Original Message:
Sent: 4/6/2024 12:18:00 PM
From: Maggie Jusiel
Subject: RE: why measure at the point of attack?
I haven't tried them! I should.
And yes about the wound strings. I was adding that to the mix because my experience with unwound strings is basically the same except divide by 4 in quality and divide by 10 in number.
------------------------------
Maggie Jusiel, RPT
Athens, WV
(304)952-8615
mags@timandmaggie.net
------------------------------
Original Message:
Sent: 04-06-2024 12:15
From: Nathan Monteleone
Subject: why measure at the point of attack?
I was gonna say roughly the same thing but Fred said it better 😂. And I'm strangely hungry for fish now.
I too encounter a lot of mismatched bass strings. I realize it's a bit off topic but have you tried the pitch lok string couplers that Scott Jones used to sell? They aren't a perfect solution but they are an absolute godsend when time/funds aren't available to replace the stinky bichord(s). I will be very sad when mine run out because I don't know that you can still buy them. I'll have to figure out how to make more.
------------------------------
Nathan Monteleone RPT
Fort Worth TX
(817) 675-9494
nbmont@gmail.com
Original Message:
Sent: 04-06-2024 11:56
From: Fred Sturm
Subject: why measure at the point of attack?
Maggie,
Concerning wound bichords, that is an entirely different kettle of fish. It is quite difficult to manufacture wound strings so that the bichords are perfectly matched as to the mass of the windings. Many bichords have such discrepancies in partial ladders that it is impossible to make the unison sound clean.
It is quite easy to see this with an ETD that is listening to an individual partial (like RCT) by zeroing both strings at the normal partial, then doing octave up and having it listen to another, higher partial. Discrepancies of over 10¢ are not unusual. (This is difficult - maybe impossible - to do with an ETD that automatically senses the pitch of the note you are tuning).
Compromising between matches at the two partial levels will generally yield the best place for the unison to be: not perfect, but generally bearable.
Regards,
Fred Sturm
"A mind is not a vessel to be filled, but a fire to be kindled." Plutarch
Original Message:
Sent: 4/6/2024 11:39:00 AM
From: Maggie Jusiel
Subject: RE: why measure at the point of attack?
Hi Fred,
I'm with Peter on this one. I think the reason you haven't found this is you picked random samples. WHEN I experience this, it it probably around .5% of the time, but with 250ish strings, that can add up enough for me to notice. Not counting bass strings, I find this doesn't happen often on newer, high end pianos. It's usually on older pianos or those with scaling problems. This tells me it could, in reality, be mismatched strings. As I mentioned to Nathan, this happens to me all the time with mismatched bass strings. I can get very clear differences read on an ETD (Tunelab & iRCT). Most of the time these differences are very small (around a cent, give or take), particularly in the unwound strings, but not always. Sometimes the difference can be as much as 2 cents in unwound strings and 5ish cents in wound strings. When I have mismatched wound strings and am reading them with an ETD, sometimes the unison will match one of the strings while the other is tuned sharp or flat to make a good sounding unison. Sometimes the unison is pitched between the two strings, with one reading sharp and the other reading flat. It happens often enough that I tend to be extra careful around the wound strings with unisons. I can easily tune a high bass section to my liking, then find the beat rates off after putting unisons together. ETD readings match this.
But, like I said, this averages out to around 2 strings per piano (some have no problems; others have several). Picking random samples will likely show no problems.
Maggie
------------------------------
Maggie Jusiel, RPT
Athens, WV
(304)952-8615
mags@timandmaggie.net
Original Message:
Sent: 04-05-2024 16:52
From: Fred Sturm
Subject: why measure at the point of attack?
Peter,
Speaking for myself, I find "the tendency for the entire unison to change pitch (usually flat) upon adding (particularly) the third string into the mix." to be a non-issue. DIdn't find it when tuning aurally for 15 years, haven't found it using electronics for nearly 30 years.
Aurally, having read Virgil Smith's assertions, I tested them by listening to beat rates. Single string against another single string, compared to three string unison to the same other string. No difference in beat rate that I could hear. Also was unable to confirm the phenomenon using an Accutuner once I started tuning with it - measuring pitch of each string, zeroed to the display, then checking the unison, I found no difference.
