Pianotech

Subject: Rescaling Early 20th Century Grand

1.  Rescaling Early 20th Century Grand

Posted 10 days ago
I am preparing to restring a 1925 Gerhard Heintzman 5' 4" grand piano.  The existing scale averages about 130 pounds tension per string in the tenor, around 140 in the low treble, and under 120 in the high treble.

I have used Scale Ripper in the past to even out scales, but have never encountered a scheme like this one.  The string lengths are short -- significantly shorter than the 5' 2" Knabe I restrung two years ago.

If I evaluate the existing setup in PScale, I find that the first 16 notes of the tenor are slightly above what Parsons has identified as the preferred range for inharmonicity.  Adjusting the stringing scheme so that the tension averages 140 pounds across the plain wire area raises the inharmonicity slightly further in the tenor, and drives the inharmonicity in the top octave-and-a-half of the treble quite a bit beyond what Parsons indicates as the preferred upper limit.

If I play with the original scale in Paulello's Typogram, I can bring stress percentages into the ranges Paulello suggests by substituting wire types in the low tenor and the top octave of the treble.  It looks like there might be something worth exploring here.

What I do not know is whether I should simply smooth the tensions of the existing scale (the stringing has no half sizes), or whether I should follow my usual procedure of bringing some degree of overall evenness to the tension levels across the whole plain wire section, and then explore how the different wire types can be used to achieve appropriate stress percentages (% of break point)

There are clearly some gaps in my knowledge!

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------


2.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
Edited by Chris Chernobieff 10 days ago
A great designed little piano with tons of potential. But, you're not going to improve the scale too much. It's pretty darn good as is . Heintzman knew what he was doing.
Here it is on a spreadsheet. Tension, InHarmonicity, and Impedance.

The problem with the piano is the soundboard Rib scale. IT'S HORRIBLE!!  It might as well be a dining room table in there.  The panel was a non tapered solid 3/8 through and through.

Notice in the diagram that the largest rib #7 is not the longest rib #8. Just one of its many problems. Plus, I suspect that their procedure of installation is faulty. At least the board I removed was. It was flat as a pancake in the piano, but crowned up significantly when removed.
I made a new soundboard with a classical structure and Improved the rib structure. Incredible sounding piano now.
Putting a hybrid rescale on a terrible rib scale don't make sense to me.
--------
​​My philosophy on restringing software is to not use them.  Better to make your own. Learn the math formulas and how they work first, then create your own speadsheet with the proper knowledge to back up your rescale. Rescaling is NOT just looking at smooth lines on a computer screen. My 2 cents.
-chris
#caveman


​​

------------------------------
I have a piano in my Nuclear Fallout Shelter, and my competitors don't. How silly is that?

chernobieffpiano.com
865-986-7720
------------------------------



3.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
When this piano was built, Heintzman & Co. and Gerhard Heintzman were distinct companies. The Gerhard Heintzmans were of reasonably good quality, but the Heintzman & Co. pianos were definitely superior.  I suspect the data you have posted is from a better piano than what I am working with.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------



4.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
Floyd, you could help us out here by giving us a few stats about the tenor part of the piano, since that is where your greatest concerns lie.  Can you share the speaking lengths and wire diameters for the first plain wire and the 20 - 25 notes above that?  That way, myself and probably some others can throw the data into our own scaling programs and give some commentary based on more data.

Given the low tension scale this piano has, it just seems ripe for rescaling using hybrid stringing.  Most likely the breaking percentages (or Practical Breaking Load percent PBL%) as the typogram is referring to) are too low for much of the piano, and the substitution of Type 1 in the very low tenor, and type O for much above that will result in a substantial improvement in tone by raising the breaking percentage.

Moreover, hybrid stringing allows you to have your cake and eat it too.  While your impulse would be to bring the tension up by increasing the wire diameter, that comes at a cost in these small pianos.  Once the wires get to a certain size, the tone sounds choked and not so pleasant.  Along with that, increasing the wire size will lower the breaking percentage - going in the wrong direction away from good tone.

Where the rebuilder has only one wire type (as in Roslau or Mapes for us), there aren't many other choices.  In the very low tenor, you can add wound bichords.  The plate won't allow you very much to increase speaking lengths that would be meaningful.

The focus on tension in smaller pianos here leads one away from the most effective method, which is hybrid wire.  I will briefly make my point with a 5' 2" Baldwin M that I rescaled a while back, using the first plain wire (note 30) as my example, with a SPL of 906 mm.

