Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum FAQ - Tuner-Technicians Forum Scale Design - tension

So I've read that Mason and Hamlin are high tension scale designs.

I've heard that Steinway and Sons S, M, L are low tension designs.

Anybody know where others stand in this range?

Bosendorfer, Grotrian, Schimmel, Yamaha, Steinway B, C etc?

Thanks.

Ulrich Sauter just told me that his pianos are 'low tension'.

[obviously designed for *high attention*.... ]

Norbert

I think I heard the Walter is low tension, where's Del?

I think he said here once that low tension scales tend to be mellow, and high tend to be more brilliant.

The Estonia is a rather low tension scale if I recall too. Please correct me if this is wrong.

Dan

The Walter 190 (with the tenor strings in the 155 to 165 lb range) is relatively low, though not as low as it would be if I were doing it today.

The Estonia (according to the published literature its tenor strings are in the 140 to 150 lb range) is extremely low.

Del

I've sort of wondered about this term, and whether it refers to an absolute amount of tension, or the percentage of tension to the breaking strength or the elastic limit.

You can change the absolute amount of tension by changing the string gauges, but the percentage remains pretty constant and can only be changed by a change of design. Hubbard used the terms long scale or short scale with harpsichords, and indicated them by the length of the strings on the C above middle C. A short scale would have low tension as a percentage of the breaking strength.

I can’t speak for harpsichord designers but, by convention, when discussing piano scales this refers to absolute wire tension.

While the percentage–of–breaking strength (pbs) is an important parameter to keep in mind (it's always best to avoid designing a scale in which strings are regularly breaking) it is not particularly critical through the tenor section of the scale. Even in a relatively high–tension scale like the Steinway D (with tensions in the 190 – 210 lb, or 85 – 95 kgf, range) the strings will be pulled to only about 35 – 40% of their rated breaking strength.

As well, changing the wire size to alter string tension has very little effect on the string’s percentage–of–breaking strength. For example, take a typical note F-33 in a concert grand: with a length of 972 mm and a wire diameter of 0.043” the tension will be 190 lbs and the string’s pbs will be approximately 37%. Increasing this wire diameter to 0.045” (keeping the length the same) will increase the tension to 208 lbs but the pbs will increase only to approximately 38%. Going back to the original wire diameter and increasing the length to 1,017 mm will also increase the string tension to 208 lb. This change will have a greater effect but still the pbs will only go up to approximately 41%.

It will not be the effect of the change in the percentage–of–breaking strength that will cause the tone quality to go sharper and more strident, it will be the higher impedance to motion across the bridges caused by the additional stiffness of the strings and the “tie-down” effect of the overall string plane.

Wrapped strings, of course, are a whole other issue.

Del

Del,
Why as your thinking evolved towards the lower tension scale? Interesting comment about the Walter, I had read a post by somebody once who complained about the 190's lowish tension scale, and how (this individual) felt that it would benefit from a higher tension design. If I had to make a choice I tend towards lower tension pianos (Estonia, Walter, Bosendorfer), but it does seem to be taste specific. What do you like and dislike about lower, and higher tension designs?

[/qb] [/QUOTE]Del,
Why has your thinking evolved towards the lower tension scale? Interesting comment about the Walter, I had read a post by somebody once who complained about the 190's lowish tension scale, and how (this individual) felt that it would benefit from a higher tension design. If I had to make a choice I tend towards lower tension pianos (Estonia, Walter, Bosendorfer), but it does seem to be taste specific. What do you like and dislike about lower, and higher tension designs? [/QB][/QUOTE]


Well, my taste has changed with experience and time. And perhaps I've become some less tolerant of crude and stagnet piano design over the years. I've never been particularly attracted to loud, raucous and unsubtle music and I find increasingly that I don’t care for hard and harsh–sounding pianos. And I’ve reached the age at which I no longer have the time or inclination to put up with them.

I now rarely go to piano concerts that are held in large halls because the piano must generally be voiced so hard (to ‘project,’ you know) it looses its inherent dynamics. Where is the subtlety? The dynamics? I do go to concerts in smaller venues where the piano can actually be voice to perform as a pianoforte and where there is some connection and intimacy between the performer and the audience. I may miss out on seeing and hearing some of the superstars this way, but so what? I’ll enjoy much more the connection with an artist who may well be just as talented — often more — but who lacks the personality drive to enter the business of profession performance.

If one is much of a fan of classical music at all you have to wonder what the original composers would have thought of this tendency toward the monochromatic, loud, fast and slavish reproduction of the works they labored over. If you’ve ever attended a concert performed on a fortepiano you know their instruments weren’t capable of anything like the enormous volume levels being demanded of the modern piano. Even with the advent of the so–called ‘modern’ piano during the mid to late 1800s the voice of the piano was much, much softer and more subtle that what we have ended up with today. (The felt wasn’t felted as hard and the hammer presses were not capable of either the pressures or the heat used to make modern hammers.)

One day before I have to hang up my mouse I’d like to design a modern pianoforte for the home and/or small concert hall. I can envision a long (say, 250 to 300 cm, or 8’ 2.5’ to 9’ 10” in length) and slender (say, 135 cm, or 53” in width across the front of the rim) instrument having a long but low to medium tension scale (i.e., one using relatively thin wire). Coupled with hammers of moderate mass and high resiliency the tonal dynamics would be to die for. It would be subtle and warm yet, believe me, it would be loud enough, when called on, to blast most reasonable folks out of the room. An, as an added bonus, the action would be light and quick due to the reduced reciprocating mass of the action system — less leading would be required to balance out the hammers. Demanding, perhaps, but infinitely rewarding.

High tension scales tie the soundboard system down, making it more difficult to develop a nice solid fundamental. The power is there, obviously, but it is concentrated in the upper harmonics. This is accentuated by the more dense and harder hammers required to drive those (usually) more massive and higher tension strings.

