Jump to content
Mander Organs
Sign in to follow this  
father-willis

Celestes

Recommended Posts

When dealing with undulating ranks, where should the de-tuned rank be placed?

I have read that the unison and sharp/flat rank should be separated by (at least?) one other rank. On the other hand there are compound two-rank celestes stops which must have both sets of pipes on the same slide, together. Does it matter? Advice please.

Share this post


Link to post
Share on other sites

Undulating ranks are certainly better placed well apart.  At Tonbridge School, Marcussen have placed the Solo undulating ranks next to each other but on elevated pipe blocks,  in much the same way as a Cornet may be mounted.  The Swell Voix Celeste is also stood up high on an elevated block, directly above the slide on which it stands.   On another organ where we added a Celeste,  the Swell Box had been enlarged to accomodate a 16' reed bass and the Celeste went in there, too.  As a result, the Celeste will beat happily and consistently with the Sw Open, Lieblich or Gamba, giving three distinct choices of undulant.

Whilst it's true that on some organs, drawing the Celeste alone will automatically bring on the unison string stop as well, this is usually done by a linkage in the stop actions. It doesn't necessarily imply that the two ranks must stand next to each other on the soundboard.

I hope that adds a little insight.   Anything which gets away from the 'snivelling strings' syndrome has got to be a good thing!  

Share this post


Link to post
Share on other sites

However, many 2 rank celestes do have both ranks quite close and function perfectly well. My celeste is separated by just one rank (bourdon 8'). If the two undulating ranks are placed too far apart, then any temperature difference (caused by winter heating etc.) might noticeably slow or speed up the undulation.

I've read of the idea that pipes playing close together tend to pull one against the other, but I've yet to hear a physics-based explanation unless their mouths are close and facing each other.

The C /C# pipe planting has more to do with weight and space distribution on the soundboard than sound - though I hope to illicit some interesting comments to the contrary. Chromatic soundboards seem to sound the same to my ears.

Share this post


Link to post
Share on other sites

It's been pointed out, not least by skilled organ-builders, that planting the undulating rank next to the in-tune rank causes the two to pull into tune with each other, thereby negating the celeste effect.  Thus, it is common practice to put the undulant next to the Open Diapason, away from the salicional, gamba or what-have-you.  This would spoil the effect of Open and Celeste called for, for example, in Whitlock's Folk Tune.  However, Whitlock was a Compton fan and probably thinking in terms of his Compton at Bournemouth. In such an instrument, each rank might stand on its own soundboard and therefore it would be quite practicable to draw the celeste with the diapason and achieve the effect Whitlock specifies.

On the subject of celestes, I (nearly) always find that a sharp celeste is a much nicer sound than a flat one.  When the 1908 Walker at St. Leonard-at-the-Hythe, Colchester was ably restored by ex-Walker man Ken Canter in 1973, he tuned the celeste flat and I didn't like it.  He said that Walker celestes were normally tuned flat, but sharpened it for me and the old effect returned.  I still think it's one of the nicest celestes that I know.

Can anyone tell us whether certain builders tuned celestes sharp or flat?

I think I have previously mentioned on this forum the organ at Burnham-on-Crouch, Essex, where I went to hold keys in 1971 when John Budgen (Bishop & Son) called in to give it a check-over prior to the opening recital by Gordon Phillips.  The church organist asked for the celeste to be tuned flat, but there wasn't enough length in the pipes, so it had to be left sharp.  Afterwards, the organist said what a difference it made to have the celeste flat.....

Share this post


Link to post
Share on other sites
51 minutes ago, David Drinkell said:

On the subject of celestes, I (nearly) always find that a sharp celeste is a much nicer sound than a flat one.

If the two ranks are of identical design, construction, winding and voicing how can there be any difference if no other stops are being used to give a sense of the pitch of the instrument?

Share this post


Link to post
Share on other sites

The celeste is tuned slightly sharp (about 3 beats a second) or flat to the other rank, producing a pleasant heterodyning when the two are drawn together.  In some organs, the celeste is also pleasant used with other stops.  The in-tune rank may not be exactly the same scale and voicing as the celeste, depending on the builder and the effect desired.

