News:

If you have difficulty registering for an account on the forum please email antespam@gmail.com. In the question regarding the composer use just the surname, not including forenames Charles-Marie.

Main Menu

How a pipe speaks

Started by David Pinnegar, April 04, 2011, 10:38:18 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

David Pinnegar

Hi!

The other day, lunchtime discussion gravitated to how organ pipes speak.

Someone said that Colin Pykett had done some excellent analysis . . .
http://www.pykett.org.uk/how_the_flue_pipe_speaks.htm
http://www.pykett.org.uk/violone.htm

In finding these files, I noticed another which is seminal. Apparently Hope Jones placed heavy reliance on both couplers and the Quintadena stop and at the time, people commented that his implementation could equate with a 5 rank mixture stop . . . quite a seemingly extraordinary claim, but Colin shows how:
http://www.colinpykett.org.uk/hjquintadenas.htm

His work also demonstrates the part that electronic simulation can play in pipe organ design and analysis.

It is interesting to note the way in which the Victorians reduced organ tonalities down to unisons coinciding with equal temperament making harmonic upperwork more discordant.

Best wishes

David P

David Pinnegar

Hi!

http://hyperphysics.phy-astr.gsu.edu/hbase/waves/opecol.html#c1

http://hyperphysics.phy-astr.gsu.edu/hbase/music/brassa.html#c1

and I'm not sure that I understand entirely
http://hyperphysics.phy-astr.gsu.edu/hbase/music/brassa.html#c1

It certainly demonstrates that the tones of organ pipes are not entirely harmonic and cannot therefore be represented by sequential harmonics with Fourier reconstruction . . . ?

Best wishes

David P

Colin Pykett

There are two distinct issues to keep in mind.

The natural frequencies of a pipe are those it resonates at on its own, when not being driven in the normal way by an air jet at the mouth..  If you put a small loudspeaker at one end of a pipe and then drive it with a variable frequency sine wave, you can pick up the sound at the other end with a microphone.  The mic will show peaks at the pipe resonances, the natural frequencies, but these are NOT exactly harmonically related.  Each resonance is successively slightly more sharp from the previous one.  The differences are due to the end corrections at each end, and the way these vary with frequency.  (They also vary with pipe scale).  Thus, in acoustic terms, the pipe gets progressively shorter as you go up in frequency.

However, if you now drive the pipe in the usual way by blowing air through the languid and into the mouth, it radiates harmonics which ARE exact integer multiples of the fundamental frequency.  This is because the oscillation set up at the mouth is periodic at the fundamental frequency.  Fourier's maths shows that the harmonics of any periodic wave MUST BE exact multiples of the fundamental.

Therefore it is entirely permissible and possible to simulate  a pipe sound exactly using additive synthesis methods of the Bradford/Musicom variety.

There is another important aspect though.  Because of the divergence between the natural partials of the pipe itself and the exact harmonics it actually radiates, this is how the pipe applies a frequency shading (low pass filtering action) to the harmonic retinue generated at the mouth.  The shading is influenced by pipe scale, with narrow pipes (strings) exerting less of a filtering action than wide ones (flutes).  This shading is in addition to voicing adjustments (e.g. cut up) at the mouth which influence how many harmonics are generated to start with.

It can be difficult to get one's head around this at first.  I have had endless discussions about it, which often end with the question "OK, then how does a bell or chime have non-harmonic partials?"  The answer is in the word 'periodic'.  A periodic wave goes on for an indefinite time with no significant variations other than the most minute ones.  This represents an organ pipe.  In this case the maths - Fourier Series - only allows exact harmonics to exist.  (Please don't continue the discussion if you can't follow the maths - it's easier to take it as read!).  But a bell is not periodic - it dies away.  Therefore it is a transient waveform, not periodic.  The difference is significant.  Fourier Series are not applicable to transients.  You have to use Fourier Integrals instead, and these DO allow non-harmonic partials to exist.  But the maths here is even worse .....

