Organ Voicing

[Leisesturm]

Greetings: Since June of last year I have been playing a Rodgers Trillium Masterpiece 928 in a roughly 400 seat sanctuary. A technician is coming on Thursday to do some badly needed voicing but wants me to tell him what I want the instrument to sound like. I have some ideas, but it is the first large instrument of my career and I want to do it right and not necessarily what I personally would like. So.. some questions:
1. How much louder than the rest of the Swell Division should the Swell Reeds be? At present they aren't really much louder than the Swell Principal.
2.Should the Pedal Principals be louder than the Great Principals? These are. Quite a bit so. In fact for general all around playing I use the 16' Lieblich and 8' Gedeckt and that furnishes plenty of bass for hymns!
3. How much louder than the rest of the Pedal Division should the (4' Choral Bass, IV Mixture and 32,16,8,4 Reeds be? Should the 8' and 4' Reeds be more prominent than the other two? At present they are tear your head off loud and I cannot use them at all.
4. How much louder than the rest of the Great should the Great Trompette 8' be? At present, like the Swell Reeds it isn't all that loud at all.

Thanks for any opinions, ideas, or other input.

Edit: bonus question: when there is a 16' Sub Bass in the Pedal, where does it fit volume wise in the tonal scheme?

[samibe]

It sounds like your pedal division is a bit too loud. Do you have the pedal channel speakers right next to the console? How similar is the sound at the console to the sound in the rest of the sanctuary? IMO, you want the principals to be the middle-of-the-range stops (not the flutes). That makes it so you can go a bit softer with the flute registrations and a bit louder with the mixtures and reeds.
1. depends on how you want to use them. I'm leaning more and more toward the prominent reeds of romantic era organs. The american classic voicing seems to be more even (the reeds are louder but not enough to make them stick out too much and the flutes are softer but not too much).
2. Usually the similar stops that occur in different divisions should be similarly loud. That being said, I have set my home organ so that the pedal is voiced with more bass than the manuals which makes the 16' pedal stops slightly louder than the 16' manual stops while the other stops are about the same.
3. They need to be usable but the actual setting will depend on what you want to use them for (add a little support, add some crunch, stick out enough to highlight an interesting pedal line, shake the building, etc.).
4. See 1.
If it's possible to set the different manual reeds to different levels independently, I would set up one that could do a convincing "En Chamade" (above a full principal chorus) and the rest at different levels so that a reed could be used to add color to (or solo over) various ensemble registrations.

[Melos Antropon]

A couple of more thoughts: As a rule, the swell principals are somewhat softer than the Great principal chorus, yes.

The 16' pedal Bourdon (Subbass) is moderately loud, (mf to f) but is mostly fundamental, so it cuts through most of the ensemble. It is loud enough for a chorus of manual stops, but soft enough to work well with quieter registrations, too.

Tony

[myorgan]

To add to what Tony has said, the 16' Lieblich Gedackt should not be loud enough to support any of the manual choruses (i.e. Gt. Diapason Chorus, Sw. Flute Chorus, or Reed Chorus). That's the job of the 16' Bourdon or 16' Diapason. The 16' Lieblich Gedackt should be just soft enough to accompany the softest combinations.

Michael

[Eric Mack]

@Leisesturm, when it comes to voicing, I'm probably the least knowledgeable guy on this thread, but I will risk a comment: As I have been studying how to voice my organ I found the manufacturer had placed a bar chart inside that shows the recommended relative volume levels of each rank. I suppose that is a starting point and it will have to be adjusted from there. In my case it is a theatre organ, so it would be a distraction for me to post that here; however, perhaps such a chart exists for classical organs or for your model in particular. I hope that helps.

Eric

This thread did inspire me to go read up on voicing. :-)

[myorgan]

Great idea, Eric. Does anyone know if a bar chart for stop comparison exists for Classical organs?

Michael

P.S. Eric, you could post the bar chart in the theatre Forum, if you feel so inclined. I'd be very curious to see it for reference purposes.

