Mixture composition: Odell Op. 649
One of the small projects we’ve been working on over the summer has been some long-planned upgrades and adjustments to the organ we built in 2011 for the Historic First Church in Orange, Connecticut. Of particular interest to some of the commenters was the changes in voicing and scaling for the Great Mixture stop, and since the answer to one particular question is a bit more involved and complex, its a good item to post about.
The design challenges I faced here overall were tricky acoustics: a smallish New England meetinghouse with some carpet and virtually no reverberation. The chambers are tight, though the organ is at least located up front and on the central axis. In the design process I had to work out how to fit 22 ranks in a space that formerly only fit 12 and I was determined not to sacrifice maintenance access.
The previous organ had the ubiquitous-for-its-time “Grave" II-rank mixture (12-15) slapped on to its postwar Moller 8-4 principal chorus with predictable effect. In our redesign, I didn't want to repeat this, and I also did not want to give up room for the Trumpet 8', because I've done the "reedless Great" thing and lived to regret it. Back when the organ was being designed and built here, the Trumpet 8’ for the Great with 16’ extension for the Pedal was the primary consumer of space, resources and attention in the division’s new design.
The design was to include a new 2’ to complement the rescaled and revoiced Moller principals. One way to go here might have been to just add an independent 2 2/3’ rank, which adds color to the plenum, but not much brilliance. I instinctively felt that a full 1 1/3' mixture in either III-rank (19-22-26) or IV-rank (19-22-26-29) configurations would overwhelm the ensemble, particularly in this small, bright room that acoustically just eats bass.
My solution, which I readily admit I've seen elsewhere, was a III-rank mixture based on 2’ pitch (15-19-22) with a split slider for the 15th. This is also sometimes referred to as an independent or “double” draw. This is a compromise, as is any situation where you're creating a compound stop that doesn't have all the pipes on the same toe board channels. While the 2’ rank was all new, I tied my hand behind my back further: because of budget and timing issues I was limited to composing the mixture from existing pipes left behind by the inimitable Garo Ray, (a subject for a whole other post). The final result was workable, but not entirely satisfactory, which everyone understood as the project wrapped up. Fortunately, the church this year assented to allowing me to make new mixture pipes, bundled along with other upgrades and tweaks.
The trick in making this compromise work is in the composition, scaling and voicing. This is where we get into the real nitty-gritty, which, with apologies for the coming digression, is old hat to experienced organbuilders conversant in what I jokingly refer to as one of the black magic areas of organ building.
Chorus mixtures are always a combination of unison pipes and quints arranged at specific intervals. The unisons can either be successive or repetitive pipes in the chorus, and likewise the quints, which are tuned pure just like any other mutation. Ideally they speak on a common wind channel to promote unity of speech and tuning.
The science in making this work is in evaluating and managing the scaling relationships between the elements; the art is voicing them so that the blend is both effective, but balanced to the rest of the ensemble and hence not unpleasant to the ear. Different builders have different customs on how they manage this; many builders simply scale quint ranks 2 notes smaller than the unisons, while others scale them absolutely uniform and manage the issues in voicing. I have heard some people mistakenly assert that all mixtures with odd numbers of ranks are inherently unbalanced and should therefore be avoided, despite copious successful evidence to the contrary. As with all good organ building practice, experience teaches you to duplicate what works and make adjustments when it doesn’t.
Since around here I am in many cases the designer, pipemaker, voicer and finisher, in this instance as the pipes were being laid out I had the advantage of knowing precisely what it was that I aimed to achieve. When I made and voiced the new 19th and 12th for the mixture stop in Orange, in addition to making sure I had good, clear pipe speech, I was very deliberate in how I managed the perceived volume and treble ascendancy, especially in the range of the compass where the two quint ranks converge:
C1 to A22: 15-19-22 :: (2’ -1 1/3’ - 1’)
A#23 to G44: 12-15-19 :: (2 2/3’ - 2’ - 1 1/3’)
G#45 to C61: 8-12-15 :: (4’ - 2 2/3’ - 2’)
In listening yesterday, I was mostly satisfied with the result. In terms of inner monologue, you may read this as me being ecstatic. I’ll take another day with it next week as well as tend to some other things, and at some point to truly judge I need to hear it in worship, which is thankfully easy enough to arrange.
Thank Heaven for patient and trusting clients.