I've got 14 reasons why I don't tune A5 directly. If the choice was only mute strip or open unison via single string, I'd use the strip.

Cheers,

I was not attentive it was a 5th; then you create a 5th "in the unison" , does not need to be tune clean first in that case, I think.

you can tune the second string of A5 to the 5th kind beat + 1 bps.
it was what I mean - if you want to get a pure 5th you tune a 1 bps in the upper note unison directly.
I miss something ?
Best

Sorry, I think we have a language barrier. I'm not getting the point. Désolé. Je pense que on a une probleme avec le transduction. Essayez en français?

Sorry Mark

I wanted to say you can create a beat in 2 strings of the same note that sound as the M3 or M or 5th beat( as whichever interval you are looking for)

Then there is no need to have yet 2 strings in unison in the note you are tuning. (while I see your point, about Weinreich...)
As a control is to be done why not doing it when the 2 strings are in unison, before tuning the last one.

One need 2 tuning mutes and not one.

Regards

I still don't understand. Tune 2 strings to beat match the interval, and then what?

That's what I do, but then just clean up the unison.

There are two things here that I think some people are missing. How it is fast, and how it is precise.

For me, this technique is fast, very fast, and extremely accurate; the most highest precision tuning I have ever done.

Then why can't people replicate the results?

Simple. It took me two years to get comfortable with it.

People have mentioned that it is not accurate, yet others site Unison Cracking or Shimming as an ultra high precision method for concert tuning. DSU is beat matching, then shimming.

When I fit a note into a P4 window, I produce SBI intervals that are so clean and consistent, I have no idea how I would be able to do that "directly" i.e. with single strings, listening to the sound of only one SBI interval while tuning.

Fitting the note into the P4 window is like listening to four intervals at once, for compromise, and then fitting the note so it is compromising with all four intervals at the same degree. Shimming is the only way I can do that; the precision is that fine.

And don't techs most times check a 'directly' tuned note with all the other interval checks after tuning the single string, and adjust if necessary? That uses a directly tuned interval as an approximation, just like beat matching.

I have to apologize because I have done a very poor job explaining the merits of this approach. It was intuitive to me. Enough for me to invest one year practicing and slowing down my tuning time in order to improve my skill at using it. I was hoping more people would be interested but I know how attached techs can get to their techniques. I have only just started using a different temperament sequence after saying I would never change. We invest time and energy into learning a technique and then that investment is very hard to throw away. I used to call people who tuned without mute strips "old school" and couldn't imagine myself doing that. I thought "way too slow and for no benefit". But then again, my unisons and stability were horrible, and I was the only one who didn't know it.

No worries though, I am committed to continue developing it and discussing it whenever and wherever people are interested.

I'm always available for questions, and I have a few more videos and recordings I've made that I will be putting up on my website soon.

Thanks all who read and commented.

Isaac, maybe if you wrote out some notes, step by step, I would understand better what you were getting at.

Best Regards,