Then, in the late 1990s, Jim Coleman wrote that he hadn't been able to measure such a difference using his Accutuner, but when he got RCT, which had a more precise reading, he was able to find that difference, in the range of about 0.1 - 0.2¢.
Once I got Cybertuner, I made my own, very careful measurements with it. I carefully chose unisons made up of three strings that I could produce stable full blush on. I tuned the three strings of each such unison carefully to get full blush, and rechecked that they were spot on. I then read the unison. I again rechecked to see that all three single strings were zeroed.
Result: most often, the three string unison was at full blush. In those cases where the display showed a small creep from stable, that creep was as often in the sharp direction as in the flat direction.
Meanwhile, Jim Ellis had made some theoretical calculations of the physics behind the assertion, and found that at most it would be in the less than 0.1¢ range, insignificant for purposes of practical tuning.
I agree with that assessment. I further believe that the tiny flattening of pitch between about 200-300 microseconds after impact and 1.5 - 2 seconds after impact is insignificant for purposes of tuning. It is only that first quarter second or so where the maximum pitch change occurs, and that period is also where we are hearing all the percussive impact noise.
In any case, absolute pitch of individual notes, within one to two cents, makes little to no difference. The "just noticeable difference" for pitch, in rough terms as measured in many studies, tends to be around 5-6¢. When played in context (intervals and particularly unisons and then octaves) it is quite a bit smaller, but that is due to interference patterns. Raw pitch is very forgiving. Differences of less than one cent are imperceptible for most practical purposes.
------------------------------
Fred Sturm
University of New Mexico
fssturm@unm.edu
http://fredsturm.net
http://www.artoftuning.com
"We either make ourselves happy or miserable. The amount of work is the same." - Carlos Casteneda
Original Message:
Sent: 04-04-2024 09:10
From: Peter Grey
Subject: why measure at the point of attack?
One aspect of this that I had sort of forgotten about (not really forgotten about but just not actively thinking about it) is the tendency for the entire unison to change pitch (usually flat) upon adding (particularly) the third string into the mix. This is particularly evident now that I'm combining some digital assistants into my work. This of course is one thing that Virgil Smith taught all the time and it got me thinking...
Part of the reason I started intentionally spreading my unisons years ago was to in fact counteract (or help) this phenomenon. By tuning the middle string "slightly" higher than the left (but of course not introducing an audible beat, then tuning the right string similarly, I found that I could often (but not always) get the completed unison to stay put at the target pitch. (Some unisons can be quite stubborn and persistent in this regard).
Anyway, those of you who regularly use the digital assistant, I ask how do you cope with this phenomenon whether you tune on the attack or to the aftersound? It quite obvious with the electronics.
Peter Grey Piano Doctor
------------------------------
Peter Grey
Stratham NH
(603) 686-2395
pianodoctor57@gmail.com
Original Message:
Sent: 04-03-2024 01:21
From: Steven Rosenthal
Subject: why measure at the point of attack?
As an addendum to the comments I made yesterday.
I paid attention to my process while tuning one grand and one upright today, both having sections 5-10 cents away from the target. I observed that I spend 75 or 80% of my time getting close to the target's pitch and getting the pin and non speaking segments balanced and the remainder settling the system at pitch. The proportions might differ for an aural tuner. But the majority of my time and energy goes to the mechanical aspects of the system compared to the final setting of the pitch.
I do use sustained tones during the set-up period, it would be counterproductive to use a series of short blows as that would interrupt the process of moving the pitch several cents, pin setting... in other words stabilizing. I did overstate yesterday that it is only the feel of the lever that informs this process, clearly it is one of ear/hand coordination. Then, when actually setting the pitch I noticed I'm still playing relatively longer tones though I'm physically making the adjustment towards the beginning but I'm listening for how it settles and then check again for where the pitch is on the attack. And there may be test blows in there if needed.