This whole issue has been an evolutionary process for me. My early studies (and studies) into scale evaluation and design were based on what has gone before. There is little written material that really evaluates the tonal consequences of scale tension schemes — we’ve kind of had to work things out as we go along. So, most of my early work was done with the higher tension scales typically found in our older pianos. (Once drawn wire was introduced, and the gray iron plate was developed to support the higher tensions possible with this wire, piano designers went crazy!) Over the years I’ve been taking another look at lower tension scaling and, increasingly, I’m liking what I’m hearing.

Del

Del,

I recently discussed my first experience going to a concert played on a fortepiano at the University of Washington. I was pretty blown away by the instrument and lamented about how loud the modern piano has become. I found the fortepiano concert to be one of the most gratifying musical experiences that I have been to. These instruments seem so much more personal than a modern piano, and so much more expressive. I especially like the sound produced by the leather hammers, kind of zingy. What was most interesting to me was that you could hear the pianist breathing while he played. I really felt like I was sharing the experience of the music with the performer, as opposed to observing a musician pounding away at a giant black beast.

I have thought alot since then about the potential of pianos to return to their smaller simpler roots. It seems to me if you got rid of the big tight strings you could eliminate the massive iron frame and alot of the supporting wood work. Maybe the frame would be steel or some type of unified body construction using steel extrusions. Who knows. I like the reliability of the modern piano and I think you could easily keep those good improvements while returning the sound and structure back to its genesis a little bit. I'd buy one.

Kirk

Del,
Great post Del. Chopin would agree I think, he said once that real music can't be made outside of the drawing room. Liszt might have felt differently though.

I think there are other reasons why we have gone to a more brilliant sound in this culture, and it's not just limited to pianos, but that's another discussion

Dan

Thank you, Del for the free seminar. Very interesting to read what you are thinking. I loved the sound of the Walter grand, although I wound up buying an Estonia. Guess that makes me a low tension guy.

Del....what an asset to the Piano Forum.

There is a point at which the tension becomes so low that the volume really falls off. You can hear it near the break in pianos which have the hockey-stick shaped bridge. There seems to be a loss of harmonics there as well. Maybe you can design that out, but there must be qualities that are inherent to higher tensions. I guess there must be some limits.

Hi,
What's the deal with the hocky stick type bridges? I recall Del saying something about the Walter not having it, I think because of the logrithmic scaling. Why would a piano have or not have it, is it generally a good thing or not, and what effect does it have?

Dan

Well, it foreshortens the scale near the break. You can make the rest of the piano relatively longer. However, I find that there is a dramatic fall-off in the tension of the strings there, unless you change the gauge of the wires often. Which I have been doing lately.

Grotriman brought this topic over to the Tuner's area, and some rumored data has been posted there. Unfortunately, it can't be correct. Also, some rough measurements I made on a couple of concert grands (Steinway D, Yamaha CFIIIs) indicate that the tension on them may not be so high either. What I can see is that where manufacturers have paid some attention to scale tension (many did not), there is a desire to have most of the treble at about 160 lb. Shorter pianos tend to have the tension drop off in the tenor, longer pianos tend to have it increase, as one would expect.

BDB's point is well taken. This is one of the major shortcomings (no pun intended) of the traditional scaling techniques. One of the most glaring offenders, simply because it is so well known, is the Steinway B. It has an average tension in the 155 – 165 lb range through the middle tenor section but from B#-26 (with 155 lbs.) it drops down to 115 – 120 lbs at F-21. Regardless the voicer’s skill this is going to be audible. It can be made less bad and you can grow accustomed to it, but it will never be good. (Well, it can be made good but it requires the installation of a transition bridge.)

The problem, of course, is that the designers of these scales felt it necessary to place the bass/tenor transition relatively far down in the scale. As to exactly why they believed this was necessary is unclear. It was probably a holdover from the harpsichord and fortepiano days and their brass bass strings though this is just speculation on my part. As may be, the early piano scale designers were quite willing to develop a more-or-less log scale pattern down to the mid-tenor (higher or lower, depending on the length or height of the piano) but then found it necessary to foreshorten the scale to avoid running into the back, or bottom, of the piano. They seem to have given little thought to possibly extending the bass section up a bit.

Using a very broad sweep of the scaling brush pianos below six feet or so will work best with 30 – 32 unisons of wrapped strings. Pianos in the six foot to seven foot range will work best with 25 – 32 unisons of wrapped strings. Pianos above seven foot will best use 20 – 27 unisons of wrapped strings. All of these transition points will be dependent on the type of scaling desired, i.e., long or short and high tension or low tension.

The goal here is scaling uniformity. It is desirable that the strings, regardless of their type, develop approximately the same harmonic spectrum and power given a uniform hammer blow. You can (and usually should) drop a bit of unison tension when transitioning to the wrapped bichords, but it should be kept to no more than 10 – 20 percent. This is easily achieved by maintaining approximately uniform string tensions and this is best achieved through the use of a log scale and by keeping the length disparity between the last plain steel string unison and the first wrapped unison relatively small.

There are many factors that govern the physical transition between the plain steel strings and the wrapped bi-chord strings but, regardless the length of the piano, these can be worked out to maintain a reasonably uniform tonal transition between them. It can only be done, however, if the designer is free to place that transition where it is best suited for the overall length of the scale. This means maintaining a log sweep to the bridge down to the point where length becomes a problem and then making the transition regardless of tradition.

Del

One should never underestimate the cheapness of manufacturers. Wound strings are more expensive than plain strings!

It's ridiculous in some respects. Piano strings, even the wound ones, are dirt cheap compared to other stringed instruments.

That's speculation on my part!

I ran the B scale through a spreadsheet, and found for a uniform tension of 160 lb. per string, the note #21 speaking length would be about 67", which is not out of line for a 7' piano. Change it from #21 wire to #22, and it drops down to about 64".

What really shows up to me is that in order to get a uniform scale, you really should use half sizes of wire at the break, and you should change gauges often. Conversely, there's not much point in making a lot of changes of gauge in the high treble. Here's my confession: The last 5'-8" Knabe that I did ended up with the top 24 notes being all #14 wire! That meant that the entire top section and a few notes beyond were the same size wire. The results were good. It sounds fine and stays in tune well, even on being freshly strung. It allowed me to lower the tension in the middle of the piano, and got rid of most of the nasty inharmonics which you usually get in Knabes.