It's all explained, but still without using maths, in more detail in my web article which David mentioned above if you have a particularly masochistic bent:

http://www.pykett.org.uk/how_the_flue_pipe_speaks.htm

To conclude, good old Helmholtz was the first to work all this out in the mid-19th century.  Good or what?  Not many people know that!

Best wishes

Colin Pykett

Barry Williams



"Therefore it is entirely permissible and possible to simulate  a pipe sound exactly using additive synthesis methods of the Bradford/Musicom variety."

Dr Pykett's erudite and interesting reply holds totally good for one pipe.  However, the 'Bradford' or 'Musicom' system achieves a synthesis which is not as effective as it might be.  If only there were a complete system and speaker for every pipe it might be a different story!

Both the 'Bradford' and 'Musicom' systems introduce an artificial detuning or random 'blip' which is musically offensive and quite unnecessary.  I have been told (possibly erroneously) that it is not possible to turn this mechanism off, in the same way that Makin, in the old days, persistently refused to let anyone hear an instrument with the 'Rotafon' turned off.

I have been told that 'Bradford' organs can only be repaired by a specialised engineer, whereas other systems can be repaired by any competent electronic engineer.

Churches looking for an electronic instrument must, inevitably, seek the best value for money.  Needlessly complex systems, whether in an electronic instrument or a motor car, do not necessarily represent the best overall value.

Many small churches seek an electronic instrument of, perhaps, fifteen or twenty stops, as appropriate to the the size of the building.  The inevitable comment from suppliers of 'Bradford' instruments is that one does not have to pay any more to have a vast array of stops and the organist does not have to use them.  The inherent lack of musicality in such an approach simply does not merit discussion and has dogged the electronic supplying industry for the past thirty years or so. It results in a gallon in a pint pot and does nothing to help the credibility of electronic instruments.

Some suppliers are willing to install instruments appropriate to the size of church and they are very successful.  I have never come accross a 'Bradford' instrument installed in a village church where the stop list was appropriate.

We are fortunate indeed to have Dr Pykett on this forum and I do hope that he will feel able to respond to the points I have raised, for many churches have no option but to seek an electronic alternative these days.  ( I would have telephoned him to discuss things, if I had a 'phone number.)
Barry Williams

revtonynewnham

Hi

This looks an interesting discussion.  I wonder how the starting and finishing transients of pipe speech respond - are they harmonic, or inharmonic, or both?

The Bradford system uses special hardware, which possibly is only available to certain engineers - I'll try and remember to ask the Comerfords next time I see them.  Musicom also uses their own design of module for the tone generators (or at least they did last time I looked at their web site) - but that may be easier to obtain.

I can understand why Makin didn't want anyone to hear their organs without the Rotofon - they sounded pretty poor & "dead" with it turned off (as does a custom Makin owned by a friend of mine locally if the reverb is turned off!)  The Compton electrostatic system was arguably the "rolls Royce" of electronic organs in its day - but now sounds quaintly old-fashioned (although to my ears, it has a charm of its own).

Every Blessing

Tony

Barry Williams

#5
The 'Bradford' system needs highly specialised engineers.  Those who perpetrate this system will let no-one other than their own people near it, unless things have changed in the last twelve months or so.

Moreover, even when offered the opportunity to work with an expert voicer, the 'Bradford' people prefer their own non-qualified voicing opinions.  I have chapter and verse on this and on the apparent conflict between the 'experts' on this system.

I tried, many, many times to hear a Makin with the 'Rotafon' turned off, as I have tried to hear a 'Bradford' with the de-tuning turned off.  I have always been denied the opportunity.  Happily, Makins have moved on.

Doug Levey has not responded to my email some time ago about the future of Musicom.

Barry Williams

David Pinnegar

Hi!

