I received various notes about a brief clip on NPR about the making of a recording of Byzantine chant in a virtual, digital, recreation of the interior Hagia Sophia. Somehow they were able to recreate a virtual interior of Hagia Sophia using the sound of a bursting balloon. Then they filled that virtual interior with the chant, thus re-producing the acoustic effect for the listeners within the enormous building.
As Clarke quipped, any sufficiently advanced technology is indistinguishable from magic.
I have the album. Unreal.
Lost Voices of Hagia Sophia
The music group has game: Cappella Romana. WOW.
The liturgical chant they record is for the Exaltation of the Cross, 14 September, introduced in 628.
How to understand what the liturgical experience would have been within Hagia Sophia? Today, chanting is forbidden within. It had to be digitally recreated.
For those of you in Columbia Heights, Hagia Sophia – Wisdom of God – was/is the mighty Byzantine cathedral, then a mosque, then a museum, built in Constantinople at the orders of the Emperor Justinian. It is quite simply enormous. The dome is higher than all the great, soaring Gothic churches, 56.6m from the floor, and 31.87 in diameter. The church has 255,800 cubic meters of space, covered in marble. The reverb is 12 seconds. The space was designed to create a waterfall of resonant sound, mirroring the book-matching, opened marble sheets on the walls which resemble waves of water. Just as layers of colors are used in icons to convey deeper realities, so in sound an icon is created. The sound and building harmonics go to where the human ear can barely reach, to omphe, Greek word meaning a voice beyond human register, thus a divine voice, which rises like axes to the dome to form a cascade back down to the floor, reflecting, like the bright gold mosaics reflect light.
Not too shabby. Take in that balloon pop!
Loading ...
I had a priest tell me once that there were ominous signs at Hagia Sophia on the eve of its fall to the Muslims, similar to those Josephus records as presaging the destruction of the Temple in Jerusalem in 70 A.D.
Pingback: LENTCAzT 2020: 07 – Tuesday in the 1st Week of Lent: Every sort of trouble proceeds from the perverse root of self-love | Fr. Z's Blog
Well, the method they used is one way to do this. They used a lapel mic to record the impulse response at various locations in the Hagia Sophia, apparently, as a balloon (a source of white noise) is popped. The recordings were used to make a finite impulse response map, which is then used to make a digital filter. The Hagia Sophia filter is then convolved or convoluted with any other sound sample. Mathematical convolution is the process of modifying one function (here, the choir input) by the other function (the Hagia Sophia digital filter). The result modifies the original sound as if it had interacted with the filtered room acoustics.
This method is not something that could only have been done in the last ten years, as NPR claims. That is nonsense. Digital optical and acoustic processing has used convolution methods for years – the method, itself, goes back to the late 1700’s. Any modern software package like Matlab or Octave can do this sort of convolution. I could have done this 25 years ago on a Silicon Graphics workstation.
This is a nice, practical achievement, however. It is a first-order approximation to what the chant would have sounded like. It is probably the best empirical data we will be able to get. I would have approached the problem a bit differently, because a popping balloon and a lapel mic doesn’t, I don’t think, capture the lower resonance frequencies. The frequency responses of popped balloons has, recently been measured and the frequency response depends on the size of the balloon, with weighting of certain frequencies, in some cases, so it is not a perfect impulse generator. It is cheap and portable, so it is often used. An article from the Acoustical Society of America measured the response of 160 balloons:
https://asa.scitation.org/doi/full/10.1121/1.3518780
It should be possible to build a model of the Hagia Sophia on a computer and use ray tracing or other techniques to calculate an impulse response. This would make a nice Master’s or Doctoral project in architectural acoustics.
The Chicken
[Not lapel mic, but mics at the ears.]
Hagia Sophia is one of my favorite churches in the world, right up there with Sagrada Familia and, closer to home, St. Mary’s Basilica in Phoenix.
Don’t forget that from 1204 to 1261, Hagia Sophia was a Latin Church cathedral, much to the chagrin of ecumenists such as myself. 1204 was a very dark year (numerically very close to my heart as well) when the Latins sacked Constantinople and made off with treasures of the East. There was a lot of bad blood since the schism of 1054 when Humbert slapped down bulls of excommunication on that precious altar.
I mourn the state of the Eastern Orthodox communion today, particularly two things: Patriarch Bartholomew’s unwelcomeness in Istanbul, and the Moscow schism. Where is Pope Benedict when we need him now, the Pope of Christian Unity. I suppose it is inevitable that “Third Rome” would make a power play, but they just drift further and further apart…
God, the earth, the heavens sound the tocsin;
Santa Sophia, the Great Church, also sounds the alarm,
With its four hundred tocsins and the peel of sixty-two bells;
For every bell there is a priest, for every priest a deacon.
The emperor chants to the left, the patriarch to the right,
And all this chanting makes the columns shake.
As they were starting the Cherubic hymn for the emperor to exit
A heavenly voice was heard from the mouth of the Archangel:
“Stop the Cherubic hymn, lower the sacred implements;
Priests remove the holy (vessels); candles blow yourselves out.
Because it is the will of God that the City fall to the Turks.
Only send word to the West and ask for three ships to come;
One will take the cross, the other the Gospel,
The third, the best of the three, our holy altar,
Lest the dogs take it from us and defile it.”
The Madonna was perturbed and the icons shed tears.
“Be still Madonna and weep no more:
With the passage of years and time, it will be yours again.”
Simeni o Thios, simeni i yi, simenun ta epurania,
simeni k’ i Ayia Sofia, to megha monastiri,
me tetrakosia simandra k’ exindadhio kambanes,
kathe kambana ke papas, kathe papas ke dhiakos.
