[CAUT] Sostenuto

[Fred Sturm]

On Feb 12, 2008, at 2:03 PM, Douglas Wood wrote:

> Today I have time to reply here. I'd like to suggest a quick set of  
> tests to assess the sostenuto that have worked well for me. I was  
> under the impression that this was what Kent asked about at the  
> beginning.

My impression was that Kent was wanting to go beyond the standard  
issue parameters, as one of his questions was whether there should be  
a point before full engagement of the sost pedal where tabs might  
cheat. Your checklist of tests is a good one, and well put, Doug. My  
own responses have been a bit complicated and perhaps confusing,  
because I think it is possible to evaluate in a more refined way, but  
only with a somewhat different approach, which includes the way in  
which one sets it up to begin with.
	It's not all that complicated to adjust the sostenuto rod itself: up/ 
down, in/out, and rest/full. What makes it complicated is that you  
have to assume certain things about the damper adjustment (tabs in  
line, degree of timing relative to the keys, upstop rail, etc). But  
assuming you have the dampers set up within reasonable parameters, I  
think that adjusting in/out visually, using a light shown directly on  
the tab/rod intersection, is definitely the best place to begin. WIth  
that very solid knowledge, you can proceed to adjust up/down, and the  
tests I mentioned with the two pedals are a good way to do that on the  
S&S (where you can't see up/down directly).
	I described use of the pedal in actual playing to try to give a  
flavor for what the performer is experiencing. What I described isn't  
really all that extreme. If one is using the sostenuto in passage  
work, the norm is that one is creating "harmonic washes" where a  
certain "pedal point" (the bass note(s) held by the pedal will be  
succeeded by another. That succession has to happen seamlessly. This  
is a big problem from a performer's point of view. One would really  
prefer to release the sostenuto a wee bit ahead of the sustain, but  
they have to be released simultaneously. Thus, you have this momentary  
"stop in the action." With normal sustain pedaling, you can time your  
pedaling to cross those harmonic divides very subtilely, holding the  
sustain pedal until the next note has just been played, or whatever  
precise timing you prefer. If you want to catch a new chord with the  
sostenuto, you have to do a complete release of both pedals, play the  
new chord, then depress sostenuto, then sustain. This takes time, but  
it can't sound to the audience like it takes time. And there isn't a  
lot of room to fudge.
	So, bottom line, to make it easier for the performer, you want the  
sostenuto to catch as low/early as you can consistent with a safety  
factor. Maybe it even catches before the hammer has quite hit the  
string in a real time setup. The performer is going to cut it as close  
as possible. At any rate, that is what I was trying to convey.