I started this discussion in the organ builders' section as it is appicable to pipes and pipe organ builders as well as to the niceties of electronic emulation . . . and it's already apparent that enabling discussion of electronics, far from being deleterous to the pipe organ community, is showing up the particular problems and inadequacies that electronic emulations produce.

Last weekend an organist visited Hammerwood to do a recording of some organ works in meantone. The small chamber instrument is in meantone and it is a charming sound with the pipes full of life but the practicalities of learning foot blowing technique, akin to bagpipes, meant that he went back to "the Beast". With the registrations chosen, however, even with the best of stops chosen and I have deliberately identified stops on the instrument which are capable of reproduction to perfection, the result of the small pipe instrument had charm and life. However, with carefully selected Diapasons and a 4ft Flute which I have specifically voiced with chiff in the style of the French Provencal organs, the effect was delightfully convincing when the other day a musician brought his keyboard masterclass students.

Out of the 200 stops, one needs expertise to choose the right combinations and it's that discernment that an organ builder applies to his specifications making up a modest but musically useful instrument. Discernment is a refined art for which customers of electronics bedazzled by the variety they can offer would show little appreciation.

On the electronic front, the Content midi modules are excellent but the chiff is a semitonal transient on all stops. Thinking this annoying, I listened carefully to the Hunter Diapason finding it accurate. But stopped pipes are not semitonal and emit a quint. Most electronics seem to give purely a quint chiff and not a semitone, or a pink noise "wind" burst. Mention of an enforced phase shift - is that what the Viscount CM100 does when it emits a noise like a skeleton rattling (not that I have ever heard one of those in real life!) ?

Colin - thanks for reasserting Fourier theory basics - of course you're right but of course representation and recreation of inharmonic repeating sounds depends upon an _infinite_ array of harmonics which are not practically available in electronic instruments. Do I recall that decaying sounds require Bessel functions in their analysis?

The importance of non-harmonic components in French reeds is important and my recognition of this caused a a retired Belgian voicer who had worked on Albi to take an interest in my work on the Beast. http://www.youtube.com/watch?v=bi2pdYou-Rs is the result and I wonder how many instruments in the UK which can approach that sort of soundscape?

Finally in analysing pipe resonances, presumably when wind is blown into the pipe, the pipe contains air at a higher pressure than in its static state. Presumably the time taken for that pressure to reach an equilibrium is the time of the transient tone and in a flue pipe, the speed of sound being faster in higher pressure gas would case the pitch to raise. Presumably this would be the reason for the semitonal chiff?

Best wishes

David p

Colin Pykett

Many questions have been posed in the last few posts, and it's not practical to try and reply to all of them.  Some take us considerably off-topic anyway.  It might be thought that, if one could fully understand all the processes involved in pipe speech and one had a digital sound engine capable of reproducing them, then perfect simulation of pipes would be possible.

This might be true, but unfortunately it is of academic interest only.  In practice we do not fully understand all the processes involved at a sufficient level of detail.  As just one example, it is impossible, even in theory, to track the way that the various partials evolve during the attack transient of a real pipe.  This is because of the short duration of the transient, which might be 50 msec for the sake of argument.  The reciprocal of this duration, 20 Hz, gives the minimum frequency resolution one can use in the spectrum analysis necessary to do the tracking.  This is, in general, inadequate for accurate reconstruction of the original sound.

Therefore, even if one had a sufficiently clever additive synthesis sound engine which could be programmed to emit any number of partials, each at any frequency we chose, and then vary the frequency and amplitude of each in any manner desired as they evolve over time, it still could not simulate the exact behaviour of the chosen pipe because that behaviour could not have been measured to that degree of detail in the first place.

Then we have to remember that we do not merely listen to the sound of the pipes alone, except in an anechoic chamber.  In practice we hear them in an ambient environment which adds its own filtering action (to put it crudely) to the sound originally emitted.  Moreover, the ambience is different for each and every listening position in the auditorium.