Original stats were 21.5 gauge wire diameter, with a PBL% of 32.49, string tension of 62.85 DaN, blank inharmonicity of .33.

What was substituted was 19 gauge Type 1 Paulello wire, with a PBL% of 48.75, string tension of 50.44 DaN, blank inharmonicity of .26.

This was an experiment for me, based on using PBL% as the most important consideration in scaling.  I knew that if I had used Type M at 19 gauge (or Mapes or Roslau) it would sound like flabby garbage. Had I dropped the tension too low, which went from in the 140's to about 111 lbs.?

The result was rather startling.  The PBL% of 48.75 put the note solidly in the goldilocks zone of good tone regarding breaking percentage, the inharmonicity had dropped from .33 to .26  AND it sounded great - open, free, and sustaining, and more in character with the notes above and below it.  I similarly made changes throughout the low tenor, tapering off to about the original around note 50.  This was also transformative.  I heard greater clarity and body and a better voice.  It was a remarkable change that demonstrated what a powerful tool hybrid scaling can be, really the most powerful tool of all, as it can be combined with the other things that we do in scaling.

I calculated all this using Arno Patin's Abacus spreadsheet, which is set up to be used with Paulello wire.

As Chris says, smooth lines are not enough.  The ear is the ultimate arbiter, it either sounds good or better, or not.

Chris, if you want to write your own scaling programs, do so by all means.  For myself, I know that there are people out there who have considerably more expertise than I in most any area of piano work.  That includes people who write scaling software.  If I can buy a program that makes me better at what I do, I am all for it.  I don't need to be the smartest man alive, just one who is wise enough to work around my own limitations in this journey of discovery.  I will say that in the 6 plus years that I have been doing hybrid stringing using dedicated software, there is no doubt in my mind that it has greatly elevated the quality of my finished rebuilds, and that starting this work was a seminal moment for my work.

Will Truitt


------------------------------
William Truitt
Bridgewater NH
603-744-2277
------------------------------



5.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
Floyd,

In addition to what Will said (hard to add to that), the tension level you're talking about makes it a good candidate for Puresound SS wire. PS really shines in a low tension environment. The same principles apply for rescaling, however the PS gives another choice in the matter given your rather extreme case.

Just throwing that out as another possibility FWIW.

Pwg

------------------------------
Peter Grey
Stratham NH
603-686-2395
pianodoctor57@gmail.com
------------------------------



6.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
  |   view attached
Here are the the measurements taken at teardown (attached).

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------



7.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
Notes 26 to 29 are wound bichords in the tenor.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------



8.  RE: Rescaling Early 20th Century Grand

Posted 10 days ago
William,
The journey of discovery, as you beautifully stated, exactly highlights my point that it is beneficial to the journey. Not necessarily necessary.

Floyd,
Some points of concern for me is that no mention is made of some important info.
Beams- None, 1, 2, 3 or more?
Ribs- On a 5'4", if it has less than 11 ribs, a red flag would go up for me. (I look at even more of course, but that's beyond the scope of this post).
String Frame- Thin support or overbuilt?  Also, I like knowing the overall tension the string frame will be receiving. Before and after.  I can't remember, is that in Pscale? A bit of experience with string frames structure obviously helps here.
Hammers- Small or large?
Scale- I prefer studying the whole scale rather than a partial view before making any decisions. Also, I find tension and inH, not quite enough. What about Z ? I find it to be a useful guide and complement to Ten/INH.
Music wire- How do you know what wire is in there now? Are you assuming it's high tensile steel? Could be an experimental one of a kind model, and that the low tension was on purpose.  The bridge could have been installed in the wrong place i suppose. Maybe, a long backscale was desired by the designer? Could be an important factor in relation to the rib scale. But again, looking a little deeper could help in making that determination. BTW, its not that difficult to break some strings to determine (somewhat) the current wires tensile strength.

