Question for physicists and acousticians

[Stanley Monkhouse]

Some years ago I arranged in advance to play the instruments of Norden, Neuenfelde, Meldorf, and Stralsund, and visited a couple of others ad hoc. It cured me of organ crawling—nothing else would ever match up.

This, and experience over the years, raised a question. I doubt that there was much in the way of central heating when Schnitger, Stellwagen and the like were plying their trade, and the large brick barns of the Baltic coast can’t ever have been warm. I have the impression that organs sound their best in the cold. Am I deluded? Of course, the paraphernalia of comfort like carpets and soft furnishings affect acoustics, so maybe it’s just this, but nevertheless I ask the question: does temperature affect our perception of sound? (I’m not talking of tuning). Do organs sound better in freezing churches?

My other observation may already have been discussed on this board, but FWIW I was bowled over by just how exciting Buxtehude, Bruhns and Tunder were on the unequal temperament organs. Astonishing. Equal temperament does not do them justice. There is no doubt in my mind about that.

[Damian Beasley-Suffolk]

I can't say anything about temperature, although I often wonder whether humidity might play a role in sound production and propagation, but you're right about temperament.

In 1998 I went to the re-opening recital of the restored 1446 van Hagerbeer organ in the Pieterskerk in Leiden. After many changes it had been returned to something like its state in 1643, with meantone tuning, and a' = 417 Hz. The last item on the programme was Sweelinck's Chromatic Fantasia. That was an experience! As the piece progressed, ears bled, teeth ground, and finger nails dug into chairs - but the final resolution was almost a physical relief. Not that it was bad - it was simply an extraordinary experience that few had had, me included. The tuning somehow added an extra edge and even (perception of) power to the instrument, and certainly demonstrated changes in emotion and waves of build and release of physical tension as the piece went through the keys. This would be very close to the experience you must have had listening to Buxtehude, and it is exciting.

Coincidently, over the weekend I was playing with different temperaments on my electronic, playing classical French music with Dom Bedos, Corette etc. As I don't know what I'm doing, it's a bit hit and miss, but I'm at least beginning to get a feel for the colours these temperaments provide.

[Colin Pykett]

The temperature factor is in fact related to humidity, as Damian said.  Recording engineers have noticed that "it could be the difference between two identical concerts, where one sparkles and has more brilliance than the other.  Yet the same orchestra performed them in the same hall with the same exuberance and skill - only the weather was different".  This quote is from a paper read at the Audio Engineering Society in 1988.  It is most noticeable in very large auditoria where the climate is ill-controlled (thus organs in cathedrals with no heating) or open air ones such as the Hollywood Bowl.  And as Stanley said, it's a different thing to tuning as orchestras always play in tune, well at least, they don't allow temperature to affect their mutual tuning over-much.

The explanation is a bit complicated.  Above freezing point, the air holds moisture as water vapour to some extent (if there is any), i.e. it is humid, and this influences how sound dies away as it travels away from a source.  It dies away more rapidly for higher frequencies than lower ones.  However when the humidity changes, this difference in propagation between the highs and lows (treble and bass) becomes more or less pronounced depending on whether the humidity is rising or falling.  But humidity also tracks temperature - when temperature rises the humidity also rises because warm air can hold more moisture, and vice versa.  Consequently a change in temperature results in a change in humidity which results in high frequencies being attenuated more or less relative to the lower ones - a sort of temperature-driven tone control effect.

--------------------

When discussing temperaments don't fall into the trap of believing that because a temperament is unequal it therefore must always have a better range of key colours than equal temperament (ET).  This might be true for some, but it is not so for all.  The important class of meantone temperaments is an example (e.g. quarter-comma as often used by Schnitger & Co, 6th-comma a little later and further south by Silbermann & Co, etc).  In all these, 11 out of the 12 fifths which define the temperament are deliberately detuned by the same amount (e.g. one quarter of a syntonic comma, one sixth of a Pythagorean comma, etc).  Qualitatively this is not much different to ET where all 12 are detuned by the same amount (one twelfth of the Pythagorean comma).  Thus the only difference of principle between the meantone temperaments and ET lies in only one fifth, which is horribly out of tune (sharp) in the meantones and thus called the Wolf.  Therefore if you avoid keys in which the Wolf appears, you are playing only on a restricted set of 11 equally-tuned 5ths, resulting in these keys being a form of equal temperament.  Thus the 'good' keys in meantone temperaments are no different to the keys in ET in the sense that there is no difference in key flavour between them.  One only gets a few, pretty gross, differences in key flavour when one moves across the boundaries between the groups of 'good', 'poor' and 'awful' keys.