Consequently, exact simulation of pipe sounds using any form of modelling is impossible, even in theory.  This includes both additive synthesis and physical modelling techniques.  This does not mean that they cannot produce attractive and satisfactory results, but we should not delude ourselves into believing that we are hearing an exact reproduction of the original pipe sound that was analysed to begin with.

This leaves us only with sampled sound organs as the only remaining way ahead.  These simply take a recording of a pipe and replay it on demand.  No prior analysis is necessary, nor any physical understanding of what is going on in the pipe.  A sampled sound organ can also reproduce the ambience of the auditorium if so desired, provided the sound engine is capable enough (several exist today which are).

There are still problems though.  A major one concerns noise on the recorded waveforms, often from the organ blower.  While this can be removed (denoising), the processes involved mean that one is no longer listening to an uncorrupted version of the real thing.

So all current synthesis techniques only offer an approximation to real pipe sounds.  The method one prefers is to some extent a subjective one - some people prefer additive synthesis, others physical modelling, and yet others sampled sounds.  But we delude ourselves if we think perfection has been reached, because (currently) it is not possible to achieve, even in principle let alone in practice.

And all the foregoing has only covered the pipe sounds themselves and how they are reproduced by a digital sound engine.  We haven't even begun to cover other major issues such as loudspeaker distortions, etc, etc, .....

To my mind electronic organs can be a lot of fun, they provide a close approach to real pipe sound in favourable (though not all) circumstances, and they are cheap.  But the pipe organ still reigns supreme in the current state of the art for the most discerning customers.  It disappoints me hugely to see famous musicians saying that "you can't tell the difference between the pipes and the digital sounds".  This has actually been said about a recent large hybrid pipe/digital installation in an important church.  While one accepts that such statements will continue to be made by the manufacturers, I think others need to sometimes reign in their enthusiasms and become more honest.

End of sermon - sorry it was so long.

Regards

Colin Pykett

David Pinnegar

Dear Colin and Barry

I have been piano tuning for concerts in recent days and there is an issue with digital tuning aids which perhaps bears comparison. I find tuners with needles not accurate enough for piano tuning with trichord strings. Of course I use my ears but a meter can provide fast confirmation of just sharp or just flat issues, saving on tuning pin movement, as well as speeding up the setting of the scale in particular. At 440 1 cent is around 1 cycle per second and meter needles really aren't very determinate on a 1 cent basis. But the old analogue tuning aids with rotating LED displays were actual phase indicators so indicating to 1/4 cycle accuracy. So the digital technology often provides an unjustified illusion of accuracy.

Anyway, we should consider moving much of this thread, particularly the last few posts into an appropriate section of Electronic Organs as it's specialising in that direction, and it would be GREAT PLEASE if anyone with more practical and theoretical insight into the practicalities and physics of how pipes speak would contribute.

There are often large numbers of guests looking at this forum and surely others have some insights?

In the world of pianos one has problems with bass strings given extra mass to achieve low frequencies which behave more like metal bars than vibrating strings. In any vibrational system one changes the Eigenfunctions when the width becomes an appreciable proportion of the length. This has similarities to the behaviour of wide pipes and narrow pipes often need quite a bit of wind to get them going, which then results in turbulance at the mouth requiring cheeks or beards to control and to prevent overblowing into the harmonics.

Meanwhile, to what extent on reed pipes, and I'm thinking of French reeds here, how much do tongues set up regularly decaying percussive vibrations of their own independant of the pitch of the pipe?

Best wishes

David P

David Pinnegar

Hi!

http://www.walter-fendt.de/ph14e/stlwaves.htm
is an interesting animation of how a pipe speaks or rather of standing waves in the pipe

This is the mode of the natural pipe but we are blowing air into it and to some extent through it. So this pattern will be compressed from the blowing end which is why the pipe will blow sharp of its natural frequency . . .

http://www.wainet.ne.jp/~yuasa/flash/EngOpenTube.swf is a similar animation

Best wishes

David P


KB7DQH

QuoteThere is a huge amount of literature about the physics of the speech/sound of organ pipes which is readily available.  I get the impression that little new work has been done in recent years.  Perhaps others will correct me.