I believe it is better to tune the celeste to the in-tune rank throughout the compass, rather than tuning the middle octave and doing the rest in octaves.  With the latter practice, the upper notes can be unpleasantly quick in their beat.

Share this post


Link to post
Share on other sites
On 23/11/2017 at 02:33, David Drinkell said:

I believe it is better to tune the celeste to the in-tune rank throughout the compass, rather than tuning the middle octave and doing the rest in octaves.  With the latter practice, the upper notes can be unpleasantly quick in their beat.

That is true.  The number of cents by which the celeste rank is detuned should increase towards the bass and decrease towards the treble if the beat frequency is to remain roughly constant across the compass, so it requires careful tuning.  However I have found that if this simple rule is applied too slavishly, the effect can become just too out of tune in the bass for my ears.  For instance, if a 2 Hz beat rate is desired the detuning required at tenor C on an 8 foot rank is 26 cents, just over a quarter of a semitone.  Should the detuned rank be carried right down to the bottom note, it has to be detuned by over half a semitone, grossly out of tune with the corresponding in-tune rank.  Faster beats necessarily result in even greater amounts of detuning.  To avoid this problem of two obviously-out-of-tune pipes sounding simultaneously, I shade the detuning towards the bass to reduce the beat rates between the two fundamental frequencies to at least half of what it should be in theory (i.e. to about 1 Hz in this example).  Yet this is not readily apparent to the player or listener because beats at the 'correct' frequencies still occur between the second harmonics, which are generally stronger than the fundamentals in string pipes.

An interesting observation is that tuning a celeste rank this way (i.e. constant beat rate across the compass) results in its octaves becoming perforce slightly out of tune.  This adds to the aural interest and richness of the subjective effect of a celeste.

I've drawn up an Excel spreadsheet which tabulates the detuning required for each note across the compass for any desired beat frequency, which the user has to insert.  If anyone wants a copy of it please send me a private email, either via the forum or using my personal email address which you can get from my website at www.pykett.org.uk (I don't want to broadcast my personal address on the web for obvious reasons).

CEP

Share this post


Link to post
Share on other sites
On 22/11/2017 at 23:44, David Drinkell said:

On the subject of celestes, I (nearly) always find that a sharp celeste is a much nicer sound than a flat one.  When the 1908 Walker at St. Leonard-at-the-Hythe, Colchester was ably restored by ex-Walker man Ken Canter in 1973, he tuned the celeste flat and I didn't like it.  He said that Walker celestes were normally tuned flat, but sharpened it for me and the old effect returned.  I still think it's one of the nicest celestes that I know.

 

Like many on this forum I imagine, I've heard this many times in that some organists prefer sharp celestes and others flat ones.  Personally though, I can't really tell the difference though I'm quite prepared to believe that the ears of a professional player are more attuned (ha ha) to some subtle difference(s) which I can't perceive.

 

On 23/11/2017 at 00:38, innate said:

If the two ranks are of identical design, construction, winding and voicing how can there be any difference if no other stops are being used to give a sense of the pitch of the instrument?

 

So without thinking very much about it, I've always taken innate's view quoted above that it shouldn't really make any objective difference.

However I've just done a bit of arithmetic and discovered that there is an objective difference after all.  Let the beat frequency across the compass of the double rank be set to N Hz (i.e.N is the beating frequency between the in-tune and out-of-tune notes).  Typically N might be 2, 3 or something in that region - it doesn't matter.  It is assumed that N is made constant across the whole double rank by careful tuning of the detuned pipes.

For both sharp and flat detuned ranks, there will then be a beat between successive octaves of N Hz.  This means that, if you play (say) middle C and treble C on the detuned rank alone, the two notes will beat at N Hz.  This will occur for any octave across the whole detuned rank.  Thus the octaves of the detuned rank are not pure (not in exact tune), as I mentioned in my previous post above.  However, here's the rub: the octaves are sharp (widened) by N Hz if the celeste is tuned flat to the in-tune rank, and they are flat (narrowed) by the same amount if the celeste is tuned sharp.