It may be speculation but it's not a far-fetched as some might think given today's economy.

When Samuel Wolfenden published the first edition of his book, "A Treatise on the Art of Pianoforte Construction" in 1916 World War One was raging and copper was at a premium. What little there was was going into land fill in France and Germany with little, if any, being available for piano bass strings. This, of course, drove up the price of copper world-wide, including the U.S. It created a large market for soft iron wire.

Del

Actually, with #21 wire and a target of 160 lbs. per string I come up with a string length of about 1630 mm, or 64.2" With a length of 1705 mm, or 67.1" I come up with 160 lbs. of tension using the original #20 wire.

In either case these are longer string lengths than I would want in any 7' piano I was designing.

Del

You're right. I wrote the wire gauges wrong. The Steinway chart has all these obsolete scales, and it gets confusing.

If notes #1 and #21 are approximately the same length and are the longest strings in the piano, with agraffes about 12" from the front of the piano, it looks like there would be about 12-15" available at the tail, which would provide a lot of room at the hitchpin end for the flexibility you want.

The reason I mentioned the 64" length for what I should have said was #21 wire was that is the maximum size on a D or CFIIIs. (The CFIIIs uses the same wire gauges as a D except for 4 notes of 13-1/2 at the top.) I would think it shows that it would be rather difficult to design a concert grand with enough mass in the strings to impart a lot of energy to the soundboard without upping the tension as you approach the break.

It depends, I should think, on the size and mass of the soundboard assembly.

Del

Well, let's see. The Steinway B is about 82.5" long. I don't have one handy but I'd guess the agraffe for A-1 is about 12.5" back from the front of the keybed. That leaves us with 70" to play with. Now, we're going to want at least 12" between the back of the rim and the leading bridge pin so that leaves us with about 58" for the strings. It sounds like the actual Steinway B A-1 speaking length of about 59.5" is about right. I'd sure not want the speaking length of B-21 to be any longer than that. It would sure complicate bass string scaling. Why not just put a transition bridge in there and make about the last seven unisons bi-chord wrapped strings? Which is, in fact, what we do with this scale.

Del

Well, I've got 9-footers this week: two Ds, the CFIIIs and a Bechstein E (only 17 bass notes!). I guess I actually see them more often than any 7-footers.

In my earlier days I had a M & H A-3 which had been rebuilt rather badly before I got it. It has 2 bichords at the tenor break. I was never happy with the plain strings there and eventually got wound strings for them and it was much better. So I can see your point. It's just that it should be possible to make a decent piano about 7 feet with 20 bass notes. Maybe not desirable, though.

What I really find is the problem with seat-of-the-pants scales is that the tension goes way up a couple of octaves below C-88, and remains so high that it is hard to get a smooth transition when the scale gets foreshortened nearing the break. More wound strings is a possibility. Lowering the tension higher up, and changing gauges as the scale is foreshortened is another.

I certainly like Del's idea for a concert fortepiano and agree that such an instrument might be better able to give us new and dramatic perspectives on old familiar works, which could also be played with smaller orchestras. These would be best suited to the classical piano concertos from Haydn and Clementi to Chopin. Del could even find some young piano bravo to highlight his piano with a series of recordings of the Beethoven sonatas or of the Mozart sonatas. Schubert would be a revelation on such a piano. Such an instrument would also be better suited to the baroque literature, considering as I do that Scarlatti and even Bach were looking forward to a true pianoforte and that so nany great baroque keyboard works have been rendered so well by many great pianists.

A pianoforte more nearly approaching 7 feet would be more acceptable to more people. Perhaps Del would consider something the length of a Steinway C with lower tension design, etc.

Assuming that Del could sell everything he made, and I think he probably could, such pianfortes would of course be quite expensive but it's also a matter of scale, the cost of make and build for how many units. And it's also a golden opportunity to consider alternative, lighter and stronger substitutes to cast iron plates, perhaps some carbonite composite ceramic that is poured into a mold and baked solid. I believe that such a material could be made as acoustically dead as the best iron plates.

Then I could see two men move a concert grand piano by merely picking it up as such a change would radically reduce the weight of the piano.

Certainly there are technical centers in universities and industry that would be willing to look into the feasibility of using some of these new materials and the cost of production might turn out to make the difference.

Of course Del probably knows as much about bellies as anyone; soundboards, ribs, bridges, the uses of aprons, etc. I wonder if he'd consider making these out of some other material as well, the goal being maximum tuning stability? I'm assuming he'd use the usual pinblock material, not clear how the action would differ (does he intend to go back to Viennese actions?).

I'll never complain about how expensive good grand pianos are again, if I ever have. I've just been finding out how much people are willing to spend on SUV's and trucks. Certainly if it could be kept reasonable, no more than $40,000 apiece, one of these pianos would be well worth it.

Del, tell us how you'd think such a piano would perform the music of Brahms, Debussy, Prokofiev, etc. How well would it do Rhapsody in Blue?

Actually, Pinblocks need reform as much as anything. I have wanted to ask Del for sometime if it wouldn't be possible, and highly desireable to go to a "screw stringer" type tuning method.

Well, Del, why do we continue to use pinblocks which can be such a hassal?

Pin blocks work just fine. There's an industry to support them, including lots of tuners who know how to work with them. Screw stringers required different skils that could be awkward if you aren't familiar with them. Strings need to be sized to a pretty exact length on them, for instance. For the end user, there would be no difference in performance, so there's no point in changing.

There are a lot of romantic about how some things that used to be used in old pianos might have been better than what we use now. The Viennese action is another one of them. Awkward to work on, inaccurate, clumsy, yet people think they were somehow better. Bösendorfer didn't even provide a let-off adjustment on theirs, so how could it possibly have been better?

The fact is that the modern piano, particularly the Steinway model which is so popular to bash around here, evolved from specific desires that people wanted. Sure, there are improvements that could be made. Evening the tension is one of them. Whether that means higher tension or lower tension, I'm less sure. Personally, I think that it is dependent on the size of the piano. But if you deviate too far, you lose something.