The following article is somewhat slightly related but proves that others out there are doing some research ;D

The picture in the article is worth having a look at if nothing else ;)

http://news.discovery.com/space/space-music-this-is-the-sound-of-pipe-organs-on-mars.html#mkcpgn=rssnws1

QuoteAs it turns out, Physicist Andi Petculescu and acoustics Professor Tim Leighton have already done just about everything short of actually blasting pipe organs into space.

Hmmm... Kinda puts loading an electronic organ into a Van into perspective doesn't it ??? ;D


QuoteEager to know the sounds of other worlds, the duo used existing atmospheric data to create a computer model of how the atmospheres on Mars, Venus and the Saturn's moon Titan. Then they took earth recordings of "Toccata and Fugue in D Minor" and engineered the audio to correspond with the fluid dynamics of an alien world.

Linked from Petculescu's website on the project, here are the MP3s for Earth,http://peppermintleafresearch.net/Andi/bach_earth_050m.mp3 Titan,http://peppermintleafresearch.net/Andi/bach_titan_050m.mp3 Venushttp://peppermintleafresearch.net/Andi/bach_venus_050m.mp3 and Mars.http://peppermintleafresearch.net/Andi/bach_mars_050m.mp3 Oh, and Public Radio International did a short audio story on the subject back in 2008, which you can listen to right here.http://www.pri.org/theworld/?q=node/22451 Here's what Leighton had to say about it all:

Quote"The atmosphere on Venus shifts the pitch up dramatically (from D Minor to F Minor), making the children sound like Smurfs, while the atmospheres of Mars and Titan transpose the music down (to keys of G-sharp minor and F-sharp minor, respectively), transforming my ten-year old daughter's voice to that of a large adult. However, while the sound on Titan carries even better than it does on Earth, on Mars the atmosphere absorbs the sound so much that almost nothing is audible when you are only 20 meters from the organ. The calculations indicate what the instrument would sound like at various locations in open -- 'air' -- concert halls on the various planets. One thing's for sure -- you wouldn't sell many tickets on Mars!"

Eric
KB7DQH
The objective is to reach human immortality—that is, to create things which are necessary to mankind, necessary to the purpose of the existence of mankind, and which have become the fruit that drives the creation of a higher state of mankind than ever existed before."

organforumadmin

Hi!


A lot of this topic has diverted interestingly to electronic simulation of how pipes speak rather than how pipes speak . . . . and so I have moved the posts veering into the electronic realm to that section.


However, mention of Colin Pykett's simulation of lost organs and ability to synthesise organ specifications before ideas are committed to pipework is particularly relevant to pipe organ builders and anyone thinking of commissioning a pipe organ. That discussion is continuing in the electronic section.


Best wishes


Forum Admin

organforumadmin

Hi!


Just found the following link about flutes . . .


http://www.phys.unsw.edu.au/jw/fluteacoustics.html


Best wishes


Forum Admin

Janner

If, to the knowledgeable, this question appears completely silly, then please forgive my ignorance in asking it, but I have searched through the various links above and found no reference to it.

One often sees organ pipes bent in various directions, presumably in order to accommodate them in restricted spaces. In some cases the bends are quite severe, to the extent of the pipe doing a complete U turn. Does this affect the sound of the pipe?

This link, http://www.phys.unsw.edu.au/jw/flutes.v.clarinets.html#time related to the one above, appears to give a good demonstration of how a wave travels up and down a pipe. However, it occurs to me that if an open pipe is bent, then the two ends are no longer in the same positions relative to each other externally, as they would be in a straight pipe. Or is this technique only used in stopped pipes?

Even if it were only used in stopped pipes, does the bend cause more energy to be absorbed in maintaining the wave? If so, would the effect be of any practical significance?

