What’s all the fuss about pitch?

Over the past decade the tango community has become aware of the topic of musical pitch in the transfers, and even in the manufacture, of old records. In brief, if you turn the record more quickly, not only will the tempo (the number of beats per minute) rise, but the pitch will also. Turn the record more slowly, and the tempo and pitch both fall.

A 10″ diameter shellac record from 1934. This is Canaro’s “Un jardín de ilusión” but issued in Europe by Parlophone

In the ideal world, the old shellac 78rpm discs were recorded, manufactured and played back at exactly 78rpm (revolutions per minute) – okay, actually 77.92rpm if we want to be precise. This was achieved through the use of a special motor locked to the 50Hz frequency of the mains electricity supply. In the real world things were more complicated. Various problems could affect the speed at which the physical master disc was recorded. Furthermore, when the record companies first made LP compilations of old tangos in the 1960s, they decided to speed them up a bit to make them sound more exciting, as well as adding reverberation and even echo (hey, it was the 60s!).

An example: Juan D’Arienzo’s Pensalo bien (1938)

In 1997 Sally Potter used the tango Pensalo bien for a memorable scene in her film “The Tango Lesson”. Here is what BMG (owners of RCA-Victor) presented us with, taken from the archives prepared for LP release in (I think) 1980 – long after the masters were destroyed. The sound sample has a piece of the introduction and then cuts to Echagüe’s voice:
This was all we knew and we danced to it with great pleasure. But imagine if you were a member of Akihito’s Baba “Club Tango Argentino” in 2001, when this dropped through your letter box:
Lower, slower, clearer, more detailed, there are so many differences it’s just ridiculous. But when we first became aware of the second version, many people (Argentines included) preferred the processed one. It was what they were used to, and sounded more exciting. Nearly twenty years later, we know better what to listen for and the first sample is plainly ‘wrong’ – much too fast and too high (about a semitone in fact): the second one is much more natural. Listen in particular to the timbre of Echagüe’s voice. In the first sample he sounds like he’s on helium; in the second, we hear the voice of a man, a creature of flesh-and-blood. As it turns out, even this version is still a little bit too quick. Here is what we think it should really sound like:
Can you notice that it’s a bit slower? (Don’t worry if you can’t). Does it make you feel different? For me, the strong D’Arienzo beat stands out even more strongly.

Wasn’t it criminal of the sound engineers at RCA-Victor to butcher the track the way they did? Yes it was; but it turns out that raising the pitch is an idea with a long and glorious tradition going back hundreds of years, as we shall see in a moment.

Pitch correction

If I know what the correct pitch is then with the correct equipment I can simply turn the record faster or slower to produce a sound that represents the tempo and pitch of the performance. In the early days record players actually had a speed control to achieve this. Here is a photograph of the speed control of a “Victrola” made by the Victor Company in 1905. At this very early time the speed of a shellac record could be anywhere between 60 and 90 rpm, but Victor persisted with speed control into the 1920s, long after the world had largely standardised at 78rpm.

Speed control on 1905 Victrola.
Image courtesy www.victor-victrola.com

How do we know what the correct speed is? Concert pitch is A = 440Hz, right (where A refers to the A above ‘Middle C’)? So surely we just measure the frequency of the notes, work out the deviation from concert pitch, and apply a correction? Well, yes, but… are you sure that tango orchestras played at Concert Pitch? And are you sure that concert pitch is 440Hz? It may be today, but it was not always so.