Still, if you want to make a revolution in piano manufacture, you aren't going to make it starting with a concert grand, or even a 7' grand, no matter what Stuart & Sons may think. You are more likely to do it starting with a good upright, like Charles Walter, or even Mason & Hamlin did years ago. There just isn't enough market for anything else.

Two questions: (1) Does the high-tension/ low-tension distinction also apply to uprights? Where would some common uprights (Boston, Schulze-Pollmann, Grotrian, Kawai, M+H, etc.) fall on the tension spectrum? I recently purchased a S-P 126 as a first piano. (2) Has Steinway changed its tension design over the years, or between makes? I agree with what Del says re: large concert halls. I recently saw Andreas Schiff playing some Bach fugues on 9' Steinway with the top off at Lincoln Center - top performer, but didn't feel quite right to me. Also heard a large living-room recital on a 1930's Steinway M (5'7") (Handel piano duets with violin). Much more mellow and intimate: piano "fit" better with music, in my opinion. I don't know if it was just the ambiance of the room and how close we were to the performers, or if the older M was tuned differently. A lot of this is very subjective, of course, but others opinions would be interesting to me.

David,

You pose some interesting questions. The idea would be to produce a modern piano — that is, an instrument with the stability and reliability we have come to expect — but with a tonal performance and aesthetic that has long since been lost. Its style would be closer to the pianoforte of the 1830s to 1850s than to the modern Steinway-derived style. The action would be modern but set up for a generally lighter and quicker feel and response. In essence it would be what I think the pianoforte might have become if Steinway had not overwhelmed the industry with their manufacturing innovations.

Size.
Well, I’m already working on a 200 cm (6’ 7”) grand that incorporates a few of these ideas though it is intended to be somewhat closer to what we think of as the “modern” piano than this. I’m not locked into any particular size. The size should fit the need and the venue as well as the desire and taste of the performer. I’ve been toying with the idea of a longer piano because I’ve been thinking about the concert grand piano market. I’ve been discussing this market with a couple of people involved with various piano manufacturers and I’ve been wondering just why anyone would want to introduce yet another conventional (read Steinway D Clone). It makes no sense to me. Where is the market? The world is flooded with concert grand pianos that are not selling. Or the prestige everyone dreams about. Even if the thing were to outperform the Steinway D, By virtue of their C&A fleet Steinway owns the concert stages of the world and will for some years to come. Why bring out yet another concert piano to compete in an already saturated market? Why not develop another market niche entirely? A concert piano for the home? The small stage. Yes, make it long, give it the length to develop that solid, low frequency growl that comes only from a very long string yet make it musically pure and tonally matched to a smaller venue. Make it narrow and slender, exotic and elegant.

Now, certainly, this concept could be scaled down to a shorter scale. A pianoforte in the 225 cm (7’ 2”) could be built that would incorporate most of these attributes with only a moderate loss of low bass performance. Much less than that and something of the aesthetic balance I have in mind will be lost along with that very low bass performance. Which is not to denigrate the performance of the shorter piano at all; I have every reason to believe the low bass performance of our Model 200 will be excellent for its size. But it will not be the equal of what is possible with a even 225 cm (or longer) piano of modern design.

Cost.
As you say, the cost of the finished instrument is highly dependent on the economy of scale. I’ve no idea what the ultimate market for an instrument of this size and type might be. But let me speculate. Let’s assume production would be less than 60 instruments per year. This translates into five instruments per month. Assuming there will be good basic tooling and fixturing along with adequate heavy-duty woodworking machinery but very little in the way of automated equipment let’s further assume it will take about 400 hours of semi-skilled to skilled labor to produce each one. (That’s almost certainly an over-allowance but let’s leave it there for now.) This means a workforce of approximately 15 or 16 people. Labor (direct plus burden) in the U.S. is going to run about $12,000 to $15,000 per instrument. Depending on how much is built in-house rather than purchased, parts and materials costs could range from a low of, say $5,000 to a high of perhaps $10,000 (keeping in mind this is a very low production venture and the costs necessarily will reflect that). Now add to this the costs of overhead, return on investment, marketing and management, profit and so forth, a final price in the $40,000 to $50,000 range might be possible. That’s assuming they were all sold direct and not through dealers. If it proves impossible to maintain this sales rate and dealers become necessary then all this goes out the window and it becomes a whole new ballgame.

Technology.
It is quite probable that several new technologies could well transform the piano as we know it. However, developing these technologies and making them production ready and reliable will not be cheap. In my little scenario above there is no room for the extensive R&D work that would be required to bring either an exotic structure or an exotic soundboard assembly to market. That would introduce a whole new level of complexity and development costs. As described above the whole thing could be developed using existing technology and the end result would be quite predictable. Once new technologies like these get introduced a whole new level of complexity and uncertainty comes into the picture. Unless the project were funded with just huge amounts of money I wouldn’t even think of going that route. (I shouldn’t think even a university could take on this type of project without some outside funding. But I could be wrong — it’s been known.)

Performance.
The idea of a project such as this would be to carry forward the concepts of the early instruments and produce a modern equivalent of what these composers and performers actually played on. I see no contradiction between an instrument that would perform with equal charm both Prokofiev and Gershwin. Or, for that matter, Dave Brubeck or Oscar Peterson. John Cage is a whole other issue.

One of the advantages of a (relatively) small project like this should be versatility. Once the basic design and structure is in place there is no reason why it couldn’t fairly easily be altered to suit the needs and desires of an individual performer. That is, from a basic tone standard, a version could be made a bit brighter and more percussive or a bit warmer and sustaining. It would take only relatively minor changes to scaling and soundboard design. Actions could be easily altered to accommodate varying tastes in touch and feel.

So, why not do it?
Well, money, of course. Or the lack thereof. A project like this would cost a considerable amount to get started. Not as much as starting up another Baldwin or Steinway, but still a fair amount. I’ve never done even a basic business plan for something like this but I’m guessing it would cost closer to a million than a hundred thousand. Not all that much compared to what is spent on some of the high-tech startups we hear so much about (nearly all of which go belly up) but quite a bit more than I have at my disposal.

