Why older halls for symphony orchestras sound better than new ones?

According to Witold Rybczynski, it is partly a matter of size and shape.  Older halls were more likely in a shoe-box shape, such as Musikvereinsaal (1870),the Concertgebouw (1888), and Symphony Hall (1900).  The orchestra is at the narrow end of the hall, and “Sound is reflected to the listener from the two parallel walls (which are about sixty to eighty feet apart) as well as from the ceiling.  Because the concertgoer is relatively close to the musicians, the atmosphere is intimate, visually as well as acoustically.”

The need for greater seating capacity, because of revenue, swells the halls to as large as 3,000 seats and renders these properties more difficult to achieve.  Most of all, it is now difficult “to reflect bass notes from the side walls.”

Different halls also have different “reverberation” times, and thus the more that a hall is used for different kinds of music, the more the design of that hall reflects a compromise.  In theory there should be different halls for playing Gorecki and Bach, so an all-purpose hall will sound ideal only rarely.

That is all from his new book Mysteries of the Mall and Other Essays, pp.188-192.


so the answer is simple acoustics and physics. such brilliant insight

It's even simpler. The old halls are smaller, so they sound better.

I bet it's even simpler than that. The old halls are old, so people think they sound better. If the experiment could be done blinded, it would not have the same results.

+1. The idea that 19th century orchestra hall architects were more knowledgeable about, or conversant with, acoustics than their modern counterparts doesn't pass the sniff test. Older halls are more familiar, and people like the familiar... much like the preference for vinyl records.

Or, the far simpler explanation: survival of the fittest. Those that are most desirable will be the most profitable or most used and thus most preserved.

Those that aren't will be changed in attempts to improve them, or used for other purposes, or torn down.

The number torn down over the past five hundred years is certainly a very large number.

And just because a venue is old does not mean it's acoustically good, much less great; it might simply be versatile.

Sophisticated knowledge of acoustics has been around for a long time: http://www.nature.com/news/2007/070319/full/news070319-16.html

19th century architects certainly knew enough to account for this in their designs, and in the days before amplification it was much more of a priority to get acoustics right. So, it is completely plausible that old halls really do sound better.

Also, the size factor should not be underestimated. An orchestra's sound will fill a smaller venue much better than a several thousand seat venue.

Do farts sound better in such halls? How about their smell?

He must be using the term "parallel walls" liberally -- if the walls were indeed parallel, how could there be a "narrow end"?

I applaud the idea of using "liberally" to mean erroneously. Bravo!

Reminds me of an old math joke: 1 + 1 = 3 for sufficiently large values of 2.

Err 1 rather ...

The joke sort of works both ways, doesn't it?

Using the singular "the narrow end" when there are always 2 narrower ends would be a "conservative", and penurious, use of the term, and of course, technically improper.

La Scala in Milan is horseshoe shaped.

That puzzled me for a bit, until I figured it meant the shorter side of a rectangle.

A rectangle has two shorter sides. (Unless it's a square.)
The BSO Hall's reputation for its outstanding acoustics is VERY longstanding. Same for the 'shoebox' scheme. Part of BSO's success is the dispersion effect of its classical surface ornamentation.
But the "the flying saucer thingumyjigs hanging from the ceilings" and elsewhere can acoustically simulate much of that.

Yes, obviously shorter side of a rectangle. As for there being two, pick one. It should not really matter.

Is it axiomatic / unanimous that older halls sound better?

Have you heard music in them?

If you find yourself near one, it really is worth the expense of a ticket. I don't know the European halls mentioned, but I've heard a bunch of BSO concerts at Symphony Hall, and no other North American hall I can think of comes close.

And while historical-authenticity arguments should be treated with grains of salt, there's no question much of the symphonic repertoire was written with smaller halls in mind. You hear the music differently.

I'm naive. My ear is unsophisticated. Concerts even at the poor University Hall sound divine enough to me.

My problem (?) always has been that compositions by the masters are so damn good that they overwhelm any difference in conductors, venue etc. Maybe I just need to get my senses more blasé.