A Brief History of Pitch

Until the 16th century there was no way to even measure pitch because there was no way of measuring time with sufficient accuracy (pitch, or frequency, is cycles per second, so to measure pitch one has to be able to measure a very short time interval). Even in the 18th century there was no standard pitch. Pitch varied not just from country to country, but from region to region and even from village to village. In practice, the pitch “standard” was the organ in the village church. This was hard to re-tune (you had to bash the ends of the organ pipes around with a hammer), so everyone else tuned to the organ. And how much did pitch vary? People playing baroque music on period instruments today tune to 415Hz, but this is a rough average for the period. Mozart’s piano builder (Johann Andreas Stein) worked at 421.6Hz (we have his tuning fork), but at the time pitch varied wildly:

An English pitchpipe from 1720 plays the A above middle C at 380 Hz, while the organs played by Johann Sebastian Bach in Hamburg, Leipzig and Weimar were pitched at A=480 Hz, a difference of around four semitones. In other words, the ‘A’ produced by the 1720 pitchpipe would have been at the same frequency as the ‘F’ on one of Bach’s organs.

Whilst organs could be tuned anywhere without too many consequences, orchestras were not so adaptable. As we enter the 19th century, they were tuned around 424Hz. All this would change in Vienna in 1814.

Alexander’s Gift
The Congress of Vienna was convened to re-organise Europe after the Napoleonic Wars. This was a lavish event, the like of which is hard to imagine today. All of the major powers sent their most important statesman, and Tsar Alexander of Russia attended in person. For ten months, Vienna became the centre of the world. (BBC Radio, In Our Time: The Congress of Vienna).

During the congress Alexander presented the Austrian Army with a new set of musical instruments tuned to 440Hz – a great deal higher than what they had used before, making the music sound brighter. The difference was most to be heard in the string section, whose tighter strings generated more overtones and thus sounded more resonant. Like a musical trojan horse, the new sound was a hit, setting in motion a period of pitch inflation that lasted for the rest of the century.

Pitch wars
Symphony orchestras and in particular opera houses now competed to have the “brightest” sound. Verdi wrote his operas with a pitch of A=432 Hz in mind, but the pitch at La Scala in Milan reached the dizzy heights of 451 Hz. British orchestras played at 452.5Hz (fever pitch?).

Pitch inflation was brought to an end by the tenors who complained that they could no longer perform the arias without damaging their voices. The French standardised at 435 Hz in the mid 19th century, the so-called “Diapason Normal” (diapason is french for tuning fork) – elsewhere it was called French or Continental Pitch. Other nations slowly fell into line, with two important exceptions. In North America orchestras continued to play at 440 Hz, which had previously been popular in Germany, whilst the British now abandoned their former pitch of 452.5 Hz for a new “Low Pitch” of 439 Hz. Choosing 439 instead of 440 was not as crazy as it sounds because the competing standards were specified at different temperatures – French concert halls, apparently, were colder than British ones.

A strange exception
British Brass bands, both military and civilian, continued with the former “High Pitch” because it gave a brighter sound which suited the music. Whilst the military bands converted in 1927 – leading in some cases to the purchase of new instruments – the colliery bands in the north of England persisted with the former “High Pitch” into the 1960s (!), finally changing only because the old style instruments were no longer being manufactured. But this was exceptional: by the early 20th century, most people were playing at either 435 Hz or 440Hz.

Pitch of tango orchestras

After the first world war the world slowly converted to 440Hz but tango orchestras were a special case because of the bandoneons. Like the organ which it was originally designed to replace, the bandoneon could not be tuned by the player, but only by a specialist. Bandoneons were manufactured at a pitch of 435Hz, and tango orchestras therefore played at that pitch.

The bandoneón: 71 buttons, 142 notes – not something you re-tune before a gig.

First to change to 440Hz was Fresedo, in 1934. Fresedo had already been twice to the United States, but the real influence on him now was the massively popular jazz band of Paul Whiteman which toured South America regularly. Following Whiteman, Fresedo now decided to incorporate the vibraphone into his ensemble. As Camilo Gatica points out, this was an American instrument with a fixed tuning of 440Hz. The band therefore had to tune to the vibraphone, which meant re-tuning the bandoneons. This is a drastic measure, both labour intensive and, to all intents and purposes, irreversible. The metal reeds are filed down by hand – for all 142 notes. Jens-Ingo Brodesser tells us that this change can be detected on Fresedo’s 1930s Victor recordings. As best as he can tell, the January 1934 session was recorded at 435 Hz, and the next one in April (i.e. after carnival, just as with D’Arienzo’s band) at 440 Hz.