While I do like the concept of tuning each string at the same point in the envelope, I think Steve N's assertion that there's no musical reason to play sustained tones is a bit of red herring in that whether or not there's a "musical" reason, there are certainly mechanical reasons to employ long tones in the process of tuning a string that will stay where you want it. Certainly if the tuning involves moving the pin in the block which is the case more often than not.
Sorry for all the verbiage.
------------------------------
Steven Rosenthal RPT
Honolulu HI
(808) 521-7129
Original Message:
Sent: 04-02-2024 03:51
From: Steven Norsworthy
Subject: why measure at the point of attack?
All,
"Where have all the pianists and composers gone? They instinctively knew the answer all these hundreds of years!"
We have not had the tools until now to fix a window of time in the first second from the attack and measure the partial frequencies averaged over that window of time. Now we do. Prior aural 'talent' cannot do this consistently. Analog signal processing tools cannot either. Now let's discuss the music and physics which coincide with the composers who knew all along.
The piano literature and piano physics unambiguously coincide and tell us the answer. The power is in the attack, and the composers knew that. Take a Chopin melodic line and you will find a slow passage has a quarter note = 48-60 range, so we get about one note per second. Isn't that interesting, because the piano decays 20 dB / sec in the middle register. 20 dB = 10x in amplitude. So after the attack we have lost 10x. In the last octave we see a 20 dB decay in 300 msec, and for low register, the 20 dB decay is about 2 sec. This is why I initially 'conceived' of the Freeze Frame concept for tuning. There is yet another physical aspect that the piano is an ever-moving target of partial frequency movement. I previously published this with graphs. Some partials will decay at least another 2-5 cents after the first second from the attack. Tuning to these is tantamount to tuning to a very weak amplitude that is now down another 20 dB or 10x so that would be tuning to something that is not even in the music that is so weak that it would cause a severe mistuning in the attack phase of the first second. Now for a funny rhetorical question, where is that piece of piano music that has one note held down for 3 seconds by itself without the sustain pedal, and the tuner says, "Oh what a beautiful piece of music that was!" Someone challenged me when I said that and showed me a Chopin prelude with a note that was a couple secs long in a melody all by itself but I pointed out that the sustain pedal is down throughout that passage. Oops!
Now if the piano partials held to their same frequencies during the decay, we have no argument and this thread is irrelevant. It isn't because the partials are also decaying significantly and at different rates, and the next neighboring half step has a different character as we see in careful spectral analysis, and we can hear that, of course.
If anyone wants me to re-post the sustain graphs of partial spread, I can do that. It is so revealing that it makes one wonder why all these years we have been tuning to a sustain that is not pianistic and fights the very laws of physics that the great composers intuitively already understood.
Best regards,
Steve
------------------------------
Steven Norsworthy
Cardiff By The Sea CA
(619) 964-0101
Original Message:
Sent: 04-01-2024 12:12
From: Maggie Jusiel
Subject: why measure at the point of attack?
Hi All,
There is something I have been thinking about for a while and I just don't understand. It's the current obsession by some ETD users with tuning on the attack. I've been peeking now and then at discussions on the subject, and there's one I'll revisit in a nutshell. Someone (I don't remember who) was demonstrating how important it was to measure pitch during the attack. In a section of his examples, he showed how results could end up being very different depending on how loud or soft the note was struck. The pitch history during the attack was chaotic. This is exactly why I would think the attack is the worst place to measure. The sustain is what the listener hears. It's where we listen for our beats. I'm not talking about the end of the sustain, but the beginning of it, still close to the attack & decay. I have heard recordings of various instruments without the attack and I was shocked at how poorly I could identify each instrument. Even a banjo sounded like an organ pipe during the sustain. While the attack is what creates the unique sound of each instrument, and sets up recognition of the medium for the listener, it seems to me that the sustain gives us our sense of pitch and blending of pitches. Why would we trust such a variable sample that can change depending on how loud or soft you play the note? This makes no sense to me. If anyone can explain that in a way I might be able to understand, I'd appreciate it.
Maggie
------------------------------
Maggie Jusiel, RPT
Athens, WV
(304)952-8615
mags@timandmaggie.net
------------------------------