I think my calculations are correct, though I will check them again.  If they are correct, this shows that there is a small difference between the two cases.  The difference is calculable, measurable and therefore objective rather than being subjective.  Therefore it is quite possible that some people might be using these small differences unconsciously when expressing a preference for one type of celeste over the other.  This is not extraordinary in any way if one looks at the situation in this manner: the detuned rank has had an unusual 'temperament' imposed on it consisting of impure octaves rather than the more usual case of impure fifths, thirds or whatever.  In one case the impure intervals (the octaves) are flattened and in the other they are sharpened.  Lots of people are very choosy about which temperaments they prefer, on the basis of flattened or sharpened intervals and the way they beat, and this case is not really any different when one looks at it like this.

CEP

Share this post


Link to post
Share on other sites

Perhaps what's really wanted is a celeste with two dedicated ranks (i.e. not used with any other), one tuned sharp and the other flat, so that the mean pitch remains accurate.  Tuning each rank sharp or flat relative to another rank as reference at half the required beat rate should do it.

Paul

Share this post


Link to post
Share on other sites

Thinking a bit further, David Drinkell's preference for a sharp celeste rank might be related to the impure-octave temperament imposed on the detuned rank as I explained above.  A sharp celeste rank has slightly narrower octaves than the exactly-tuned octaves of the in-tune rank.  This means its thirds will be slightly better in tune (narrower) than they would otherwise be, and it is the rapidly-beating thirds in many temperaments (and certainly in equal temperament) which are the most objectionable - ignoring extremes like Wolf intervals here for the sake of argument.  On the other hand, a flat celeste rank has slightly wider octaves and therefore wider thirds also.  These will beat even faster than they do in ET and consequently they will be that bit more objectionable to someone whose ears can detect the difference.  It is therefore quite possible David is picking up on this difference when he expressed his preference for a sharp celeste.

23 minutes ago, pwhodges said:

Perhaps what's really wanted is a celeste with two dedicated ranks (i.e. not used with any other), one tuned sharp and the other flat, so that the mean pitch remains accurate.  Tuning each rank sharp or flat relative to another rank as reference at half the required beat rate should do it.

Paul

Hope-Jones provided three-rank celestes in some of his organs, e.g. at Worcester cathedral.  There was one unison rank, one tuned sharp to it, and one tuned flat.  The unison and sharp ranks spoke when the stop tablet was first pressed, and the flat rank was added when it was pressed again.  (At least, I think that's how it worked, though the tablet might have had an extra mechanical detent which you moved through using extra finger pressure if you wanted all three ranks).  However he used the first (double-touch) method widely, including in some quite small organs such as that at Pilton (Devon), where the detuned rank is still added to this day to the Phoneuma on the swell if you press the tablet twice - at least, this was so when I last played the instrument.

I think similar arrangements were provided by other builders.

CEP

Share this post


Link to post
Share on other sites

I prefer the slightly brighter effect that a sharp celeste gives when added to the in-tune rank.  A flat celeste gives a more sombre effect, for which I don't care, although others may well have the opposite preference.

Share this post


Link to post
Share on other sites
15 hours ago, Colin Pykett said:

Thinking a bit further, David Drinkell's preference for a sharp celeste rank might be related to the impure-octave temperament imposed on the detuned rank as I explained above.  A sharp celeste rank has slightly narrower octaves than the exactly-tuned octaves of the in-tune rank.  This means its thirds will be slightly better in tune (narrower) than they would otherwise be, and it is the rapidly-beating thirds in many temperaments (and certainly in equal temperament) which are the most objectionable - ignoring extremes like Wolf intervals here for the sake of argument.  On the other hand, a flat celeste rank has slightly wider octaves and therefore wider thirds also.  These will beat even faster than they do in ET and consequently they will be that bit more objectionable to someone whose ears can detect the difference.  It is therefore quite possible David is picking up on this difference when he expressed his preference for a sharp celeste.

Hope-Jones provided three-rank celestes in some of his organs, e.g. at Worcester cathedral.  There was one unison rank, one tuned sharp to it, and one tuned flat.  The unison and sharp ranks spoke when the stop tablet was first pressed, and the flat rank was added when it was pressed again.  (At least, I think that's how it worked, though the tablet might have had an extra mechanical detent which you moved through using extra finger pressure if you wanted all three ranks).  However he used the first (double-touch) method widely, including in some quite small organs such as that at Pilton (Devon), where the detuned rank is still added to this day to the Phoneuma on the swell if you press the tablet twice - at least, this was so when I last played the instrument.