Once again, apologies for my ignorance. Any thoughts or explanations appreciated.

Colin Pykett

It's not a silly question at all, and not easy to answer either, well succinctly anyway.  Broadly speaking the principle is that any wave, be it electromagnetic or acoustic, will take progressively less notice of a bend (or any other form of discontinuity) as the wavelength gets longer in relation to the dimensions of the bend.  As examples, the huge wavelength of the Droitwich long wave radio transmitter at 198 kHz (1515 metres) propagates serenely through walls, huge buildings, and just about everything else provided they are non-conductive.  It doesn't notice their existence.  But we all know about the issues of UHF TV reception at c. 800 MHz (wavelength c. 38 cm) which depend critically on aerial position and on small intervening objects.  The same issues apply to the similar wavelengths used by mobile phones, wi-fi, etc as we all know to our infinite frustration.

It's much the same with an organ pipe.  If the bend starts and ends within a distance which is small compared to the largest wavelengths (lowest frequencies) of the pipe, these will not be much affected.  But its higher harmonics (smaller wavelengths) might be.  This could result in some changes in tone quality.  However most of the time they will not be noticeable for the simple reason that the pipe cannot be voiced until it has been made, bends and all, so the voicer will automatically take care of the problems when he does his thing.

Colin Pykett

KB7DQH

An intriguing analogy Colin brings up...  The one similarity between RF antennas and flue organ pipes is the need to deal with "end correction". Oddly enough the thickness of the conductor relative to its length has a similar effect on this "end correction" in RF antennas as it does with flue pipework ??? 8) ;D ;)   A "thicker" conductor will need to be made shorter than a thin one to obtain resonance at the same frequency, and also the usable fundamental bandwidth of a "thick conductor" antenna will be broader than a thin one...  Similarly the air density can be related to the "dielectric constant" the antenna operates in... An antenna in the vacuum of space on a orbiting satellite will have to be cut slightly longer in order to resonate at the same frequency  as a similar device operating within earth's atmosphere... It gets more extreme when dealing with devices designed to operate, say, within the bowel of a cow... I actually saw such a device shown to me by the engineer who designed it... the physical size of the antenna for its design frequency ended up being considerably shorter. 

As to "mitering" pipework to make it fit a smaller space and its effect on tonal quality?  Similarly, one can "load" an antenna to reduce the physical space it occupies, and the example of the 1515 meter longwave frequency is a fine example of how extreme this becomes in the realm of electromagnetic waves.  An "efficient" marconi antenna would need to be nearly 400 meters long... but placing an inductive load somewhere in that length, depending on its value will reduce the length needed to achieve resonance...but at a sacrifice in efficiency and bandwidth.   Although mitering doesn't actually change the physical length of the pipe for a given musical note, I suspect (not having worked with organ pipes) that the harmonic development would be altered by the differences in wall length created by the mitering if the pipe were "coiled". A pair of "folds" in opposing directions would tend to cancel that effect I would think.  Depending on how much sound is emitted by the mouth versus end of the pipe, (in the case of flue pipes) and the relationship after mitering could have a noticeable effect on how the pipe speaks into the space it occupies...  But as Colin points out
much of this can be corrected in the voicing process to one extent or another.

Yet another confusion of RF and audio presents itself and will lead this discussion in yet another direction :o :o :o  and this concerns the "tonal effect" if any, on non-speaking facade pipework ;) ;D

When dealing with RF antennae, it must be noted that nearby conductors which resonate close to the operating frequency of an antenna will affect the radiation pattern in some way, and in some cases destructively... and in other cases constructively... Yagi and Uda, of Japan, invented an antenna that most marvelously exploits these phenomena to produce directive patterns through careful arrangement and tuning of "parasitic" elements around a "driven" element. 

To what extent has "research" been done to determine  if energy is coupled from speaking pipes into non-speaking pipes (either in the facade, or simply neighboring ranks) and what positive or negative effects this has on the tonal scheme of an organ?