I think it would make a certain amount of sense for some existing piano manufacturer to pick up a project like this and develop it along side what they are already doing but I don’t see it happening any time soon. As an industry we don’t seem to be able to look beyond the basic Steinway and/or Old European archetype and conceive of what might be. And our vision of R&D seems limited primarily to manufacturing technologies. It would take a brave soul indeed to attempt developing a market niche of this type even if it would be one they totally owned for years to come.

Del

So I understand that the higher the bass section goes in the scale, the higher the tension. (?)

Also if there are wound strings in the lower tenor, there is a likelyhood the scale design is higher tension.

I hear lots of statments about certain pianos being low tension designs. But I don't hear of any particular piano that is a "high tension" design.

Sometimes I get the feeling that "low tension" design is (was once) a marketing tool to give the notion that a piano was not going to go out of tune so fast (being under less stress).

But where is a high tension design?

1] Not necessarily. Running the bass section up some can eliminate the hockey stick effect at the low end of the tenor bridge. This is unrelated to the overall scale tension.

2] Not necessarily. This is one way to cope with extending the tenor bridge down further than is actually called for by the tenor scaling. This can be done for several reasons. Once is placement of the bass/tenor plate bracing. In a four–section plate design it is generally desirable to keep this break fairly well down in the scale to more evenly distribute the string load across the plate.

3] Certain Kawai, Yamaha and M&H scales come to mind. (Please note: I am not saying there is anything necessarily ‘wrong’ with higher–tension scales. Pianos incorporating them have a certain tone quality that is appreciated by many. My own taste has been going in different directions. Yours may not.)

4] It can be a marketing ploy. But it is easy to check. It is also easy to confuse. Just where are we to measure? Treble? Tenor? Or bass? When I speak of a high–tension or a low–tension scale I am referring to the average scale tensions used through the tenor section of a piano, disregarding the hockey stick hook. I’ve categorized the general ranges in previous posts.)

As well, just like most piano design or construction ‘features,’ the proof is in the sound. If the piano has a sound quality you appreciate then what does it matter if the scale is HT or LT? One really shouldn’t be buying a piano because it has a HT scale or a LT scale. One should be buying a piano because of its tone quality and performance.

5] See No. 3.

Del

Well, yes and no. The conventional tuning pin and pinblock arrangement actually works quite well. While it's not perfect it is quite inexpensive, quite functional and lasts a really long time. I've not seen any other proposed alternative that comes close.

Yes, the screw stringer system has something to offer. But its high cost will probably preclude its ever being reintroduced. It would have to be redesigned and built with somewhat more precision. And, if we expect it to last anywhere close to as long as the conventional tuning pin/wood pinblock arrangement, it is going to have to be made of considerably stronger materials than the original. Both would drive up the cost considerably.

Del

Hi Del,

Why is it desirable to drop the tension of plain steel unisons before transitioning to wound strings?

Thanks!

Calin

It's not. It's the tension of the bi-chords that gets dropped. And it's the total unison tension that is under discussion, not the individual string tension.

And the amount that gets dropped depends largely on the amount of string length offset. In other words, how short the first wrapped string bi-chord unison is relative to the last tenor plain steel tri-chord unison.

Del

Hi Del,

What's the relationship between the length and the tension?
You use less tension for a longer bass string or more?

Calin

The greater the length offset — i.e., the larger the percentage of length differential with the wrapped strings always being shorter than the plain steel strings — between the wrapped strings and the plain strings the more the tension should be dropped. I have no mathematical formula for this, it just works out better this way in terms of tone quality and power blending.

Del

I think what Calin wanted was much simpler:

If you lengthen a string, and you want the pitch to stay the same, you have to increase the tension if the weight of the string remains the same.

Increasing the length lowers the pitch.
Increasing the tension raises the pitch.
Increasing the weight lowers the pitch. (We usually treat weight and diameter or gauge as equivalents, which is true for a material with uniform density. There's actually slight difference with wound strings, depending on how they are made. But then, there are other imperfections in the real world.)

I'm a little confused by the phrasing I added italic emphasis to above, Del. It may be the phrase "length differential" that I don't grasp as it implies a relationship and I'm not sure what's relative to what. My longest wound bass string is 77" from agraffe to bridgepin and my and my longest unwrapped tenor unisons are only 65". So when you say "wrapped strings always being shorter", I'm confused. Could you clarify, please?

By the way, I'm truly enjoying your ongoing "seminar" in piano design here in the forum. You may often wonder about the futility of educating us end users a little, but I'm sure I'm not alone in finding the portion I understand fascinating and the rest intriguing, and your generosity with time and thought here a wonderful gift to us.

I think he's talking about adjacent notes. If you switch from plain gauge wire to a heavier wound string, you need to shorten the speaking length to maintain a similar tension.

Yes, I'm talking about adjacent notes.

For example, the lowest tenor note (F-21, a plain-steel tri-chord) in a Steinway B scale is about 1470 mm long (±5 mm or so). The highest bass note (E-20, a copper-wrapped bi-chord) is about 1025 mm long (again ±5 mm or so).

This makes E-20 approximately 30% shorter than the adjacent F-21. Te make this string configuration blend the unison tension of the highest bass string is going to have to be somewhat less than it would be if it were scaled somewhat closer to, say, 10% shorter than the F-21 length. A length discrepancy of this magnitude will make both harmonic blending (i.e., controlling the harmonic content of the vibrating strings) and inharmonicity blending somewhat problematic. These scale breaks always present difficulties, this just makes them worse.

Del

A pleasure to read all of this very informative stuff. Thanks Del.

Perhaps in light of the (start of an) education we've just received on scales, you can comment on the A, B, C scale designs of Steinway in light of the bass to tenor transitions.

It is curious to me that (to my ears) the A and the C scale design are really nice. But these are the models they chose to discontinue in the US. The B scale perhaps being the most problematic is the model they chose to push (though I've heard very nice B-s).

Are there structural issues that make an A or a C difficult to produce?

