Accu-just HPs in Steinway D :

[Fred Sturm]

Not to argue with whether it is possible to do a _well-tuned_ back
duplex with vertical hitch pins (I agree, it ain't gonna happen), but
Baldwin certainly tried to do an approximately tuned back duplex in all
accujust pianos I've seen. IOW, placement of the accujust pins is
designed to come up with back-tuning at approximate partials of the
speaking length. Every other unison has one "out-of-tune" for obvious
reasons stated by Ron. 
	How important this is to overall tonal output, I have no firm opinion.
It's like the whole question of tuned duplexes - does it really help to
try to tune them "right-on?" My general opinion is "probably not."
Probably better that they be somewhat out of phase. But I think tuned
duplexes are far better than a "random" back length, which tends to be
close to one length (as in a single brass rod in a groove, as we see in
many less expensive pianos). Or maybe random would be just as good as
tuned, as long as it was varied sufficiently (in other words, like most
"tuned" duplexes).
Regards,
Fred Sturm
University of New Mexico

Ron Nossaman wrote:
> 
<snip>
> 
> A tuned rear duplex isn't part of the vertical hitch scheme of things for at
> least a couple of reasons. The first, I've already mentioned. You can get all
> the back scale lengths of each plain wire unison approximately the same length
> within the unison with no tied loops by having three hitch pins in a row,
> parallel to the notch. This would give you one unison with shorter back length,
> and one with longer for each set of three pins. The difference in the back
> lengths being the difference between unison placement on the bridge between
> those two unisons. Now find a tuned duplex length for one of the unison pairs
> that satisfies the other. But, the hitches are probably too wide to do this,
> especially in the treble, so this probably won't work anyway. Alternately, two
> hitch pins, with a loop and a tie, could be used for each unison. Now all you
> have to do is locate the hitches, the bridge, and the plate accurately enough
> that everything falls into place and the duplex is automatically in tune. This
> isn't a likely possibility, which is why tuned rear duplexes aren't part of the
> vertical hitch scheme of things.
> 
> Ron N