.. that just isn't true. In the last few years I have head Beethoven and Wagner in The Albert Hall in London which has glorious acoustics and is as good as the Musikverein where I heard Bruckner and Schubert also glorious. On the other hand how can we tell the difference since today's performances are all tarted up with electronics and the flying saucer thingumyjigs that are hanging from the Albert Hall ceiling ...

All of this is received wisdom on the subject, and follows well understood physical and physiological acoustic principles. In fact, I think Rybczynski himself may have mentioned it in an earlier book, though my memory is not good enough to dredge up a reference.

When good acoustic is the primary goal, modern builders have no trouble matching or excelling the older halls. There has been a backlash against large but terrible-sounding halls, so it is not really the oldest halls that sound best and the newest worse. Rather, there was a trough in sound quality about 40-50 years ago. The newest halls are the best of all, in my experience.

Here in Toronto, for example, our largest hall is Roy Thompson, seating about 2,600. From the air it looks like a concrete hat box and from inside, it used to look like a cavernous concrete barrel and sound like one too. They have since performed renovations to improve the sound, mainly an acoustically reflective arrangement that effectively drops the lofty ceiling but also paneling on horizontal surfaces and, very importantly, behind the stage itself. The inability of musicians to hear each other properly caused as much trouble as the direct inconvenience to the audience.

The Weston Recital Hall, by contrast has a superb acoustic, though this carries its own hazards: you can hear another patron cough or unwrap a candy at 50 meters. But it seats only around 1,000 and was built in 1993.

The Glenn Gould Studio, also built in 1993, has the best sound of them all, but seats only a little over 300. (There are also confounding factors: for noise isolation, the Glenn Gould is situated at the core of the much larger CBC structure, which itself sits on damping rubber pads.)

Koerner Hall sounds pretty nice (as nice as the Weston, to my ears). It's pretty boxy.

I was at Royal Albert Hall recently, and the differences between boxes were profound. We started in a box close to directly opposite the stage, and I heard the London Philharmonic well. However, the concertgoers in the box adjacent to ours were drinking and kept running in and out of the box, and so my wife asked for us to be reseated. We were, to another box more along the side. From there, I heard the distinct echo of the orchestra off the back of the hall.

So yes, size matters in these sorts of engagements.

If you are interested in measurement and physics of sound, this is a great free program:

Room acoustics is a transfer function: If I play frequency x, what do I hear at various points in the room, for all x.

When it was originally commissioned, the Royal Albert Hall had such a persistent echo effect that it was said to be "the only hall where a British composer could be assured of hearing his music twice."

"Why older halls for symphony orchestras sound better than new ones?"

This is actually a line from the upcoming film, "Tonto, Tarzan, and Frankenstein Get Cultured".

Okay, I finally get the joke. That is pretty funny.

Me no get it.

The question is missing 'do' and so sounds like how Hollywood writes characters that aren't supposed to be fluent in English.

There's a certain amount of selection effect: Old concert halls with famously good acoustics tend to get maintained, while bad ones get replaced or renovated.

But this can take generations. When I lived in Chicago, it was still said that the Chicago Symphony Orchestra had made a mistake in 1904 by moving out of the Lyric Opera's theater, with its fine acoustics, and into crummy-sounding Orchestra Hall. From Wikipedia:

Acoustical History

Suboptimal acoustics within Orchestra Hall have been an ongoing concern throughout its history, and have been adjusted in major overhauls of the main hall in both the 1960s and as part of the Symphony Center transformation between 1995 and 1997.[4]

Critical reaction is that the 1995-97 acoustical revamp was largely successful, though with room for further improvement, particularly in the upper registers.

Chicago had two popular rock concert venues near my old condo in Uptown, the Aragon ballroom and the Riviera. They are conveniently located near the El, bus routes, and Lake Shore Drive, so they get a lot of business. But at least in the 1980s and 1990s they had terrible, terrible acoustic. For example, every Ramones song sounded exactly the same except for the minimalist I Wanna Get Sedated. The only bad so musically adept that they sounded good was Los Lobos.