If everything looked "normal" and the structure was "sound" then why settle for 140lbs and a low BP%?.
I mention these points to illustrate that it's important to look at the physical characteristics of the instrument for making good scaling decisions.
Good luck with your project.
-chris

------------------------------
I have a piano in my Nuclear Fallout Shelter, and my competitors don't. How silly is that?

chernobieffpiano.com
865-986-7720
------------------------------



9.  RE: Rescaling Early 20th Century Grand

Posted 9 days ago
  |   view attached
4 beams, 11 ribs, string frame looks comparable to on-hand pianos of higher tension scales, more details of the string scale attached, and the original wires are no longer on hand for testing. Original total trichord tension: 23202 lb.  If average tension adjusted to about 140 lb, total trichord tension becomes 24324 lb.  If average tension is adjusted to about 165 lbs, total trichord tension becomes 28248 lb.  I have not weighed the hammers yet.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------



10.  RE: Rescaling Early 20th Century Grand

Posted 9 days ago
Here's an example of being able to have control over the math.
In the breaking Percent formula, most likely the Roberts formula is used.
The constant 1.667 should match the material. I just tested Mapes Gold .031, 24". Snapped at 440 hz. This equals 243 lbs of tension.  The BP% formula should equal 100.1%.

It doesn't.

It records at 86.8% using 1.667.  Obviously my sample has a different tensile strength than the Roberts sample.
I then adjust the constant. The real constant for this current batch is 1.625. That represents a 100% failure of my sample of music wire.

Usually, I have to fidget the number to match each batch.

Food for thought!
-chris
#caveman



------------------------------
I have a piano in my Nuclear Fallout Shelter, and my competitors don't. How silly is that?

chernobieffpiano.com
865-986-7720
------------------------------



11.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
I have created three drafts of the plain wire stringing using a hybrid approach.  I will post each draft separately, so as not to have a huge wad of attachments with one post.

The first draft uses the original stringing scale, simply substituting wire types to bring the breaking point % into a reasonable zone.  PScale tells me that with this scheme, inharmonicity is elevated somewhat above what Parsons identifies as the target range.

Breaking point % seems to trend appropriately.  I was surprised how tidily this seemed to work out.  Tensions don't fit nicely into what I perceive to the the current approach to scale design.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------

Attachment(s)



12.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
Here is Draft 2.  I started by ignoring everything except Breaking Tension %, then tamed down some really high tensions that showed up around note 47, where the wire type transisitions from Type 0 to Type 1.  Evaluating this draft in PScale showed that the notes 30 to 31, and notes 47 to about 54 came out with inharmonicity elevated above what Parsons has identified as the target range.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------

Attachment(s)



13.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
Here is Draft 2.  It is a minor modification of Draft 2.   I chose Type 2 wire for the bottom 3 notes in order to bring down the inharmonicity into the target zone.  For the same reason, I reduced wire sizes in the area above note 47.  I looks kind of odd to me to see three peaks in wire sizes, but that's where my path led me.  Breaking point % seems to trend well.  And interestingly, in both Drafts 2 and 3, total plain wire tension drops from the original design, which I thought was already a very low tension scale.

I have very much enjoyed and appreciated all the response I have received thus far on the two threads relating to this project.  This is a whole new area of exploration for me.  I welcome any comments on the data I am posting here.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------

Attachment(s)



14.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
I'll need to repost some data on Draft 3.  The charts do not look right.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------



15.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
Here is the corrected data on Draft 3.  I had not transferred the revised tensions into the spreadsheet.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------

Attachment(s)



16.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
I would not use type 2 at all...type 1 is a low as I will go, and as low as Paullelo suggests for a modern piano. Instead, try wrapping those notes to see what you can do. Soft iron wraps are a possibility if you need to reduce weight without going below .15 or .20mm wire size used for the wrap. Based on feedback from string makers and some experience with unsuccessful small diameter wraps, I consider .20 as my minimum wrap wire size in the wraps.

If you need soft iron wraps, Heller is the only supplier at this point. His shipping from Germany is amazing...inexpensive  2 day shipping from Germany.

------------------------------
Jim Ialeggio
grandpianosolutions.com
Shirley, MA
978 425-9026
------------------------------



17.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
I agree with Jim about Type 2 - the word is that it just sounds too different.  It may give you the BP% you want.  I haven't looked at the numbers (too busy), but  I would play with reducing the wire gauge along with Type 1 to see if it gets you where you want to be.  If not, then wrapped strings will be the way to go.  I endorse Jim's idea of the soft iron wraps - they sound incredible, but will be pricier.