I think similar arrangements were provided by other builders.

CEP

Tuning undulants however is not always done mathematically - they are a stop solely for effect, not part of structured organ tone. The rate of beat should increase in the treble to maintain the shimmering effect, so they are tuned by ear. My preference is to set a beat at mid C slightly slower than an equal temp C-G interval so long as the ranks will take it. Most English Salicional/Dulciana type undulants I tend to find are too quick in the middle registers, however more biting strings can benefit from a quicker speed. It really is all done by ear.

Share this post


Link to post
Share on other sites

In German organbuilding, there is the term “Terzschwebung”, and I understand that this is the most frequent method of celeste tuning. It refers to the tuning process: Both ranks are pulled, and a major third is played; in the sharp rank, the upper pipe is silenced, while in the unison rank the lower one is. Then both remaining pipes are tuned to a pure major third by way of sharpening the celeste pipe. For a flat celeste, the silencing would be done the other way round. That way, the beats per note will increase with the pitch in a pleasant way, and it’s quite easily done.

Is this the usual method with British builders as well?

All best wishes,

Friedrich

Share this post


Link to post
Share on other sites
1 hour ago, sprondel said:

In German organbuilding, there is the term “Terzschwebung”,

This sounds like a beautifully “pure” and musical approach! Thanks for your explanation, Friedrich—I suspect many will have been enlightened by it.

Share this post


Link to post
Share on other sites
23 hours ago, sprondel said:

In German organbuilding, there is the term “Terzschwebung”, and I understand that this is the most frequent method of celeste tuning. It refers to the tuning process: Both ranks are pulled, and a major third is played; in the sharp rank, the upper pipe is silenced, while in the unison rank the lower one is. Then both remaining pipes are tuned to a pure major third by way of sharpening the celeste pipe. For a flat celeste, the silencing would be done the other way round. That way, the beats per note will increase with the pitch in a pleasant way, and it’s quite easily done.

Is this the usual method with British builders as well?

All best wishes,

Friedrich

I think I have encoutered this and to English ears it doesn't sound very natural and is rather too fast to produce the type of effect we are used to. Again though, it depends on the tone of the string rank and the effect desired.

Share this post


Link to post
Share on other sites
1 hour ago, ajsphead said:

I think I have encoutered this and to English ears it doesn't sound very natural and is rather too fast to produce the type of effect we are used to. Again though, it depends on the tone of the string rank and the effect desired.

Maybe there are other “pure” intervals that could be used to similar effect.

Share this post


Link to post
Share on other sites

I agree with ajsphead.  At Liverpool all the undulating ranks are tuned with slightly different speeds according to their voicing and the effect they give.  I tune the unda maris with the slowest beat, keeping it even to the top.   solo strings, which are the most biting, the fastest, getting slightly faster in the trebles.  the strings are angelicas so tuned flat.

Interesting idea from Friedrich, will have a go sometime and see how it sounds. 

Share this post


Link to post
Share on other sites
On 6.1.2018 at 17:03, ajsphead said:

I think I have encoutered this and to English ears it doesn't sound very natural and is rather too fast to produce the type of effect we are used to. Again though, it depends on the tone of the string rank and the effect desired.

I just found an older comment in another (now defunct) forum in which an organbuilder suggested that the »Terz« approach worked best when applied to the first 24 or 30 notes, from there continuing in pure octaves, so that the beats won’t increase at the former rate. More than one other contributor back then wrote, however, that they tuned individual notes, just by ear, the only condition being that the ranks are positioned sufficiently far apart from each other.

All best wishes,

Friedrich

Share this post


Link to post
Share on other sites
23 hours ago, sprondel said:

the only condition being that the ranks are positioned sufficiently far apart from each other.

I wonder what happens with a 2-rank stop, such as Unda Maris I-II 8'

Would these ranks be positioned well apart from each other, bearing in mind that they would, presumably, be on the same slider?  I'm sure someone here will know for sure.

Share this post


Link to post
Share on other sites
8 hours ago, John Robinson said:

I wonder what happens with a 2-rank stop, such as Unda Maris I-II 8'

Would these ranks be positioned well apart from each other, bearing in mind that they would, presumably, be on the same slider?  I'm sure someone here will know for sure.

