Are there conventions or standards for Unison detuning? Ranges?

Jake,
According to this Northern Sounds post:
http://www.northernsounds.com/forum/sho...ostcount=2
it appears that a grand in "concert state" ideally has no unison detuning whatsoever, if I'm interpreting this correctly.

However, the author also admits that "concert state" doesn't last very long at all. 

Maybe the ideal is never actually achieved in practice, though, and we have to set it to something more realistic in Pianoteq so that it doesn't sound too sterile?

I actually have a piano technician's reference book and I can't find anything about deliberately setting the unison detuning to a particular amount - just seems that you do your best to tune it exactly, which tallies with that post.

Greg.

I like the C3 "worn out," though I think it's a bit too worn out. LOL I'm guessing it's purposely outta tune somewhat? It's got a lot of character. The "normal" pianos are a bit too pristine. If they could have a preset somewhere in between utopia and worn out, I think we'd have a real winner! Or make "worn out" an adjustable parameter.

Is "worn out" simply C3 with a certain amount of unison detuning and hammer hardness? Or is it more than that?

So there *is* mention of slight detuning in at least one piano tuning manual.  So, perhaps "concert tune" does not *necessarily* mean that the unisons are exactly in tune afterall - it just means substantially in tune, but with the appropriate/normal amount of detuning (if that is desired)  In that case, we still don't have an answer for Jake's question. If one *was* going to intentionally detune, how is it done!

Greg.

Very slight amounts of detuning don't sound "detuned", though, do they? It makes the note evolve very slowly, without sounding at all "honky tonk".

Greg.

You are absolutely correct.  And "evolve" is a great description -- a 2-cents difference is tiny, IMHumO, but surely enough to enrich the texture (though the decay of the piano sound robs you of truly hearing what happens as the pitches beat in and out of phase with each other, very slowly). 

When I do my multi-piano "gamelan" (pianolan -- gameliano?), I tend to detune each _piano_ (and pan them all over the place) by significantly larger differences (10 to 50 cents, though smaller amounts are acceptable ;^), and then I'll pit them against each other at different octaves (or 4ths/5ths).  The resulting "shimmer" is outstanding, though perhaps a bit much for most ears.  (If I were to detune the unisons in this setup, then the result would be too messy.  Thus, I'm just tickled _pink_ that "Pro" allows you to detune the unisons on a per-note basis.  8^)

Uh-oh -- here comes another idea!  We need a "Pro For Crazy People" version, in which case you could do per-STRING editing.  Ooooooooooo -- I'd never leave the house, then...

;^)

This list of messages from a piano tech site has several brief stories from technicians who have been asked to detune the unisons. Most of the techs hate it, some do it just because of crazed customers, and a few say that it isn't unusual and have had it suggested by colleagues. Haven't found any that speak of specific ranges or amounts, though:

http://www.ptg.org/pipermail/pianotech/...html#49001

FWIW, I asked a piano tuner about this, and he too said that some tuners do detune, whereas others don't. (he personally does NOT do it, and thinks it is old fashioned)
He said that one should look more to the hammers for issues of tone. He said that on RARE occasions, two strings which are perfectly in tune will cancel each other out, and on these cases, they do have to be detuned slightly. But this is only for the offending notes - NOT the whole piano.

Yes, I hear the longer sustain in Pianoteq as the unison width is decreased. This contradicts that Northern Sounds post, btw.

Greg.

Some random comments of mine:

1.  Since there is no consensus amongst tuners, then it's highly unlikely that there will be a standard for detuning (deduced from Greg's comments).  To add to this, my rebuilder/technician friend says he always tunes unisons as closely as he can - this in response to my comment that Pianoteq talks about slight detuning).

2.  I'm wondering if detuning an acoustic piano has the same effect (faster decay).

3.  It's becoming apparent to me that the physics of the piano are far more complex than most of us have realized.

4.  There are many opinions which purport to be expert, but I'm starting to have doubts (I've not warmed to the northersounds site - probably because I don't think samples make any sense, and they are a site pushing samples).

5.  We must keep in mind that a physical modeling of the physics of a piano may not respond completely the way an acoustic piano does.  Individual adjustments may be similar, but when all of the parameters interact, the equation becomes extremely complex (jope raised an interesting possibility.

6.  It seems that we have much experimenting to do - waiting for Jake's continuing work on this.

Glenn

RE: that particular Northern Sounds post, keep in mind that it appears to have been authored by someone who was working directly with a Steinway technician.  For me, this gives it a lot of weight. (although I can't help wondering about the comment that unisons in tune result in LESS sustain - intuitively, Pianoteq's behaviour makes more sense to me)

Greg.

Yes, I know - that is Pianoteq's behaviour, but that is in direct contradiction to the Northern Sounds post - in that post, the author says that when the unisons are in tune, the sustain is *less*.

I'm saying that *intuitively*, Pianoteq's behaviour is what I would have expected.

Greg.