How significant is this change in pitch? After all, it’s much less than a semitone (the interval between two notes) – the first two samples above of Pensalo bien are about a semitone apart. Instrument tuners divide a semitone into 100 cents; the change from 435 Hz to 440Hz is only 20 cents: a fifth of a semitone. No big deal, right?
Decide for yourself by listening to the third sample above, which is corrected from 440Hz to 435Hz. Most people can detect a shift of 10cents with a bit of practice and so can hear the difference, and a three minute tango recorded at 435Hz but sped up to the new standard will be two seconds shorter. And it’s this difference in pace which is the most significant. D’Arienzo in the years 1935-1938, correctly pitched at 435Hz, is not as fast as we thought it was.

1939: 440Hz becomes the standard – sort-of
In May 1939 an international conference in London recommended a tuning of 440Hz. Although it seems that at least some bandoneon manufacturers continued at 435Hz, the tango orchestras had already decided to retune – we still don’t know why, but clearly this decision was in tune (as it were) with the zeitgeist. Analysis of the D’Arienzo transfers of CTA by Frank Jin has established that the D’Arienzo orchestra recorded its last session at 435Hz on 3rd March 1939, waxing the milonga Meta fierro and the tango Dos guitas. The next session of 18th April was recorded at 440Hz. These dates would be consistent with the orchestra recording one session immediately after carnival (which took place in Feburary), with the bandoneons then being sent to the tuners whilst the band took its customary post-carnival break – the only holiday of the year for a working tango orchestra.
From this moment on the instrument importers now re-tuned the instruments from 435Hz to 440Hz when they arrived from Germany, before the instruments went on sale. The most famous and important manufacturer was Alfred Arnold (AA) and their exclusive importer in Buenos Aires was Casa America. Whilst AA would switch to 440Hz, the outbreak of the Second World War brought an end to the export of musical instruments by the end of 1941. Production ceased altogether in 1942 when the factory was converted to war production (I once read that it made diesel pumps); very few instruments were actually manufactured at 440Hz. Production did eventually re-start after the war but the Argentine market rejected the new instruments, finding their quality inadequate.

Working on bandoneón reeds – photo Hanna White, courtesy Christoph Pass

The situation was complicated by the record companies, Victor and Odeon. Musicians brought their own instruments to recordings with one important exception: the piano. All the bands had to use the piano in the studio and this could not be constantly retuned between sessions: since Fresedo must have used a different piano to the other bands recording at Victor in the years 1934-1939, he must have used a different studio as well.

But whilst most bands retuned in 1939, research by Jens-Ingo Brodesser has established that recordings made at Odeon remained at 435Hz until the end of 1943, almost four years later. Possibly this was due to their use of Bechstein grand pianos, imported from Germany in 1938 (proudly advertised on the labels of, for example, Demare’s 1938 recordings): one British pianist noted that the 1921 Bechstein Model B piano she inherited was a quarter of a tone lower than modern concert pitch. When Laurenz changed from Victor to Odeon in the middle of 1943 he had to switch back from 440Hz to 435Hz. For this purpose one suspects he used a second bandoneon.

So, from 1944 onwards all the orchestras played at 440Hz, right? Well, yes, but… Pick up a bandoneón in Argentina today and you’ll find that it is tuned to a higher pitch, somewhere between 442Hz and 445Hz. What happened?!?!