A single drawstop can draw whatever you want it to with electric stop action. With mechanical stop action, one drawstop can still be made to draw more than 1 slide.

There are also many instances of the 2 main designated beating ranks being on adjacent slides and the effect working perfectly well. It does occasionally require a little more skill from the tuner.

Share this post


Link to post
Share on other sites
14 hours ago, ajsphead said:

A single drawstop can draw whatever you want it to with electric stop action. With mechanical stop action, one drawstop can still be made to draw more than 1 slide.

There are also many instances of the 2 main designated beating ranks being on adjacent slides and the effect working perfectly well. It does occasionally require a little more skill from the tuner.

Thanks, I suspected as much.  I have seen a cross-section image of a soundboard on which sit two adjacent strings including a celeste rank, so I assumed it must work.

Share this post


Link to post
Share on other sites
On 11/22/2017 at 23:44, David Drinkell said:

... On the subject of celestes, I (nearly) always find that a sharp celeste is a much nicer sound than a flat one.  When the 1908 Walker at St. Leonard-at-the-Hythe, Colchester was ably restored by ex-Walker man Ken Canter in 1973, he tuned the celeste flat and I didn't like it.  He said that Walker celestes were normally tuned flat, but sharpened it for me and the old effect returned.  I still think it's one of the nicest celestes that I know. ...

 

I quite agree, David.

On our Walker instrument here at the Minster, we have a beautiful pair of mild strings in the Swell Organ, rather in the mould of the old 'Father' Willis Salicional and Vox Angelica. Our Vox Angelica was originally tuned flat (as a glance at the pipes will show). However, I don't like the effect either, and have never regretted asking our organ builder at the time to re-tune it sharp.

Neither have I regretted ousting the G.O. Dulciana (which was never used) for a gorgeous second-hand Viole de Gambe. Aside from the fact that this rank is infinitely more interesting tonally, it is of far greater use, both in service-work and recitals.

 

Share this post


Link to post
Share on other sites

In a lot of cases, I think that the presence of a Dulciana is an admission that the other 8' stops are too loud, although the Dulciana is useful to "warm up" the 8' flute (pcnd's Viole de Gambe probably does this even better.  Lawrence Swinyard once likened Great gambas to "army cocoa", which I always thought was a beautiful description, although I never quite understood what he meant).

There are some gorgeous unenclosed Dulcianas about - Arthur Harrison's are generally outstanding (in saying this, I guess I'm perilously close to agreeing with pcnd that his organs are generally too loud!), and I liked our host's provision of one on the (ruckpositif) Choir Organ at Cripplegate (which I first played just after it went in), despite Sam Clutton's opinion that it was about the most useless thing for those circumstances.

Over here, the Choir Organ in older instruments is nearly always enclosed. On the organ at the Anglican Cathedral in St. John's, Newfoundland, which I played for thirteen years, some modest pepping up in the nineties involved scrapping the Dulciana, Viole, Wald Flute and (Hope-Jones) Quintadena in favour of Nazard, Tierce and appropriate flutes to carry them.  The 8' flues in my time were Violin Diapason and Chimney Flute and I never had cause to regret not having anything else. (There was a full-length 16' Dulciana in the Choir box, as well as another one a few feet away from it in the case and played on the Pedal - money must have been no object in 1927 when Casavant rebuilt the organ!).  There was a Gemshorn 8' on the Great, which was a Dulciana in all but shape and was useful in quiet combinations, especially as the Great had octave, sub and unison off couplers, but the Choir Dulciana was never missed.  Here at Fredericton, I find little use for the enclosed Dulciana on the Choir Organ, although I use the Viole d'Orchestre quite a lot to inject "quiet ginger" (Norman Cocker's term) into various combinations. There's also a very quiet Aeoline on the Echo and a Dolcissimo on the Swell which is more of a wheeze than anything else....

Share this post


Link to post
Share on other sites

A dulciana on the G O  of a 2 manual  is useful as an accompaniment to a swell oboe or other suitable solo stop.

On the instrument I am responsible for I have had the C O dulciana tuned sharp to beat against the violin diapason. It also works well coupled up to the S O & used with the viol d' orchestra. Other combinations are available!

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×