Have fun with this puppy.  You are going to make it sound better, and this path will be an eye opener.  Go big or go home.  :-)

Will




------------------------------
William Truitt
Bridgewater NH
603-744-2277
------------------------------



18.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
It seems strange to me that BP% (I'm ignorant about these things, so I'm assuming BP stands for breaking point and not a petroleum company) is the determining factor for tone. It would seem BP only relates to a safety margin, i.e., it's important to keep the BP low enough that the likelihood of breakage is minimal. What about string deformation? Strings have to pass over points that will deform the strings and affect tone. How much deformation can a string take and still give a clean tone, one without beats? String leveling is also affected in that getting strings level and mated to hammers is important, so string tensions and deformation come into play for critical tone regulation procedures. 

The old Baldwin Hamiltons were purported to have a scaling with high string tensions. If that is indeed true, the result was lousy. Hamilton tone was terrible, IMHO. Those pianos were certainly built to last, however. So they sounded lousy but you couldn't get rid of them because they just kept on ticking. Has anyone ever attempted to "fix" those Hamiltons?

I'm retired and not likely to do any string scaling, but this thread has been interesting. 

Richard West








19.  RE: Rescaling Early 20th Century Grand

Posted 5 days ago
Richard, my understanding is that in working with the percentage of breaking point in terms of tension, it is not a matter of more being better.  Apparently there is a sweet spot that corresponds with "minimal internal amortization and spectral balance".  Especially in the tenor, the goal in the piano I am working with is to get the tension up to 50% of breaking tension.  That doesn't seem risky to me.

------------------------------
Floyd Gadd
Regina SK
306-502-9103
------------------------------



20.  RE: Rescaling Early 20th Century Grand

Posted 4 days ago
I am curious to learn from those who know:

How long has this "sweet spot" for BP% in relation to tone been known among those of you who do this regularly? It seems like a relatively new parameter.

What say you?

Pwg

------------------------------
Peter Grey
Stratham NH
603-686-2395
pianodoctor57@gmail.com
------------------------------



21.  RE: Rescaling Early 20th Century Grand

Posted 4 days ago
My knowledge base began with Arno Patin 6 plus years ago, who had worked with Stephen Paulello in France before emigrating to the US.  He was the Paulello North America Supplier for a number of years.

I think it began with Stephen Paulello and  the inventor of Pure Sound wire (Peter?), who wanted to create low BP wires to replace the original wire in older antique pianos, because it sounded better than modern wire.  How that evolved into hybrid stringing of modern instruments, I do not know.

Will

------------------------------
William Truitt
Bridgewater NH
603-744-2277
------------------------------



22.  RE: Rescaling Early 20th Century Grand

Posted 4 days ago
When I began studying piano scales around 1974, People like Jim Coleman Sr. and Darrell Fandrich were using break point percent as a guide to sound quality. I adopted it and tested many things to refine my practice. It seems to be the only really reliable parameter that is also handy. One still needs to intuit what size cores to use in the wound section. And controlling longitudinal mode is of paramount importance over inharmonicity. No accurate formula for L
-mode(s) yet exists. They are too dam slippery to predict with narrow exactitude.

------------------------------
Edward McMorrow
Edmonds WA
425-299-3431
------------------------------



23.  RE: Rescaling Early 20th Century Grand

Posted 3 days ago
BP% in relation to tone goes back to who knows when.  Early on it was the an important factor used to determine scaling by trial and error.  Non-pythagorean scale designs are found in many early harpsichords which used stretched scales to take advantage of the higher tension that could be achieved in the treble from the work hardening of the wire. Older instruments were scaled to come as close as possible to BP.  Since the wire used in the early 20th, and second half of the 19th century was quite different from what we have today, I imagine that this difference effects the scale choices that the original builders made, so that simply copying what is there in terms of length and diameter will not produce a comparable result using wire currently available....which is why restrung late 19th century pianos never sound as sweet as they did.

------------------------------
Jason Leininger
Pittsburgh PA
412-874-6992
------------------------------



24.  RE: Rescaling Early 20th Century Grand

Posted 3 days ago
Jason:

Thanks for that explanation. I have to say that I'm only just now realizing the importance of  the new string options that are available today. Perhaps I missed this in previous posts, but what are the current preferences for BP% to achieve the best tone for the various types of wire? I dabbled in string scaling using the P-Scale program years ago when there basically was only one choice, Mapes or supply house string. Sixty-six percent was the only breaking point I worked with. 

Is there an article or publication that discusses the various current wire types and their characteristics?