FWIW the transitions on the L and M from bass to treble are very problematic for me. I assumed the pianos were just too short to have a good blend between the bass and tenor sections.

Thanks for the clarification, Del.

Hi Del!

Thanks for answering. If you drop the tension of the first bass string at the transition, you get more inharmonicity and maybe less power. Are you thereby trying to reach the same inharmonicity as the last plain wire note?

One thing I don't really get: why lower the tension more the shorter the bass string is? A short bass string already has more inharmonicity than a longer one. So it would make more sense to me to use lower tension if it is long (close to the length of the last plain wire note), to compensate for the fact that the plain wire string has more inharmonicity and the bass string less.
Am I understanding things the wrong way?

Calin

None of these pianos have what I would call a transparent transition.

The Model C comes off pretty well simply by virtue of its length. It has a relatively low-tension scale for a piano of its size. And it has a relatively small hook at the end of the tenor bridge. In fact, just looking at it there doesn’t seem to be any at all; it’s not until you plot it that it becomes apparent. (These comments refer to the old, NY version of the piano. I don't know what, if any, changes may have been made to the current German version.)

With the Model A it depends on which version you’re referring to. Some were less bad than others. There were three substantially different designs for this piano. The piano being built in Germany is, I think, a descendent of what we call the A-2, a 6’ 1”. It would benefit from a well-designed transition bridge.

There is no inherent reason why the bass/tenor transition cannot be, for all practical purposes, acoustically transparent even in a short piano. We’ve done this quite successfully in grand pianos as short as 4’ 5”. By comparison, pianos in the 5’ 7 ½” to 5’ 10 ½” pose little challenge. It’s just that it does take some creative design manipulation of the various string scale and soundboard assembly parameters. It is easier to accomplish this starting with fresh design but it is also possible through the careful redesign of existing product as well.

Del

The issue is not inharmonicity, it is the tonal, or acoustical, blend. We don’t really hear inharmonicity. Yes, it affects tuning but, on it’s own, we don’t hear it. What we do hear is the relative power and the harmonic mix in the acoustic waveform produced by the strings associated with adjacent notes. Of these, power is considerably less important than is the harmonic mix in the acoustic waveform. Which is why we can get by with the moderately lower unison tension (the individual string tensions are, of course, higher) on the excessively shortened wrapped bi-chord notes.

I’m not yet prepared to try to explain just why this tension drop is desirable. From experience I know it to be the case, from theory I’m not sure just why. I’m contemplating some experiments that should cast some enlightenment on the question.

Del

[highly desireable to go to a "screw stringer" type tuning method.

I'm sorry to dig backwards into the discussion, but I'm not sure what a "screw stringer" type of pin is and does. Larry Fine says they are machine screws, but how do they hold the tension of the string? Did any other manufacturer besides M&H use them? Thanks.

Thanks to Del and other experts for providing invaluable seminar like discussions on piano design.

I am interested in sustain because almost anything hit producing a sustaining tone sounds good. (Short sustaining instruments may sound good as well,like drums and certain makes of pianos.)

I'd like to know which are the dominant factors in determining sustain in descending order: The string tension, the strings themselves, the soundboard, the inner rim wood hardness, and the outer rim wood hardness.

I read C.C.Chang's piano webpage and I think that he mentioned that the initial fast decay was due to the vertical mode of motion of the strings while the ensuing singing tone with slower decay was due to the horizontal mode of string vibrations. The coupled motions of the strings also have to be such that there are no net twisting or shifting effect on the bridge thus two strings are only allowed out of phase motions horizontally and in-phase motions vertically. I'd like to hear from industry experts if these are accepted theory.

I also noticed that in tuning unisons, there seems to be a spot where the tone appears to be dead with fast decay. Moving slightly away from this point yields a much more pleasant tone to my ears. Some people referred them as in-phase or anti-phase coupled motions of the strings but from what I read from Chang's site anti-phase vertical motions of two strings are not allowed as well as the in-phase horizontal motions of two strings.

Also no matter what wood is used for the rim, to me they should be all hard enough compared to the vertical suppleness of the soundboard to cause near 100% reflection. Could energy loss due to radiation and friction be more dominant than the hardness of rim wood in determining sustain?

Sorry for all the technical questions not necessarily related to string tension but I work with electromagnetic waves and equations and can't help to draw some analogies in an attempt to understand more about pianos.

Screw stringers were the first Mason & Hamlins. They held the string by a metal hook, rather than a tuning pin. The hook had a screw thread at the other end, which went through a hole in a ridge in the plate. There was a nut on the other side of the hole, which was what you turned to tune the piano. One revolution of the nut was about a 24th of an inch at the end of the string, as opposed to over 3/4" on a tuning pin. This was slower during manufacture but easier to tune once you got it up to pitch. However, the hooks were held in position by a slotted piece of brass, like a comb, and the teeth would tend to break off, making them difficult to tune.

Hello Del, and thanks for sharing your knowledge!

Here are a few more questions:
How does lowering the tension influence the harmonic mix of a wound string?
I read that generally higher tension gives more poreminent harmonics in a string. Is this so?
When you are lowering the tension of the first bass string, are you trying to make it have more or less harmonics?
Is there a formula of calculating/predicting the harmonic structure of a string's tone (such as for inharmonicity)?

Calin

1] The dominant factor? There are several including the harmonic mix in the vibrating string. But the principle factor is the impedance relationship between the strings and the soundboard assembly. In very general terms, and assuming a given set of strings on a given piano:
-- A ‘light’ soundboard assembly will absorb high frequency energy at a faster rate than will a ‘heavy’ soundboard assembly.
-- A soundboard assembly with ‘low’ stiffness will absorb low frequency energy at a faster rate than will a ‘stiff’ soundboard assembly.

Generally speaking a piano with a ‘low-impedance’ soundboard assembly — i.e., relatively light and flexible — will have a powerful, percussive attack with shorter sustain. A piano with a ‘high-impedance’ soundboard assembly — i.e., relatively heavy and/or stiff — will have a less percussive attack and a longer sustain.