I mention this because it is an objective phenomenon related only to physics, as Stanley requested.  But it is a factor which does not seem to be well hoisted-in by some of those who prefer unequal temperaments.  I won't comment on the subjective issues.

As Damian also said, digital instruments are great for getting to grips with temperament if it's something you want to do.

[Contrabombarde]

Utterly fascinating Colin as ever - so are you saying the best conditions for listening to organ music are a cold wet day or a cold dry day?

My grand piano (a 1900 Bechstein, but with recently rebuilt action and new hammers) definitely feels lighter and more inviting to play when the living room is on the chilly side and except in mid summer I try to keep the room at around 17 or 18 degrees C for that reason. That's probably as much to do with the action though.

[Colin Pykett]

18 hours ago, Contrabombarde said:

Utterly fascinating Colin as ever - so are you saying the best conditions for listening to organ music are a cold wet day or a cold dry day?

My grand piano (a 1900 Bechstein, but with recently rebuilt action and new hammers) definitely feels lighter and more inviting to play when the living room is on the chilly side and except in mid summer I try to keep the room at around 17 or 18 degrees C for that reason. That's probably as much to do with the action though.

The issue is to keep humidity as low as reasonable, and if air conditioning incorporating full climate control is available it can be adjusted.  But to answer your question, a dry day ought to be better, since it's the increased density and higher thermal conductivity of high-humidity air which both conspire to kill the high frequencies more than the low ones.  I doubt there would be much difference noticeable due to the humidity effect in small rooms though.  It's only where the path from the pipes to one's ears is long enough such that the differences in attenuation with frequency become significant.  In a large building with a long reverberation time (several seconds) the multiply-reflected path lengths contributing to the later parts of the reverb tail would correspond to thousands of feet since the speed of sound is about 1100 ft/sec, and over these distances the effects we are talking about here would definitely become noticeable.  A 5 second reverb implies path lengths due to the multiple bounces approaching a mile or more!  Clearly these could only exist in large buildings such as cathedrals, and it is in these where the effects described by Stanley in his original post can be noticeable.

It begs the question of whether performers ought to put a rider into their recital flyers along the lines of "weather permitting".  Then they could cancel it if the temperature and humidity turned out to be out-of-spec.  (No, not really ... ).  However I'll make a note in my diary to serve this up again on 1st April to see if forum members have forgotten about this thread in the interim.

[Stanley Monkhouse]

Last para: brilliant! This could be another reason why Buxtehude et al sound better in winter in North Germany/Denmark than they do here.

[Colin Pykett]

Well, I did used to work in sonar ...

[David Pinnegar]

I disagree with the expressed opinion on Meantone that it's hardly different to Equal. Couperin deliberately used sour combinations of sounds so as to create crisis points from and to which his music lurched. The opening Kyrie of the second of the Masses for Parishes and Convents is a very good example and those with electronic laboratory facilities might well try it both in  Meantone and Equal. Meantone creates sounds which are wonderfully on edge whilst Equal merely sounds nice. Sanitised.

Best wishes

David P

[Colin Pykett]

I think you might have missed the points I was making.  These points are all objective, not subjective, and based on nothing more than arithmetic.  What I was trying to point out (quite probably inadequately) is that, in terms of the practicalities of tuning, there isn't much difference of principle between the temperaments.  In both meantone and ET one tunes 11 fifths so that the frequency ratios of their constituent notes are the same.  The actual ratio used depends on the temperament you are tuning, the values in terms of commas being those mentioned in my earlier post.  The arithmetic then means that the remaining fifth (the 12th fifth which completes the circle and doesn't need to be tuned) automatically becomes a Wolf in meantone, whereas it automatically stays the same as the other 11 in ET, i.e. all 12 have the same inoffensive ratio with no Wolf.