Richard West








25.  RE: Rescaling Early 20th Century Grand

Posted 3 days ago
http://stephenpaulello.com/en/les-5-types-daciers

Will Truitt

------------------------------
William Truitt
Bridgewater NH
603-744-2277
------------------------------



26.  RE: Rescaling Early 20th Century Grand

Posted 2 days ago
Richard, I think historically, at least in the treble, builders aimed for the note to be at pitch around 1 semitone below break point.  All the specialty wire coming available today seems to be just an imitation of the original stuff, or in other words, its goal is try to sound like the original wire....which is one way that it is marketed.  It would be nice if someone would just make the original stuff,.....but that would take a lot of physical work and charcoal. Also, earlier builders didn't need complex formulas to determine the string scale, all they needed was a monochord to break some wire on, practical skills in geometry, and their ears.

------------------------------
Jason Leininger
Pittsburgh PA
412-874-6992
------------------------------



27.  RE: Rescaling Early 20th Century Grand

Posted 2 days ago
Jason, we still have our ears as the ultimate judge of how good our scaling is.  We also compare what we are hearing with what our spreadsheets are telling us is good or not so good.  The formulas and spreadsheets were not available to the early makers, so their scaling had to be done empirically.

The metallurgy behind the making of the modern hybrid strings is much more sophisticated and controlled than what they did back in the day.

There are many, many bad scales from yesteryear out there.  In contrast, a piano technician with a modicum of understanding can create a decent scale in a matter of a few hours, using available software.

Will Truitt

------------------------------
William Truitt
Bridgewater NH
603-744-2277
------------------------------



28.  RE: Rescaling Early 20th Century Grand

Posted 2 days ago
William, I would be hesitant to imply that the wire from the past was not sophisticated,.....unless that is you understand the kind of hands on expertise and judgement required to purify charcoal iron, and temper and anneal it with the aid of very little if any scientific measuring devices and obtain repeatable excellent results....which they did. Also, I am not implying that all scales or designs from the past were good or better than today, just trying to make the point, that some of the very best designs, scales, and finest musical instruments were executed in this way, by these kinds of men, with those methods. Hoping to make known to others who are not interested in complex formulas, that there are in fact other extremely successful ways to draw up scales, without spreadsheets and computers. The proof is there. Hope this makes sense.

------------------------------
Jason Leininger
Pittsburgh PA
412-874-6992
------------------------------



29.  RE: Rescaling Early 20th Century Grand

Posted 4 days ago
I would suggest you visit Stephen Paulello's web site and wander around.  There is plenty of information there by Stephen, and you will get a better understanding from him than anything I could offer.

stephenpaulello.com

Will Truitt



------------------------------
William Truitt
Bridgewater NH
603-744-2277
------------------------------



30.  RE: Rescaling Early 20th Century Grand

Posted 4 days ago
I've used Type 2 on a late-19th century piano for 2 notes and they sounded so warm.  I don't think they stood out.  They just sounded great.

------------------------------
Douglas Laing
Tuner/Technician
Tampa FL
727-539-9602
------------------------------



31.  RE: Rescaling Early 20th Century Grand

Posted 3 days ago
<They just sounded great

Hmm..worth a try. I really dislike wraps on the long bridge, and especially do not like bichords there. Some times rebar sized plain wire is used on those last few notes. But maybe, given the "all options are not ideal", the 2 may be the least disagreeable compromise. I may give this a try next time.

The thing to watch with the type 2 is, that the very soft wire has a greater the tendency to corrode, relative to M. Nickel plating addresses that on M, O and 1, and is what I stock, but 2 is not offered plated.  


------------------------------
Jim Ialeggio
grandpianosolutions.com
Shirley, MA
978 425-9026
------------------------------



32.  RE: Rescaling Early 20th Century Grand

Posted 2 days ago
Yes, I did. During the late 1980s I redesigned the Hamilton. The original scaling was, as noted, very high. I designed a new tenor bridge and moved the V-bar to correct some hammer strike problems. I reduced the tensions from upwards of 200 lbf (per string through parts of the tenor) to the 180 - 190 lbf range. Still very high for a piano of this size but it was the best I could do without changing the string frame casting more than the company was willing to pay for. Doing it again I'd lobby harder for a new string frame.

ddf

------------------------------
[Delwin D] Fandrich] [RPT]
[Piano Design & Manufacturing Consultant]
[Fandrich Piano Co., Inc.]
[Olympia] [WA]
[360-515-0119]
------------------------------