[Eric Mack]

Michael, I posted the Theater Organ Voicing chart as requested. I'll also post to a Facebook group and see what others say.

[myorgan]

Thank you. I followed that thread with great interest!

Michael

[Leisesturm]

Thank you all! I was able to see @samibe's reply last night and literally slept on the advice. The rest I saw this morning on my phone as I was headed to the appointment. Great stuff. It looked to the technician like the organ was back to factory defaults and was in no way voiced to the church at all. It is unknown whether this happened after an initial voicing or whether it was never voiced at all. The instrument was installed in 2008. We have decided to take on the voicing in stages. Today's stage involved lowering the level of the entire Swell Division, reeds and all. This also lowered the level of the Pedal 16' Lieblich Gedeckt since it is borrowed from the Swell. I never noticed before how faithfully it responds to the Swell Pedal. It can get quite soft now and still have some presence with the expression shoe open. The entire Pedal Division was lowered even more than the Swell was. The Great and Choir were left alone. These changes are on top of an earlier lowering of the master volume of the entire organ that I had made shortly after taking the position.

Interesting about the role of Swell Reeds in Romantic vs American Classic tonal schemes. This is a decidedly American Classic instrument. I'll live with it this way for awhile and see what happens. There is an independent Festival Trumpet that functions as the En Chamade Trumpet. It's volume was lowered as well. I had a Memorial Service with lots of music following the voicing so I was able to test out some combinations. Tomorrow I will have some time to really listen to what has been done. First impressions:
I see the logic of what the technician has done. I can live with it as it is better than how it was. The 16' Subbass and the 8'Gedackt in the Pedal are now identical in level with each other and also to the Great 8' and 4' flutes. I don't know if that follows some sort of convention but IMO it makes both Pedal Flutes just that bit uncomfortably loud still. Before today the flutes of the wide open Swell and the Great Flutes were equal in level. Now the Great Flutes are a step above the Swell Flutes at their full expression. Swell and Choir are about equal in level.

I can see why large organs take many months to voice. I think the St. Thomas organ is set to take a year or more of finish voicing. Just the gross changes we were making took over an hour.

[voet]

I have had my instrument for nearly a year and I am still making adjustments. One thing that I find helpful is playing a wide range of literature. As I play pieces I know what changes need to be made. I find it difficult to find everything just by going through each stop from top to bottom. Also, after doing this for a while, we get aural fatigue. I am fortunate to be able to make changes myself at the console, but if you cannot do that, I suggest having a notebook at the console to make notes for your next session with the voicer.

When I began playing at my last church, I disliked the Great mixture. The guy who had been servicing the organ wanted to replace it. He was really a hack, so I brought in someone else with more skills. There were just a few notes in the mixture that were too loud. When the new guy adjusted them, the mixture was perfect. Sometimes just one or two notes need to be tweaked to make things better.

Best wishes to you, Leisesturm, on your journey to improve the sound.

[jbird604]

Voicing any large organ OUGHT to take a lot of time, as it is a very detailed and grueling process. Your ear needs time to rest between sessions, and trying to drill down to too many details in one sitting may result in not paying enough attention to the nuances. So be patient and let this work out over a period of time.

There are many opinions about stop to stop and division to division balances. I have a general plan in my head that I tend to apply to most any organ I voice, but sometimes I find that a given organist will want something done that doesn't quite fit in my neat little plan. It's obviously important to consider the organist's preferences, but you also don't want to create a "quirky" organ that a visiting organist might find frustrating or un-musical.

As to the 16' and 8' primary pedal flutes, I tend to voice them a tad softer than some people do. I'm of the opinion that the primary 8' flute in the pedal needs to be soft enough that it can be combined with the Lieblich Gedeckt when used as a pedal underpinning for a soft celeste or string. This means I set it quite a bit softer than the 8' flute on the great, which is usually a fairly weighty stop, second only to the 8' principal.