What I then went on to say was that if you play in the restricted set of 'good' keys of meantone you are not using the  Wolf interval.  You are only playing on the notes which constitute the 11 other fifths.  If you were to use keys which included the Wolf you would not get as many of the lovely thirds, but dissonances instead which some people regard as hideous.  Because the 11 non-Wolf fifths whose notes constitute the 'good' keys, all having the same interval ratio, they form one of the class of equal temperaments.  Because it is equal there is no discernible key colour between these 'good' keys.  However, to say otherwise is understandable because it is a widely held misconception, but it is nevertheless a misconception.

Only when you stray beyond the good keys into those which call on the Wolf interval do you get a sudden change in key colour - indeed the injection of 'sourness' as you put it.

I know that it can be difficult to get one's head around this, but please bear in mind that it's not just little old me who makes these points.  In the mid-1980s the late Charles Padgham in his well known treatise ('The Well-Tempered Organ') said of quarter-comma meantone:

"It is sometimes forgotten that mean tone temperament is a form of equal temperament except for one fifth, the Wolf.  All the other fifths have identical tempering.  Thus the 'good' keys are identical except for pitch, and true 'key flavour' does not exist among these good keys as it does in the irregular temperaments."

And for sixth (Pythagorean) comma mean tone - the so-called 'Silbermann' temperament - he said again:

"There is also no different key flavour amongst the good keys".

In more recent times Fred Sturm is just one well known writer and professional piano technician (a name which scarcely does him justice I might add) who has made the same points, and (as I have done here) he has also drawn attention to the widespread misunderstandings which have resulted in many people holding the view that the meantone temperaments possess a wider range of key colours than they actually do.

Oh, and while we are at it, let's dispose of another myth.  It is often claimed that ET was slow to catch on because it is difficult to tune.  That's true - one has to carefully temper 11 fifths.  But all the meantone temperaments also require that 11 fifths are carefully tempered, so they are just as difficult to tune as ET.  The one exception is that quarter-comma meantone has several pure major thirds and minor sixths which can be used along the way to check that your fifths are accurately tempered.  I maintain that's one reason why quarter-comma might have lasted for as long as it did.  (All this, of course, refers to the ancient methods for tuning by ear and without timing beats, not using electronic tuning devices).

Having said all this, I am completely with you in trying to persuade people to explore the beauties of unequal temperaments.  In this aspect we sing from the same hymn sheet.  But at the same time one has to be a bit careful that we do not help to propagate long-held and widely-held misconceptions such as these, which unfortunately are demonstrably wrong.

 

 

[David Pinnegar]

But the key colour doesn't come from the fifths. It comes more from the thirds and in 1/4 comma there are four of them so there's quite a lot of intervals to hit which tweak the ears - 

and one of the most extraordinary pieces to play in meantone is Mozart's 2nd piano sonata. The middle movement is an exploration of everything sour, sorrow, frustration, anger, inconsolate grief, and everything about being buried in the grave and coming to life again in the third movement. The Mozart piano sonatas are particularly revealing in meantone.

It's not just the fifths nor just the thirds either, those close and wide semitones too. 

I'm sure I've mentioned before Mozart's fantasias for mechanical clock

and which performed so fit the character of the context for which they were written - https://www.academia.edu/37951978/THE_COLOUR_OF_MUSIC_IN_MOZARTS_TIME_A_journey_from_Couperin_to_Chopin_Examination_of_reconstruction_of_Mozart_Fantasias_K594_and_K608_for_Mechanical_Clock

Best wishes

David P

 

 

[Colin Pykett]

As I said in my last, we're actually on the same page of the hymn book.  I see exactly what you are saying, though remain unconvinced that I've explained myself adequately in return, for which I apologise.  But I'm not sure there's much to be gained by pursuing it any further, at least from my point of view.  I think I'll spare the forum any more grief!

[pwhodges]

21 hours ago, Colin Pykett said:

The issue is to keep humidity as low as reasonable, and if air conditioning incorporating full climate control is available it can be adjusted.  But to answer your question, a dry day ought to be better, since it's the increased density and higher thermal conductivity of high-humidity air which both conspire to kill the high frequencies more than the low ones. 

Wrong way round; the absorption of sound reduces with increased humidity as water vapour is less dense than air.

Simple account

Scholarly account

Calculator

Which is good, because most instruments have wooden parts which suffer if kept too dry!