Eric
KB7DQH
The objective is to reach human immortality—that is, to create things which are necessary to mankind, necessary to the purpose of the existence of mankind, and which have become the fruit that drives the creation of a higher state of mankind than ever existed before."

Colin Pykett

"To what extent has "research" been done to determine  if energy is coupled from speaking pipes into non-speaking pipes (either in the facade, or simply neighboring ranks) and what positive or negative effects this has on the tonal scheme of an organ?"

I felt an instant urge to respond to this because of a strange coincidence that I was thinking of exactly the same thing a day or two ago - don't know why though! - although it is an issue which crops up from time to time.

As Eric will know, the issue is one of coupled resonators.  The closer they are in tune, the stronger the parasitic (receiving) one will resonate and the more energy it sucks out of the transmitting one.  I doubt it's much of an issue though for organs.  However it is used in architectural acoustics in treating seats in an auditorium so they are acoustically the same whether people are sitting in them or not.  Or in the wall treatments which sometimes incorporate resonant cavities to control reverberation.

What does matter for pipes in close proximity in organs, though, is if their mouths are too close.  Then they can pull each other into phase-lock and you then lose the desirable chorus effect of multiple separately tuned pipes.  It can make them difficult to tune also.  So they are often staggered in height on a soundboard to increase the distance between their mouths.

Another issue is that phase cancellation can occur so that you hear nothing at all in certain circumstances!  See:

http://physicsbuzz.physicscentral.com/2009/09/surprising-physics-of-pipe-organs.html

But, again, this isn't due to coupled resonators as such.  More likely to be due to coupling via the wind supply or because the mouths are close as per the above.  Such coupling can encourage the two pipes to go into a minimum-energy speaking mode.

Colin Pykett

Janner

Gentlemen,

Thank you for your explanations, and for the associated links.

Colin, your example of the TV reception is very apt; I just had not applied the principle to organ pipes, and I take your point about the voicing.

The link to Physics Central is extremely interesting. For me the demonstration further down the page, using the five metronomes, is excellent; thank you for that. In another time and place, it might be interesting to experiment with things like putting liquid of varying amounts in the drink cans, or setting one of the metronomes to double or half the rate of the others, but I digress.

Quote from: Colin Pykett on May 23, 2011, 07:00:19 PM
What does matter for pipes in close proximity in organs, though, is if their mouths are too close.  Then they can pull each other into phase-lock and you then lose the desirable chorus effect of multiple separately tuned pipes.  It can make them difficult to tune also.  So they are often staggered in height on a soundboard to increase the distance between their mouths.

Most interesting.

Thank you again for your replies.

Colin Pykett

I don't want to be guilty of the worst sin of any forum - never knowing when to stop!  But on the subject of bends in organ pipes, it did occur to me to suggest that one might soft-pedal the theory and look instead at some of the extreme examples which one gets in brass orchestral instruments.  The trombone for example doubles back on itself just as some organ pipes do.  And valved horns, euphoniums, tubas, etc - well!  Not only is the main bore multiply coiled, but the plumbing around the valves has short pieces of pipe bent sharply into all conceivable shapes.  And the remarkable thing is that it makes not the slightest difference to sound propagation, because of the indisputable fact that these instruments actually work.

Some organ pipes are also multiply coiled, such as the lowest notes of 16 or 32 foot reeds.  In these cases I am more impressed by the quality of the mitreing and soldering and the pipe maker's art in general, rather than worrying too much about the theory!

Thanks 'janner' for making us think about an interesting issue.

Colin Pykett

David Pinnegar

Hi!

From the point of view of horn speakers, provided bends are gentle or mitred there is arguably not much difference. But if the joints are square without reflectors, then it is said that high frequencies don't go around them, so they act as low-pass filters. Mitring organ pipes therefore is key unless we're looking at flute tones only with fundamental notes.

Best wsihes

David P