This is harder to do when dealing with a unified organ. In many analog organs, both Rodgers and Allen, there might be only a single flute rank, and it was not possible to make the pedal 8' flute softer than the great. Surprisingly, this is also true with my current organ, the Allen Renaissance R-230, because the pedal flute is simply borrowed from the great, therefore must be the same level. Best I can do is to use DOVE software to roll off the loudness of the pitches below middle C quite rapidly. This lets the stop work as a manual stop, with sufficient body in the middle of the keyboard, and as a pedal stop in softer registrations.

This is only one example of the detailed pondering that may go into doing a truly artistic job of voicing a large digital organ. So keep at it. At least you have tamed it enough now that you feel a little better about it. It will only get better and better as you spend more time working with it.

Keep a notebook at the console and write it down every time you find a certain stop to be too loud or too soft for your purposes. Then have the voicer work on that in subsequent sessions until you get all the bugs worked out.

[Leisesturm]

So the voicing project is going well. At this point I would like to ask if anyone knows of written material on the art of Organ Voicing. In particular the note by note or maybe octave by octave voicing of stop families. I notice the lowest octaves of just about all the ranks on the organ petering out. This is not because of amplifier or speaker issues. There is plenty of both, its how the stops are voiced. I have seen and heard organists playing Bach fugues and you can hear all the counterpoint their left hands are playing even with both hands on the same manual. I've never had an instrument that can do that. Mixtures are supposed to help with that and if I am being honest I think there is greater clarity of the polyphony when using them but its very subtle. I just wonder if there are some definitive guidelines for which stops should be voiced with the middle octaves prominent or maybe the lower octaves prominent or should all the voices be leveled off to be as even as possible? Thanks.

[myorgan]

Leisesturm,

Unfortunately, I would be of no help to you here. Allen ADC organs come fairly well-regulated, and fairly even throughout their range from the factory. If not, it's a fairly simple fix-unlike your Rodgers.

That said, however, I would think an Allen voicing chart for a larger organ would provide some good guidelines, and Allen put out a voicing record that provides good guidance as well regarding the relative balance of the stops.

Outside that, I used recordings of various organs from organstops.org for voicing the timbre of stops, but it appears to be down for some time now. Best with your efforts.

Michael

[John Drabik]

It seems unlikely you'll find a "generic" Organ Voicing treatise for electronic organs, but Colin Pykett (pykett.co.uk) and others have some thoughts on the subject. The problem(s, multiple) with electronic organs stems from the fact that they are in no way related to construction methods that affected pipe voicing. You *will* find Voicing books for pipe organs; from your question though, it sounds as if you may better off starting with a good dictionary of organ stops (e.g., Stevens Irwin's classic - if you can find one that got printed correctly!). Then, something that is common to pipes and electronics is a better understanding of Acoustics (see the foundation works by Helmholtz and others). Next, study some of the pipe organ construction methods, and pay particular attention to "pipe scales". These laid out specific mathematical reductions in the size and shape of each pipe as you walked through a rank. And be aware that pipe scales won't apply to electronics (keep reading).

Electronic organs do no such thing. Older models (analog, VO organs using synthesis, etc.) usually scale several related electronic components at a time across several notes, because electronic parts do not have the required gradations in value (e.g., resistors of, say, 10, 12, 15, 18, 22, 27, 33 ... ohms). These steps are not related to pitch steps across the temperament (some books, like Audell, describe "of such-and-such a size, halving on the 19th step" and terminology like that. In between, each and every pipe was scaled by an amount of, say 1/19th of a given dimension like mouth height. Electronic organs don't have that luxury, and rely on lumpy-value resistors and capacitors, for example. Capacitors are in many ways worse than resistors, since they typically are not as accurate percentage-wise. For example, it's not uncommon to find a .22uF capacitor that reads as low as .15 or as high as .33 - thus overlapping other capacitors range, and none of the builders took the time to measure each and every capacitor to the 1/24th or better value needed to actually scale the electronic circuit to a pipe-scale equivalent. This has improved in recent years but in general the value printed on a capacitor is less "reliable" than that on a resistor. It's also relatively cheap to find accurate resistors, while accurate capacitors are much more expensive (e.g., mylar vs simple ceramics). So first you don't have the gradations of value to do accurate scaling, and second you can't guarantee the values especially for lower cost parts which most organs use (or you'd be paying a lot more). (Of course, some of us find the inaccuracy to be one of the things that makes some older organs sound good).