This characteristic can vary across the soundboard assembly. I.e., the soundboard assembly might exhibit high impedance characteristics in the bass and low tenor region, low impedance characteristics in the upper tenor/lower treble region and, again, high impedance characteristics in the upper treble. In fact, this scenario is not at all unusual and is found in many pianos. Especially as they age a bit.

2] This theory gained some popularity a few years back with an article that appeared in Scientific American magazine. There were several problems with the factual base of the article and several unsupportable assumptions made. (It’s been a few years and I’m not about to try to dig it out and study through the thing again. So please don’t as for specifics. If you’re interested it appeared about 20 years or so back.) Whether the conclusions of this original article are valid or not they only partially explain what happens within an individual unison; they do not explain the general tonal character of the piano. For that you have to consider #1, above.

3] I have not read Mr Chang’s web site so can’t comment specifically on what he writes. However, all of these theories consider only the specific unison being examined or tuned, not the overall function of the piano. For that you have to go back to #1, above.

4] The rim supports the soundboard assembly. The soundboard assembly is vibrating. Therefore energy is being felt at the parameter of the soundboard panel at the junction between it and the inner rim. If the inner rim (or the whole rim assembly, for that) is relatively light and flexible it will readily absorb energy from the vibrating soundboard assembly. This energy is, for the most part, lost within the rim assembly as heat. If the rim assembly is relatively dense and massive it will not readily absorb energy from the soundboard assembly. More energy will then remain in the soundboard assembly where it is converted into sound.

There noticeable, and often significant, difference in both the quality of tone and the useable sustain time between pianos with different types of rims. I have personally observed this difference between otherwise identical instruments side by side. Some years back Baldwin was attempting to save a few dollars and switched from hard maple to poplar in the rim assemblies of their three smaller grand pianos, the models M, R and L. The difference in performance was immediately noticeable and unacceptable. The tone in the pianos using poplar rims was conspicuously sharper and shorter. They switched back to hard maple for the inner rims which, in the opinion of some, was an acceptable compromise. I’m told they are now back to maple for both the inner and outer rims.

Yes, there are energy losses within the soundboard assembly. But this is also true within the rim assembly. Most of the energy coupled to the rim assembly is also lost due to the internal friction of the wood.

Del

Regarding Del's previous post:

First of all, thanks for your insightful and informative posts.

So, a point-of-information question: what do you mean by soundboard assembly? Is that the bridge crown, bridge, soundboard, and ribs?

Could you give your opinion as to choice in design of each of the elements for impedance characteristics? What issues are there with using different materials in each phase in terms of energy loss (impedance mismatch)? Also, if you're also including the ribs in your consideration, I'd appreciate your opinion on the use of different woods for the ribs in different sections of the soundboard.

I’ve gone into most of this in the not-so-distance past. At least I think I have. But to repeat:

Lower tension scales tend to develop more energy in the lower partials. Hence more potential energy in the lower harmonics of the sound wave envelope.

In general strings with relatively higher tension tend to develop more energy in the higher partials, less in the fundamental.

In developing a scale with generally lower tension I would be trying to emphasize the energy content in the fundamental and lower partials and, by extension, in the acoustic energy waveform.

There is no mathematical formula available to predict this that I am aware of.

Del

By ‘soundboard assembly’ I mean the assembly comprised of the soundboard panel, the ribs, the bridges, all of the incidental hardware such as screws, pins, etc., and how the system is assembled to form crown. In other words everything having to do with the soundboard and how it reacts to the strings.

As to the second part of your questions — I give all-day seminars on this subject and barely scratch the surface. (Which, by the way, you are welcome to join when and where they are presented.) The issues you raise are far more complex that I’m prepared to even attempt to answer within the context of this list. At least not all at once. Break them down into bite-sized pieces and I might be able to handle them over a period of weeks, perhaps months.

That, or wait for the book (which is what I should be working on right now).

Del

Hello Del!

Can you give us some details about your book?
When is it going to appear and maybe what chapters you plan on including?

Calin

I just thought I'd pop in & thank Del and others for their excellent explanations concerning elements of design that tend to mystify mere mortals, such as myself. The ability to articulate (as text) such complexities so others can understand the subject matter is a talent few possess .... Del, I hope you've had opportunities to mentor young techs .... You appear to me to be a very good teacher.

Of the similar pianos I was dealing with this week, the most pronounced hockey-stick hook was the Bechstein E, a very old-fashioned design. The Steinway Ds and the Yamaha CFIIIs were much less so, even though D and the E which are in the same building are probably contemporaries. I suspect that the Yamaha theoretically has the best design, but I think that there is a gap between the theory and the practice. After all, it was the first Steinway A that had the intermediate bridge. The later versions did away with it. I don't think Steinway would have removed it if they didn't feel that it was an improvement. It just shows that there are different tastes in these things.

I also spent a lot of time with a Steinway B. Not much of a hook there, either.

I would like to echo Katie's sentiments. On topics such as this, when Del, BDB and the rest of you smart folks start dialoging, I invariably get educated.

I appreciate the time and effort required to pen clear and concise answers to tough questions on a highly complex subject. Speaking for myself at least, I assure you it is not an easy task to do well.

Del, please do write that book some day. Undoubtedly, it will be of great interest to all that love pianos and an enduring legacy that will remain useful and relevant for a very long time to come.

1] It’s been a while since I’ve analyzed a Yamaha CF, but the last one I took a serious look at was fundamentally a Steinway D. Cleaned up in a few places, but still . . . .

2] Yes, the original Model A used a transition bridge that was later removed. But — and this is a really big ‘but’ — it transitioned from tri-chord plain steel strings to MUCH shorter tri-chord wrapped strings. About the worst transition you can make from a scaling standpoint. When properly designed and scaled the transition bridge is a very effective method of blending the low tenor with the bass. Effective enough that I’m coming to almost prefer it in shorter pianos.

