Ah, my favorite topic!
Installing a soundboard is very labor intensive. For too long, my normal procedure was to merely copy the original. Then came the Heintzman grand. After installation I did a standard tap test and was sorely disappointed. It just didn't have the liveliness that a new board should have. Thank goodness I recorded the info. The board had a 3/8"+ thick panel, and the ribs had what I now call a high profile (7/8" x 7/8") throughout. The result is an uneven section modulus and an uneven proportioning.
(Blue line is section modulus curve, red line is rib transition smoothness, treble on the left)
Thus, and so far, a three year research project began. I have been collecting soundboard data, listening to pianos, and observing rib scales and related problems etc etc. At first, one would naturally think that all ribs are the same and don't do much, therefore dimensions don't matter. But, just like string scales, there are many decisions to be made when it comes to a rib scale. Many different rib scale plans.
Steinway uses the plan of an 80%+ rib profile, 7 belly rail ribs ( the rest are spine ribs), with the top belly rail ribs getting smaller and narrower in an inconsistent fashion ( why widen the fields in the treble?). On the rare occasion when I come across a Steinway that sounds fuller and more alive than most others, the rib scale profile (so far) has been between 71% -76%. But they always have poor rib scale transitions.
Here's my latest rebuild a Steinway A pre 1900 original scale.
Now, when you have a rib scale this horrible, you need riblets, you'll have dead spots, etc etc. To answer Jim, my belief is that some ribs are worker harder than others, when they all should be working as a team.
For comparison, here is the Steinway A rib scale modified.
This Steinway A went back to the Nashville technician, who has sworn to give me an honest feedback. Good or Bad. I am eagerly waiting.
Yes Jim (LOL) I will be producing recordings. I already have two professional pianists that will gladly perform, lined up.
Here are some fundamentals that I have settled on that give good and consistent results and deliver a sound quality that I like.
The rib count should equal the panel square feet. The sum of the width of the ribs should also equal the panel square feet.
12 square feet panel
Rib widths equal 12" ( Thus 1" wide ribs)
This does a few things. Creates evenness across the scale, provides smooth transitions, smooth section modulus curves. Allows for a 70% rib profile by adjusting heights and maintaining the section modulus TARGET. Allows for even and smooth Volume to length curves. Plus, it's simpler to deal with one parameter than juggle two.
A characteristic that I have noticed in the last few boards since adapting this method is a surprising carrying power. The piano seems to be more powerful the farther away you get from the piano. The Mason and Hamlin BB I recently did seemed to ROAR when I stepped across the room. I've heard many different pianos, never heard one do that, like that before. Certainly was motivating for me.
The ultimate goal of course was to come up with a product that impresses the clients, so the business will grow. Time will tell.
OK I understand the derivation of section modulus but what do mean by "rib transition" and how are you calculating that. BTW here's a graph of a 17 rib board, section modulus, the widths, as you can see below, are not equal. (bass on the left)
***Long Post***Well I did say "allows", not "only" LOL.
Comments are regularly made by many regarding problems exhibited by old boards. In this case the problem octaves in Steinways.
The cause, it seems to me, is blamed incorrectly on a method of construction, namely the hygroscopic compression crown. Thus the proposed solution is to replace that method with something else. That usually results in laminating ribs, adding more ribs, i-beam ribs, narrow and tall ribs, and uneven rib layouts. I think all of these plans just introduce other types of problems. Namely, and especially in Steinways, a different un-steinway sound. I like the description that it creates a thin sound (too much stiffness).Players use to describe some Steinways and other by-gone pianos as having a fat sound.
My belief is that the root cause of the problems in compression crown boards has never been addressed. I am simply making an observation, based on analysis of numerous boards and constitutes as an opinion. I have seen some common traits such as: uneven rib transitions, incorrectly sized ribs for that location, poor rib volume distribution, wrong rib count, and as mentioned before, making the ribs narrow in the treble.
I am also collecting data that supports my case, for example:
I played and studied a REED and Sons Upright that was a 100 years old with a rib structure similar to what I am talking about.
It had a nice fat sound, was in fantastic condition with no cracks and good downbearing. I did not notice any problems in the treble such as dead spots. There were some false beats.