Newer electronic organs may use a collection of pre-recorded or pre-modeled (for synthesis) voices from classic organs such as those in Germany, Holland, or even some in the USA and England. While complex software can adjust model parameters, that's a painstaking process that is driven more by the sample set *and* the skill of the original voicers (dating back centuries in some cases), than by predictable scaling models; the limits of Allen punched cards is a good example. Pre-recorded samples are worse in a way - you can't change much about the sample, but you may be able to adjust attack, decay, or volume of them a bit (and thus violate the original voicing). This also leads to difficulty with those wanting noticeable chiff - the samples may have been recorded from low-chiff pipe sets (under the assumption it's easier to add it back than to take it out?). All but the best chiff "side-car implementations" are dicey because they don't have a usable set of predictable voice model parameters to work from (and each harmonic component of the chiff is related to some aspect of an actual pipe's construction, e.g., wind pressure, foot size, wind path to the languid, nicking, beards, bars, height/width, pipe wall thickness, etc etc etc. Electronic organs like some older Rodgers and Allens used AC-coupled harmonics from a given note (the "E" connection being most notable), and the capacitors resulted in a rapid fade of the leading attack sound(s). My Providence 330, for an 8' pitch for example, incorporates a 4', 2 2/3', 2', and "E" chime connection to generate the "poof". Fortunately, it's not a lot of extra circuitry but then, it's not a perfect chiff either, and it wears on the ears after a while because it is so predictable, unlike a pipe. Put that pipe on a tracker chest under a great organists fingers, and the difference between real chiff and phony stuff is quite remarkable (listen to E. Power Biggs play one of the Bach Trio Sonatas if you doubt this. Or, his opening rendition of the Passacaglia and Fugue (the pedal part) if you want to hear both chiff and air sound that work to great effect, granted, on a real organ!). Some pipe organ builders went to great length to eliminate chiff (and samples from their pipes may never get that back). Others sometimes included similar ranks, some without chiff and some with chiff (although baroque organs in particular usually had noticeable chiff on almost every rank).

To get around the problems in electronic organs, many manufacturers provided some kind of per-voice volume control, some "feed thru" control, adjustable (but rather flat) air sound, and a couple of chiff harmonics (real chiff has dozens or hundreds of harmonics. Replicating that, with actual control of each component, would be enormously expensive). They also provide "voicing guideline". These change from vendor to vendor (due to the way they built their instruments, e.g., individual oscillators vs tone-generator chips, how they broke down voicing resistor/capacitor (RC) and/or inductor (L) circuits, etc. Again, for an organ like the 330, the recommendation began with organ at full volume, a particular chord set in the mid range (C-E-G), and a couple of extra people to aid the process (one to adjust controls in the back of the organ, another to sit in main seating area to provide feedback). Organs with "software voicing" get rid of the need for the person at the back of the organ perhaps, but you still need a trained ear, common sense (look into things like the "diapason chorus", "flute chorus", and Irwins or other books on relative pipe scale and volume), and that feedback person in the audience area.

Sorry for the long-winded, and yet still elementary, explanation. The short answer is: if you're not able to devote considerable time to learning how to voice an instrument (and all the factors that go into that), whether pipe or electronic, you should start with the default settings from the vendor and then adjust very carefully, in small steps, and with plenty of feedback from people in the audience as they listen to a variety of actual music being played, not just C-E-G chords.

[jbird604]

Sometimes one must take rather bold action, even without clear direction! (I sound brave, don't I?) On that note, when I got my Allen Renaissance organ a couple years ago now, I thought the voicing was abominable, with a number of characteristics that I found unattractive. An apparent "hump" in loudness in the middle of the scale, a lot of seemingly random and meaningless tuning deviations, the odd note here and there that was louder or softer than its neighbors. And other abnormal details, some large, some small.