Afinación brillante: “brilliant pitch”
In the 1950s tango suffered its own episode of pitch inflation. The new pitch was referred to as afinación brillante – “brilliant pitch”. As we know, an increase in pitch makes music sound more exciting. However, there was a problem: raising the pitch of a bandoneon requires removing material from the reeds. Prior to this, the pitch stability of AA’s reed plates had been legendary: no matter how hard one blew the reeds, the pitch remained the same. However as the reeds were tuned higher and lost mass, they also lost stability: when blown really hard (i.e. played loud), the pitch would drop. (Years later, Astor Piazzolla would consciously exploit this as an expressive technique). This drop in pitch meant that the bandoneon had to be tuned slightly higher than the piano. Héctor del Curto, who played with the Pugliese orchestra, informs me that the Pugliese orchestra played with the piano at 441Hz, with the bandoneons tuned just a touch higher (between 441 Hz and 442 Hz). However, other orchestras went higher. Carlos Lázzari told his nephew Facundo Lázzari that the D’Arienzo orchestra tuned at 442Hz, and I’ve read that some orchestras went even higher. Whilst 435Hz and 440Hz were “standards” for the típicas, afinación brillante is not actually a standard.
When did this change take place exactly? We don’t know. Elvino Vardaro’s nephew Osvaldo tells us the Di Sarli orchestra retuned in 1956, the year in which he joined the orchestra.

Concert pitch today – standardised?
Concert pitch is still not absolutely standardised today. Yes, most orchestras play at 440Hz, but there are exceptions, and these include many of the most famous orchestras. The New York Philharmonic plays at 442Hz, and the Berlin and Vienna Philharmonics at 443Hz.

Many people in the world of opera believe that today’s Concert Pitch of 440Hz is still too high, changing the “colour” of the voices. In 1988-1989, Placido Domingo, Luciano Pavarotti, Birgit Nillson and Renata Tebaldi petitioned the Italian government to return to the pitch of Verdi’s time, 432 Hz. This movement is viewed suspiciously because it has some dubious supporters, but the arguments are compelling. I cannot believe, said Tebaldi, that Italy, which gave to the world great voices, can no longer produce [them]… If we went back to the correct tuning, I am sure that we could return the Italian opera to its Golden Age.

Second Postscript: Pitch Inflation in Modern Pop Music
In the 1970s record producers starting using the Varispeed function on their tape machines to speed up the performance of a pop song by 1 or 2%. This was typically done for the “radio single” version of a song.
In 1989, A&M sped up Janet Jackson’s Rhythm nation by 40 cents (A = 450Hz) with unexpected consequences. The frequency of the shifted note “E” in the bass line now coincided with the resonant frequency of a popular model of laptop hard drive, crashing laptops which played the song.

Many people helped with the information in this article. In addition to Jens-Ingo Brodesser, Frank Jin, and Camilo Gatica, I’d like to thank some people whom we dancers generally overlook, because we don’t know about them – the bandoneon restorers and tuners:

My apologies if I’ve forgotten anyone – just let me know.


  • By how much were 78s sped up when they were transferred to LPs in the 1960s?
    In my experience it was commonly around 2/3 of a semitone, but the topic is still being investigated.
  • You say that “various problems” could prevent the master record being produced at 78rpm. Such as?
    Instability in the mains frequency was one problem; another was the resistance given to the cutting head by the wax, which increases as one gets closer to the centre of the disc. Tanturi’s 1937 recording of A la luz del candil slows down by half a semitone from the beginning to the end of the disc. Don’t believe me?
    It’s crazy, isn’t it?! No-one noticed, and they printed it. So much for the quality control at Odeon.
  • You said that 77.92rpm was locked to the 50Hz mains electricity. What about in the US, which uses 60Hz?
    They used a different motor and gear ratio to produce 78.26 rpm, a negligible difference (0.4% – less than 1 cent).
  • What’s all this stuff on Youtube about 432Hz being a cosmic frequency?
    It’s nonsense. 432Hz is nicely divisible by 2, 3, 4, 6, 8 and 9, but so what? If you want a cosmic frequency, then it should be the one with Middle C = 256, this being an integer power of 2. This corresponds to A=430.54Hz. This is not a beautiful number, but even so I predict that it will become fashionable eventually.