Psalli zerva o vasilias, dexia o patriarhis,
ki ap’ tin polli tin psalmudhia esiondane i kolones.
Na mbune sto heruviko ke na ‘vyi o vasilias,
foni tus irthe ex uranu ki ap arhangelu stoma:
“Papsete to heruviko ki’ as hamilosun t’ ayia,
papadhes parte ta yiera, ke sis keria svistite,
yiati ine thelima Theu i Poli na turkepsi.
Mon stilte logho sti Frangia, na ‘rthune tria karavia,
to ‘na na pari to stavro ke tallo to vangelio,
to trito to kallitero, tin ayia trapeza mas,
mi mas tin parun ta skilia me mas ti magharisun.”
I Dhespina tarahtike me dhakrusan i ikones.
“Sopase kira Dhespina ke mi polidhakrizis,
pali me hronia me kerus, pali dhika sas ine.”
I’m intrigued that “fall to the Turks” can be expressed by one word in Greek: turkepsi, as if this was a phenomenon wearily familiar to the last Byzantines. (The last Romans they would have said.)
Christ is eternal; the lie of Islam fleeting. One day the Hagia Sophia will return. Even the Muslims know this; the beauty of the Church is hidden by curtains and fake walls…not chiseled and disfigured.
The music is hauntingly beautiful …
Probably, no one besides me is interested in the technical details of this project, but there are a number of good papers available showing how they did it. Here are slides from a talk by the Standord CCRMA group:
https://ccrma.stanford.edu/~njb/research/AES129_Balloon_Slides.pdf
and here is the technical paper:
https://pdfs.semanticscholar.org/808c/7fc70f22b524b1e1aaa2bed61169be3932d1.pdf
According to their research, the balloon pop compares favorably to sine sweep methods (which is what I would have used), although there is some question about its ability to accurately reproduce the iOS 3382 standard at high frequencies, but for the most part, it is a nice cheap way to get room acoustics data. This could be done for any church (would be nice to do with some older European churches to make up an acoustic database).
The Chicken
The Masked Chicken :
Probably, no one besides me is interested in the technical details of this project
A friend of mine, who recently passed away, studied advanced mathematics, assisting the French Professor of Mathematics who pioneered all of these things many decades ago.
And in his talks to me on the subject, because he did like to keep up to date as he could, he seemed to suggest to me that the focus in more recent times has not been on the fundamentals of acoustic soundscapes, but on improving the capability of microphones to both accurately analyse them by eliminating as much % possible of irrelevant noise.
Continuous ongoing improvement of signal-to-noise ratio, but focused not so much despite my previous points on microphone tech improvements, but rather upon mathematical and material analysis of human hearing itself in signal-to-noise scenarios.
From that, albeit that this is all very far from my subjects, I would guess that much that was technically feasible 25 years ago may have become technically easier than it was.
@Grant M
Who wrote that poem? I thought it might have been Cavafy, but I can’t find it on the internet.
Eastern liturgical music is thrillingly beautiful–and that balloon pop! Amazing.
@Charlotte Allen
I first came across the poem in the Penguin Book of Greek Poetry, which ranges from Homer to Kazantzakis. This poem was included with various anonymous pieces from the time of Turkish dominion over Greece.
I rediscovered the poem here by googling last mass in hagia sophia poem.
https://www.google.com/url?sa=t&source=web&rct=j&url=http://grbs.library.duke.edu/article/viewFile/14639/3803&ved=2ahUKEwj_spT24v7nAhUQWX0KHbBPC1wQFjAWegQICBAB&usg=AOvVaw3dPP93iMmheRKiKwuKfak0
Clicking in the link downloads a pdf file of the song in English and Greek, together with an article by Marios Philippides entitled Tears of the Great Church: The Lamentation of Santa Sophia. It begins:
During the period of the Ottoman occupation, the so-called Tourkokratia, the Greeks expressed their concerns in folk songs, whose numerous variants were gradually collected and published in the nineteenth century to form an impressive corpus. Some songs reach back all the way to the last years of Byzantine Greece before its fall to the Ottoman Turks. One song in particular achieved a great deal of popularity and perhaps qualifies as the most popular demotic song among Greek-speakers of the nineteenth century and the first decades of the twentieth. The poem is well known, but it has not received the scholarly attention it deserves. Entitled [The Song] of Santa Sophia, it is thought to describe the situation shortly before the fall of Constantinople to Sultan Mehmed II Fatih on May 29, 1453. This song survived orally and was finally recorded in the nineteenth century.
Tremendous music. Cappella Romana have several pieces on YouTube. Maybe the Columns really were literally shaking with the chant in the old days.
Or you can go to grbs.library.duke.edu (grbs = Greek, Roman and Byzantine Studies) and put “tears of the great church” into the search box.
The poem is from the Byzantine folk tradition, called The Lamentation of Santa Sophia. There is a scholarly essay here, with the text included as an appendix. It’s a .pdf download from Google Books, I believe.
Cappella Romana is truly one of the Northwest’s jewels; they don’t perform an extensive season but each of their concerts is worth every bit of trouble getting to and from Portland.
Tsk; I must not have refreshed the page before typing– I don’t mean to repeat what’s already been noted, honest.
@Grant M (and Marc from Eugene):
Thanks for the information!
As one who has no knowledge of such things, I have a technical question. Assuming there would have been many people worshipping in the Hagia Sophia when the chants were sung, would they have sounded different from these recordings, since the filter was created using the sound of a balloon popping when the space was virtually empty? Can the filter be adapted to account for the difference? Or is the difference simply insignificant?