Recently, I was looking through a book about the history of jazz in Chicago and there was a quote from my late father-in-law, the tuba player for Chicago Lyric Opera, saying that when he played tuba in big bands at the Aragon Ballroom in the 1940s, the acoustics were miserable, and oldtimers told him they'd been that way since the 1920s.

So, try hard to get the acoustics right when first constructing an auditorium because, historically, it's proved hard to fix afterwards.

TBF every Ramones song *does* sound the same

Problem also applies to chamber music settings, where venues are often way too large for the music being performed. I can only enjoy this music up front, loud, intimate.

Wigmore Hall in London is another extraordinary hall and, again, not too big. I heard Alice Coote in April and thought she was looking at me the entire time, as I'm sure did everyone else.


The Los Angeles Coliseum has terrible acoustics for football - it isn't loud, so the opposing team is able to hear each other and run plays. The article gives an example of 79.5 decibels there, and 136.6 in Seahawk statium, a 56 db difference!

The Coliseum is big enough for a 400 meter track. It hosted the 1932 and 1984 Olympics and might get the 2024 Olympics. But track doesn't play nice with other sports, so the Coliseum isn't well-suited for football. I can remember sitting at about the negative 30 yard line for a UCLA-USC game.

In general, multi-sport stadiums are out of fashion.

As are high capacity stadia. Los Angeles has a number of old fashioned very large capacity stadia: the Rose Bowl, Coliseum, Dodger Stadium. The trend in recent decades has been to build smaller arenas and charge more per seat.

Why fleece a city's taxpayers for one new stadium when you can fleece them for three?

Largest attendance ever at a World Series game, 1959 Dodgers vs White Sox game 5, over 100,000 at the Coliseum. Never to be surpassed.

Meanwhile all the stories you've heard about how great the acoustics are in Greek and Roman theaters are true. I regularly shock students when we visit Pompeii's Odeon by delivering a dozen lines or so of the Aeneid in an easy speaking tone while they sit on the back rows and hear every syllable -- when you use languages with final-syllable inflection you work hard to make sure that every syllable comes through.

I first experienced this myself in the Roman theater in Orange, France. I've heard the sound is even better in Epidauros.

How about a utilitarian take? If you have to accept just 80% of the quality, but get to triple the number of audience members, aren't the larger modern halls better?

Interesting. But my sense is that Carnegie Hall, which still sounds great, was noticeably less acoustically beautiful after the renovation -- without any change in the basic structure.

My undergraduate thesis at Brown (unpublished, of course - I was an undergrad!) was titled: "Modifying the Sound Field of Almunae Hall to Improve the Spatial Impression". The idea was to create artificial (using digital delay lines and an array of loud speaker) early "reflections" off of the side walls, thus effectively narrowing the width of the room. I didn't have enough time/data to show a statistically significant result, but there were hints at a reliable signal.

The theory (promulgated by Yoichi Ando and Manfred Schroeder, among others) was that a contributor to a positive subjective impression in a concert hall was for the first reflections to reach the listener to be coming from the side walls, rather than the ceiling. The reason: whereas ceiling reflections arrive at the two ears virtually simultaneously, lateral reflections arrive at one ear earlier than the other, so that the signals are less correlated, thus contributing to a spatial impression, essentially what one gets from a stereo recording (vs. mono). This, combined with masking (the tendency of earlier arriving sound to decrease the salience of later ones, at least for some refractory period), thus favors narrow/high halls, at least at typical dimensions (reflections from really high ceilings are potentially not masked, thus reducing non-correlation of the two ear signals).

Measurements, using dummy heads, with in-ear microphones, of some prominent "good" and "bad" (by reputation) halls confirmed the core physics here - that narrower halls created less correlated inter-aural signals, but connecting this to psycho-acoustic preference proved difficult, even for non-undergraduates! This is likely because there are many other components of the subjective experience - S/N, reverberation time, transfer function, visual queues, etc. - and disentangling the various sources (not to mention quantifying preference) is difficult and noisy. The theory has not been disproved, however, and seems plausible to me.

The old halls are prettier, so the music sounds better. (For those who are a little slow to recognize irony, it's all in your head.) It is not believable that we have actually lost knowledge of acoustical engineering since the 1800s.

Comments for this post are closed