Paui

[David Pinnegar]

High humidity seems to lead to fog . . . so I did a quick google and https://www.sweetwater.com/insync/effects-of-temperature-humidity-live-sound/ seems a very rational explanation.

Best wishes

David P

[Colin Pykett]

1 hour ago, pwhodges said:

Wrong way round; the absorption of sound reduces with increased humidity as water vapour is less dense than air.

Simple account

Scholarly account

Calculator

Which is good, because most instruments have wooden parts which suffer if kept too dry!

Paui

I thought this might come up!  I've had to explain it often in the past.  There are three issues:

1.  Absolute humidity is not the same as relative humidity (RH).  RH is the amount of water vapour present in air expressed as a percentage of the amount needed for saturation at the same temperature.  Assume we have a volume of air with a given mass of water vapour in it i.e. the absolute humidity is constant.  Now vary the temperature.  At low temperatures the RH becomes higher than at high temperatures because at low temperatures the air saturates more easily, even though the amount of water (absolute humidity) remains the same.  Put another way, the same amount of vapour gives higher RH in cool air than warm air.

2.  Sound attenuation at a given frequency (e.g. measured as dB per metre) gets higher for low RH (warm air) and lower for high RH (cool air).  Lots of references here going back to the year dot - e.g. Evans & Bazley 1956.  So there is less attenuation if the temperature drops.

3.  Attenuation at a given RH gets lower as frequency gets lower and vice versa.

Upshot: Attenuation is low for conditions of high RH and low frequency, and high for low RH and high frequency.

Therefore as I said previously, one gets more treble (high frequencies) if the RH is high, which for constant absolute humidity, means at low temperatures.  Thus music sounds more brilliant in the cold, as Stanley and the recording industry have observed.

I haven't looked in detail at the links you gave as unfortunately some of those I have looked at in the past have got it wrong.  Some of what's in Wikipedia is also wrong.  I think the main problem is the counter-intuitive relation between absolute and relative humidity and their variations with temperature. 

[pwhodges]

OK; that makes it clear that our difference is more because you are addressing changes in temperature with constant (absolute) humidity, rather than my thinking of changes in humidity with constant temperature. 

For interest, note also that figure 2 in my second link suggests that at frequencies below about 300Hz and 20°C, the absorption is higher in the very driest air - but RH that close to zero is not common...

Paul

[Colin Pykett]

Actually Paul, looking back over this, I didn't really answer your question properly - apologies.  If we go back to Contrabombarde's original question:

On 08/01/2020 at 20:23, Contrabombarde said:

Utterly fascinating Colin as ever - so are you saying the best conditions for listening to organ music are a cold wet day or a cold dry day?

I assumed (without thinking about it too much) that he was talking about rain - liquid water rather than vapour.  So I assumed this would increase the density of a given volume of the atmosphere.  However if I'm honest I don't really have much of a handle on what this might do to sound passing through it as there are too many variables (precipitation rate, etc).  And in any case we wouldn't actually have rain in the auditorium housing the organ even it was raining outside - unless the roof leaked badly.  But the presence of rain would also saturate the air with vapour, probably inside the building as well as outside, which brings us back to where we came in.

Re your further point about Figure 2 in your link, it illustrates the anomalous variation of absorption for damp as opposed to dry air.  Though on first sight these (theoretical?) curves don't seem to correlate too well with the experimental data of Evans & Bazley I mentioned earlier.

But we've probably gone way beyond what Stanley Monkhouse had in mind when he set us this disarmingly simple exam question in the first place ... so I think I'll draw stumps for now.

[Stanley Monkhouse]

Colin, I didn't have anything in mind except my "feeling" that organ music in cold churches seemed more satisfying than in warm ones. There might be something in it. It's difficult untangling physics from neurobiology - ear, cochlea, eighth cranial nerve, auditory pathways, emotional perception. It leads me to another question that physicists might illuminate, but I hesitate to voice it until this one has died down. Thank you all for your expertise and erudition.

[OwenTurner]

Probably you all know, but I remember getting very confused when studying acoustics at university; attenuation is a measure of sound energy loss, so higher attenuation means more loss at a given distance. I thought of it the opposite way for quite a while. Hopefully I've got it right now!