That's exactly my line of thinking too.

Greg.

There is no contradiction between a real acoustic piano and Pianoteq tuning behaviours. But the things are a bit tricky, I will try to explain it here. You may want to read first the following posts:
29-03-2008 18:54:43
http://www.forum-pianoteq.com/viewtopic...1548#p1548
26-09-2009 07:44:07
http://www.forum-pianoteq.com/viewtopic...6128#p6128

For simplicity, let’s suppose there are 2 strings and let's talk only about the fundamental frequency. The strings are coupled by the soundboard, so you shouldn’t think about them separately but together. There are 2 modes of vibration, lets call them D and R:
- D is the direct sound, it is a mode in which the strings are almost in phase,
- R is the remanent sound, it is a mode in which the strings are almost in opposite phase.

How do D and R behave with unison tuning? This is the core of Weinreich’s “coupled piano strings” article: the more in tune, the faster is D and the slower is R. D gets faster because the more the strings cooperate (being more and more in phase), the more efficient is the energy transmission to the soundboard, thus the faster is the sound decay (and also the louder is the sound). For the same reason, R gets slower because the more in opposite phase, the less energy is transmitted, thus the slower is the sound decay.

So, when you have both modes in the sound (aD+bR), what will you hear when reducing the unison width?
Case 1) If D is prominent (a/b is big), then the total sound will get shorter. This is particularly the case with very well regulated hammers because in that case, the hammer being very parallel to the strings, R is almost zero.
Case 2) If R is prominent (b/a is big), then the total sound will get longer!
As you see, both cases - and of course all intermediate situations - are possible when reducing the unison width.

In Pianoteq, you can control the ratio a/b with the direct sound duration slider: the more you push it to the right (case of perfectly parallel hammers), the greater is a/b. Hope that helps!

(When I said earlier that reducing the unison detuning while reducing impedance creates a bass without as much thump, I may have misspoken. In a sense, there's more thump--a heavy, thick sound as opposed to the sense of hitting a hard piece of wood. Less thock, more thump, I should have said...)

You're not alone.  I've tried playing with the "direct sound duration" before without _truly_ understanding what it does -- this particular iteration of the explanation does substantial justice to the concept.

Just be glad we can't futz with the deeper parameters behind this concept just yet -- you wanna talk about getting lost in a labyrinth...

Excellent explanation, Philippe!  Spoken like a true mathemusician.  ;^)

For a long time, I was confused by the Direct sound duration, since I thought that sliding it to the right would just increase the duration of the sound before it went to the soundboard. But it didn't. Then Philippe mentioned once that increasing it delayed the sound of the unisons getting detuned, which made me use it more for controlling the sound of the attack and early decay.

But I didn't until now see why it sometimes extended the length of the decay and sometimes abbreviated it.

I have uploaded the mp3 that was rendered in Pianoteq (wave from PT, then converted to mp3).

Entitled "Detune Test Rendered"

It is pretty much the same as the previous one.

Glenn

I owned a G2 from 1975 until 2004.  For the first two or three years it drove me nuts with seasonal pitch changes. There may have been unison differences, but they were probably masked by the overall change from low to mid to high that occurs when the soundboard swells and shrinks (with humidity rise, the centre of the soundboard moves the most, raising the midrange relative to the low and high ends).

After a few years the seasonal changes were not nearly as bad and after four or five years they were not noticeable at all.

Now for a parallel (if you can follow a bit of structural engineering).  Roof trusses are made from so-called kiln dried lumber (less than 19%).  During the winter in northern climates, the exterior air is very dry and doesn't hold much moisture, so the potential for drying shrinkage is high.  Ordinarily all member of the truss would shrink similarly, and the truss would be stable.

However in cold climates, while the top chord is above the ceiling insulation and remains cold, the bottom chord is at the bottom of the insulation, and is thus very warm.  Note that the humidity of the bottom chord is very low because although it is warmed by the warm interior ceiling, it is effectively "outside" where the air is very dry.

As a result, the bottom chord dries out much more than the bottom chord in the winter, and this causes the trusses to arch upward in the midspan (note - when the bottom chord shortens, the truss arches).

In the spring, moist air returns with higher temperatures, and the bottom chord regains moisture and the truss settles back down.

This wouldn't be a problem if there were no interior partitions in the building, but when the truss moves upwards by two inches or so (50 mm), it raises havoc with the partitions.

After two sometimes three) seasons of cycling, the trusses finally stabilize and there is no more movement.  The similarity to the piano soundboard is striking.  Incidentally, they are both made of spruce.

It is my contention that when young, wood will take on and release moisture with seasonal humidity changes, but in time this decreases and is not a problem.  I call this curing or aging, not drying.  The cell walls harden, and simply won't take up moisture.

Perhaps our illustrious Guillaume can make comments on piano soundboards.  Incidentally some researchers believe that the difference between the great violins (Stradivarius etc), and well made modern ones is that Strads used very old wood.

Glenn