There were minor design flaws, but nothing horrible. Here's the data.Clearly suppose to be all 1" widths, but human imperfection as it is.Length and Volume a bit out of proportion.Height and widths uneven.Section modulus and Delta curves uneven and rough.For a fun exercise, here's how i would clean it up and make it more precise and efficient.Section Modulus and delta curves silky smooth.Length and volume curves in proportionHeight and width smooth and even.-chris
This is where we have to wait for a decent recording to see what various sounds you are valuing in your aesthetic actually sound like. For instance, in my book, the "roaring" of your BB, could possibly be, in my take on it, a "booming" sound...which I find highly objectionable. I remember a recording of a Marc Allen piano a couple of years ago, that had that booming bass. The description in the journal of how he achieved it was interesting. But...that booming bass was large and ungainly, and somewhat out of control...big, but out of control. I'm not saying this is the case, and perhaps for you "booming" means high fundamental which pulses one's solar plexus. Words are not helpful past a certain point, and I think we have progressed past this point.Same with the "fat" sound. "fat" is what I refer to as an old Steinway structural inability to create a pitch focus, particularly in the tenor. Thick textures are indistinct and registration undifferentiated. I absolutely hate that sound, and take measures to make sure that my pianos never make this sound...hence my tendency to create a more controlled sound. Decisions are aesthetically driven. But once again, these are just words, and "fat" may mean something entirely different to the two of us...so I look forward to a sound sample, because I think you are making some interesting points.Re the pretty graph lines, I have lost my affinity for these graphs, as I have seen wonderful sounding string scales that look like a saw-tooth that had ADD...all over the place. The lack of prettiness in the lines did not translate to experience of the sound, which was quite nice. That's not to say the lines are useless. But, I have gone through great trouble to create designs that looked smooth on paper, only to find that yeah...yadayada...it sounded like a piano, like all the other less beautifully smoothed designs.I know you are working on the sound aspect, so I look forward to that bit of data..
Narrow/tall ribs is dependant on stiffness with the idea of reducing weight. That's great for ceiling joists, but is counterintuitive to spring design. The stiffer a board is the quicker it burns up energy and dissipates it. A flexible board will take that same input of energy and slowly use it and release it. Thats increased efficiency and is easily demonstrated. 1" wide ribs assures that the proper mass is attained while engineering the most flexibility that a board can accept. Keeping the same amount of mass, while lowering the section modulus is a different twist. David Love in his rescale of my Weber did this. I'm still contemplating whether this is an advantage or a disadvantage. Maybe it just offers a different tonal palette. I like to compare boards to one another. Especially the same make and model. One parameter I use is what I call "The sum of Z". This is just adding up all the section modulus values. The original Weber was 1.72 (in3) with a mass of .76 in3 per inch. My approach is to stay fairly close to the original design with reputable makes. So my modification was to increase the Z (flexibility)slightly to 1.80 (in3) and lower the mass a little to .75 in3 per inch. The .75 in3 per inch is a reliable amount of mass based on study and observation of old boards and how they survived time. Love went the other way and stiffened the Z to 1.50 in3 and also lowering the mass to .75 in3 per inch. Since I heard the old board, saw how the old board did over time, i don't understand the justification for changing the Z to such an extreme. I recall now that I was criticized for under designing the Weber, but i actually increased its strength slightly over the original. It was ironic though, that Love and I both had changed the mass to .75 in3. But that is what makes this fun when other ideas come into the mix, because it gets you thinking in ways you wouldn't think before. The Z Delta is my favorite parameter. Delta measures the rate of increase or decrease in size from rib to rib.I find it very useful.
Small drum:The Smallest Drumkit in the World - maybe!
You can hear it yourself. The more flexible drum head (the timpani) sustains longer than the small drums (stiffness).This confirms my variable rib demo also, the stiffer the rib is, the faster it vibrates and shorter the sustain is. To move the rib fast takes up the input of energy quickly. When i make the rib flexible, it vibrates slower and longer with the same input of energy.It's the same rib.Now to be fair, my varible rib demo adjusts the flexibility/ stiffness by adding and removing weight. I think this is a fair representation of the 2 different rib profiles in action. A narrow tall rib will treat a downbearing weight lightly, and a short wide rib will treat that same weight heavily.If i'm missing something in the reasoning, i'm open to hearing why.I do think in the plurality, since there is a plurality of ribs. But for purposes of comparing one board to another, i have various parameters that i have reduced to a single number. This works in my system as its a system of isolating one factor, while doing everything else the same way. And leap frogging off the original. Thats why i don't mix talk