So I thought I had little to lose by just wiping out the irregularities I seemed to be hearing. Now, Allen MOS organs and to a great extent even the ADC and MDS organs have a perfectly "flat" scaling of all stops (modified only by the settings of the bass and treble pots, on the DAC of MOS organs, on the TG boards of ADC and MDS models). So it seemed peculiar to me, given how nice most of those sound, that this Renaissance organ would have what sounded like a lot of humps and bumps and quirks in the scaling. I suspect that a well-meaning but untrained "voicer" might have introduced some of these quirks, but some of them are probably inherent in the out-the-door factory voicing due to some design spec drawn up in the early days of production, that might or might not still be valid.

So once I had DOVE connected, I saved a copy of the voicing as I received it, but then I went to work. On nearly every single stop, I started by making the note to note level perfectly flat from bottom to top. (I also edited all the tuning curves and removed weird random deviations, moving most of the curves in the direction of flatness, while not doing away with all tuning distinctions among the ranks.) But first, I made the stops all flat in scaling.

I also adjusted the bass and treble controls of the separate "frames" in the audio control to the flat settings, so the tone controls would not be influencing the tone coloring until and unless I decided to tinker with them. On your Rodgers, you have bass, treble, balance, and level controls on each divisional audio adjustment page, so I'd suggest setting those flat too, when doing a from-scratch voicing.

Next, I adjusted the relative volumes of all the stops IN THE CENTER OCTAVE to be correct, by normal standards. That is to say, the Octave is slightly softer than the 8', the 2' is slightly softer than the 4'. Flutes slightly softer than principals, strings even softer, power reeds louder, solo reeds softer, etc. In other words, I "voiced" it like I was voicing an ADC organ (though it was of course more complex, because I was working with 32 separately adjustable ranks, not just a dozen or so stop groups).

Then I began to listen critically to the stops at the upper and lower ends of the scale. While "flat" scaling might seem optimum, in practice, you may want to roll off the loudness of those 2-footers and mixtures in the upper octaves. And you may want to make the pedal foundational stops heftier in their lower octaves. And when a stop, such as the 8' great flute, has to serve both great and pedal, you may want to taper the lower octaves of it quite severely. I would apply this tapering with the "rubber band" method in DOVE -- setting the desired level on certain notes, then using those notes as anchors for the "ruler" tool to evenly taper the notes in between.

Now and then I find that I really DO need to make an individual note louder or softer than its neighbors, if it clearly doesn't fit in the scale, either because of audio system resonances, or perhaps some quirk of the samples used. But I used that possibility very sparingly.

I thought the end result was much prettier and more pleasant to listen to than the original voicing. It certainly seemed more "disciplined" and more like a well-maintained pipe organ than a typical old clunker with issues, which may be fun to play sometimes, but not what you'd want to play all the time!

My point (and I do have one)......... If you feel that the lower octaves of some or all the stops are petering out, then use the controls to raise those levels. That what it's for! Before you spend vast amounts of time doing that stop by stop and note by note, you might first try just boosting the volume of the subwoofer channel, or boosting the bass controls on the manual channels, to see if that adds the heft you want. You just have to be careful not to over-boost the bass anywhere lest you blow out the subs on the low notes of the pedals.

So do what you gotta do!

[voet]

Excellent points, John. Thank you for your detailed account of your voicing methods.

I would only add that, in some rooms, certain frequencies are either amplified or dampened. In very large reverberant rooms (think large European churches) the high frequencies need to be boosted, while the low frequencies have more presence. In a practice studio this voicing would be far too aggressive for most people.

I believe that one of the reasons some of the first pipe organs of the organ reform movement built in the US were unsuccessful is because voicers did not adequately consider the relatively dry accoustic invironment vis a vis the European model they were emulating.

[Dutchy]

I agree with voet: excellent. Thanks for sharing your experience!

[Leisesturm]

Thanks John, that was amazing as always. Thanks to this thread, Palm Sunday and Easter Sunday are going to (with God's help) be very memorable where the organ is concerned.