Regards,
Fred Sturm
University of New Mexico
fssturm at unm.edu

> They will establish the necessary function of the sostenuto. There  
> is probably room for improvement for someone who is pushing the  
> sostenuto function to its limits, but I've never had that come up.  
> That may be only because the most frequent question I get is "Why  
> doesn't that middle pedal do anything?" Sounds like Fred, at least,  
> has had some experience with using the sostenuto near its limits.  
> Sort of like Uchida and her quarter pedaling with the sustain...
>
> In any case, here are the tests:
>
> 1. Depress sustain pedal with normal pressure, all the way. While  
> holding it down, depress the sostenuto pedal. All dampers should  
> stay where they are or lift a little. The latter is required for the  
> older fixed-tab systems. Release the sustain pedal--no dampers  
> should move.
>
> 2. Slowly release the sostenuto pedal. The dampers should "fall like  
> rain" all across the scale as Eric S. so nicely puts it.
>
> 3. Test each note with the key. I usually play octaves in each  
> hand--4 notes at a time--with 2 octaves between thumbs. Move slowly  
> enough to be SURE that all 4 dampers engage with the sostenuto--look  
> and listen. This test is necessary because the key end felt and  
> damper undelever bottoms are not necessarily in straight lines.
>
> 4. Release all keys and pedals. Press the sostenuto firmly. Bang  
> every note with a damper--none should hang up.
>
> 5. Release everything. Then work the sostenuto pedal alone gently  
> and vigorously. Listen for any excess noise and watch the damper  
> heads for movement. There should be only the noise of the bushings  
> on pedal rod, sostenuto dowel and monkey. If these are excessive,  
> lubrication may be in order.
>
> If the piano fails any of the tests, then the items below become  
> very helpful. I would really like to hear about any performance  
> issues with a sostenuto that passes these tests. I do understand  
> that it is quite possible, but it would help us all to know how  
> reliable these fairly simple and quick tests are in performance  
> situations.
>
> I would like to highlight one other thing. The sostenuto rail in  
> most NY, and some Hamburg, Steinways is mounted on the action, as we  
> all know. The damper system is mounted on the crossblock. The action  
> and the crossblock can move relative to each other with humidity  
> changes. Unfortunately, I live in the Pacific Northwest, so I have  
> very little data on just how much they move. But repeating these  
> tests with large humidity (and concomitant tuning) changes is  
> probably wise. It is my understanding that these humidity-related  
> changes are the reason for the space between the tabs and the rod to  
> be a bit larger than would otherwise be desirable.
>
> Doug
>
>
> On Feb 6, 2008, at 7:40 AM, Kent Swafford wrote:
>
>> Here are procedures for making certain that the sostenuto mechanism  
>> is ready for concert use, edited together from comments by myself,  
>> Jim Busby, Jon Page, and Fred Sturm. This is put together for my  
>> benefit, so that when I forget all this within a few months, I will  
>> have this material to review. Thanks!
>>
>> Dampers must be in good regulation, and all dampers must seat  
>> properly. The damper upstop rail must be regulated perfectly.
>>
>> [FS] With flashlight well placed, looking at in/out position of the  
>> sostenuto rod, but also at whether the sostenuto pedal is well  
>> adjusted, with damper pedal at rest, depress the sostenuto pedal  
>> fully. Is the blade horizontal or nearly so? How much does it  
>> overlap the tabs? Get a second perspective by first depressing the  
>> damper pedal, then the sostenuto pedal. How much does the blade  
>> overlap the tabs?. Does it raise the dampers the same as the damper  
>> pedal? more? less? This can lead you to whether you need to take up  
>> lost motion, adjust the stop felt, and tell you about in/out  
>> position.
>>
>> [FS] The Steinway rod can be adjusted with action in to a large  
>> extent, except raising the rod. You can usually get at the screws  
>> and loosen them enough to jar the rod forward or backward a bit,  
>> then tighten and try again. And you can tap downwards to bend the  
>> brass brackets to lower the rod.
>>
>> [FS] Raise the dampers (press the pedal) just enough so that the  
>> tabs are even with the sostenuto rod at rest, IOW even with the  
>> round part of the rod. Get your eye vertical over the gap between  
>> the tabs and rod and look at the gap (move the flashlight from  
>> section to section). Is the gap sufficient (so that there won't be  
>> occasional "hangers"), but minimal, about 1-1.5 mm? Is the gap  
>> even? This tells you a lot about the in/out position of the rod,  
>> and about the evenness of the line of tabs in that plane. Tabs that  
>> are obviously out of line can be slightly adjusted from above  
>> (without removing the stack), with judicious use of a slim bladed  
>> screwdriver, pressed against the bottom of the wire/top of the top  
>> flange.
>>
>> [FS] It's a good idea to note whether dampers are lifted the same  
>> from the keys as from the pedal. If there is late timing, and the  
>> dampers lift less from the keys, you may run into trouble of  
>> dampers not catching consistently.
>>
>> [JB] Check the obvious “normal” use of the pedal. Play groups,  
>> chords, etc. You can see if they lift too far, and if it works well  
>> here, that’s a start.
>>
>> Depress damper pedal. While continuing to hold the damper pedal,  
>> depress the sostenuto pedal. While continuing to hold down the  
>> sostenuto pedal down, release the damper pedal. All dampers should  
>> remain raised. (This is a more severe test than depressing each  
>> key, then depressing the sostenuto pedal, and releasing the key to  
>> see that the damper stays raised.)
>>
>> [JB] Eric S. says that for this first test it’s important to  
>> release the dampers slowly (after being held by the sos pedal) and  
>> the dampers should “fall like rain” randomly from every section. If  
>> most of one section falls before another you have a problem.
>>
>> Depess sostenuto pedal. While continuing to hold the sostenuto  
>> pedal down, depress the damper pedal. While continuing to hold the  
>> damper pedal down, release the sostenuto pedal. The immediate  
>> flipping of tabs should make a noticeable sound as the blade  
>> returns to its rest position. (If the blade is held too tightly in  
>> its mounting brackets, there might be an inappropriate delay in the  
>> return of the blade to its rest position. The tabs should not be  
>> able to keep the blade from returning to its rest position.)
>>
>> Depress and release the damper pedal. Depress the sostenuto pedal;  
>> no dampers should rise. In addition, there should be no noise  
>> resulting from slight contact between the blades and the tabs of  
>> any dampers. If noise is heard from the blade contacting tabs as  
>> the sostenuto pedal is depressed, you may need to rest your finger  
>> on each damper in succession while depressing the sostenuto pedal  
>> to find the damper tabs that are makding contact with the blade.
>>
>> [JP] No downwards pressure here; lightly rest your finger so as not  
>> to press the tab lower.
>>
>> [FS] You can learn more. Press the damper pedal a very slight bit,  
>> and work the sostenuto pedal. FInd out how much you need to lift  
>> the dampers with the damper pedal before the blade starts catching  
>> against tabs. There should be a little bit of safety factor, but  
>> very little. Then work backwards from the top. Damper pedal fully  
>> depressed, press the sostenuto. It should catch all. Damper pedal  
>> not quite fully depressed, press the sostenuto pedal. FInd how much  
>> safety factor there is in this direction. In any of these tests,  
>> you don't want to jam on the sostenuto pedal, just a light touch to  
>> see when and where you get rubbing contact, or from above, when  
>> there starts to be jamming. This can also tell you what sections  
>> might be different from others (which sections catch first from  
>> below, which from above), all of which tells you where you might  
>> want to raise or lower the rod. And it tells you how even the tab  
>> line is horizontally, if there are several that behave markedly  
>> differently from the others (catch much earlier, while the others  
>> are missed).
>>
>> Depress and continue to hold down the sostenuto pedal. Deliver a  
>> double-forte blow to every note that has a damper. All dampers must  
>> return to their rest positions on the strings as each note is  
>> released. (This is a particularly severe test. This test may fail  
>> if the sostenuto pedal is not depressed as far as is possible.)
>>
>> [JP]  depress the note and and lift the damper head with your other  
>> hand to see if you can force the tab to 'cheat'.
>>
>> Pull straight up on the damper. If you flex the damper wire, you'll  
>> get false positives and/or false negatives.
>>
>> [JP] I have found that a visual inspection of blade insertion onto  
>> the tabs at the ends and aligning them with a straight edge (fore&  
>> aft / up & down) precludes slipping tabs. A light source from a  
>> flexible goose neck works well.
>>
>>
>>
>>
>>
>>
>>
>
>