3] Visually, your observation is correct — there doesn't appear to be much hook to the low tenor bridge of the Model B. But from a scaling standpoint the hook is both evident and tonally significant. The scale is pretty good down to about F-33 and then begins to shorten up significantly. If it maintained essentially the same length multiplier on down the lowest tenor strings (F-21) would be about 1850 mm (72.8 mm) long. As it is F-21 has a speaking length of approximately 1475 mm (58.1”). The hook is there, it’s just masked by the length of the piano.

Del

If only so you don't have to spend so much time responding to our requests for more information!

BTW - Where are your seminars given?

Thanks.

It was prompted by a little book I pulled, wet and mangled, out of a trash pile at a company I used to work for. Titled “Piano Scale Making” it was published in 1927 and in about 80 pages of sparse, but mostly concise prose, it explains how to ‘design’ and build a 5-foot piano. “A sample of which has been built and pronounced good.”

Now, I’m a sucker for old books — especially old books about pianos — so I rescued the poor thing. As I gently separated the old, yellowed and disintegrating pages I became increasingly intrigued with the piano being described. I had the growing impression that I’d rebuilt one just like it. It represented the culmination of the simplification and homogenization that had been taking place over a period of some ten to twenty years within the piano industry leading up to mid-1920s.

I was particularly interested because I have long considered the very small piano to be one of the greatest challenges facing the piano industry. And one that has not been well met. We have lots of cheap very small pianos (we did in 1927 as well — it’s not a new phenomena) but we have no really good very small pianos.

When one considers the challenges of piano design it is common to immediately consider the king of instruments; the concert grand. But, by comparison it is easy to design a large piano. There is so much forgiveness in those great, long strings. It is, by comparison, much more difficult to design a truly musical very small piano. (By my definition this means a piano 160 cm [5’3”] or shorter.) It is difficult enough that to date no one has done it. Oh, there are very small pianos that are less bad than others but none that are really great musical instruments. Still, it is possible. At least within the context of the musical scale normally used by good pianists. We have come very close through extensive remanufacturing but there are limitations to what can be done with existing design.

Unfortunately it seems to be easier to buy up instant heritage and move production to regions of ever-lower labor costs than to aggressively investigate and develop the fundamental design of the piano. So. I decided to tackle the subject of piano design from the perspective of the very small piano. It is here that we encounter all of the challenges of piano design. A tiny design error in a large piano becomes a major issue in the very small piano. And, because piano design principles remain the same regardless of size what I write about in describing and explaining the design and construction of the very small piano is also applicable to the larger piano.

I am currently about 2/3rds to 3/4ths of the way through the book. It covers most everything I can think of from developing the shape of the rim and developing a string scale to designing the plate. It is not intended to be a step-by-step instruction book but a treatise on both how and why things are done. I am trying to pass on a practical basis for understanding how the piano works more than a tutorial on piano design. The book is intended more for the piano technician and interested piano amateur than for the university piano researcher. At least this is my goal.
Del

Mostly at various Piano Technicians Guild conferences and conventions. But I have also done a number of one-day events organized by individual PTG chapters. I've also done seminars in Europe, Asia and Australia. (I'm always open to suggestions.)

These are generally promoted primarily to piano technicians but they are always open to anyone (including non-technicians) who hears about them and wants to attend. And who is willing to pay the registration fee which varies depending on the type of event. Registration at the annual PTG convention is several hundred dollars whereas a one-day seminar is generally $50 to $75. (Depending on projected attendance.) Sometimes including lunch.

Del

Del,

I understand your point that no one met the challenge of "musical" at 5'3" or under, but would be curious what make or model you think came closest--how close did they come to musically acceptible? What did they do right, and, more interestingly, where did they made their mistake or compromise--or what do you think they should have explored and experimented with to even out their design?

There are a lot of questions up there. Way more than I have time or space to answer here — besides, that is what the book is all about.

I haven't yet come across any very small grands that I've thought really fit the criteria. Even in so-called high-end piano lines the very small piano has been considered an afterthought. An instrument no real musician would purchase and use. Yet increasingly good pianists are finding it necessary to look for smaller instruments as homes become smaller and other stuff intrudes.

Del

Then I'll just look forward to your book. Keep us updated on its progress toward release and about the distibutor. There doesn't seem to be much on-line nor in book form discussing issues of design differences between makes and how they benefit/detract/compromise in any kind of systematic and meaningful way.

An obscure book indeed! I was able to find it in the Library of Congress catalog, but that's about it.

I wonder if the piano described was like one that I restrung some years ago. That particular one was an Everett, with 32 bass notes, although I later saw an identical piano marked Story & Clark. It stood out because you could see exactly where the wire gauges changed by the notching in the bridge, which to me, is as it should be. It was a fine little piano.

Not exactly but probably close. This instrument had a 30 note bass section with three sections overall. It was 4' 11 3/8" long (the lid would have a 5/8" overhang bringing it up to an even 5'). I expect its intended market was just a bit below either the Everett or the Story & Clark.

The design is very generic but functional. And it has most of the design weaknesses of most of the small pianos of the era. Which makes it a good starting point for discussion.

Del

Del,

A belated thank-you for explaining so patiently to my naive questions on sustain. It is people like you who make this forum so educational and valuable --- from parents who simply want to buy a starter piano for their children to professional pianists to people like me who are just intrigued by the magnificent instrument. I am so glad that a book is in the works and am looking forward to its publication.

BDB,

I found the quote I was thinking of over the weekend. On page 23 of “A Treatise on the Ari of Pianoforte Construction” (1916 edition), Samuel Wolfenden writes:
“The thickness of the covered strings below this point [the highest wrapped string on the bass bridge] often depends upon the number of pounds of copper the maker can afford to use.”

With the price of copper now considerably lower than it was during WW I labor makes up a much higher proportion of the cost of wrapped strings and they are more likely priced on a per-string basis rather than on the actual amount of copper used. But still....

Del

It's like steel pillars in automobile design. Whether the pillars at the corners of the top were wider "for more strength" or narrower "for better vision" depended on the relative price of glass versus steel.

I was intrigued by the information about longitudinal resonance of bass strings in Five Lectures on the Acoustics of the Piano. Mr. Conklin had criteria for bass string design which his recordings backed up as making an audible difference.