Icons of Sound
Aesthetics and Acoustics of Hagia Sophia, Istanbul

A collaboration between Stanford University's Center for Computer Research in Music and Acoustics and Department of Art & Art History

What is an Auralization?

Imagine that you are floating in the air, far away from any sort of surface that could reflect sound. Now imagine that a bird is flying through the air some distance away from you, and emits a call. Sound from the bird will reach your ears after a short delay.

Now imagine that you are in an enclosed room, with the same bird. When the bird calls out this time, sound will reach your ears both directly from the mouth of the bird and also from reflections off of the floor, walls, and ceiling of the room. The echoes you hear will change depending on where you are in the room, where the bird is, and the layout of the room itself.

Finally, imagine that you are blindfolded and have no visual cues about the space that you are in. If we can accurately produce sounds such that the sound reaching your ears is the same as the sound that would have reached your ears in either of the above scenarios, then sonically speaking there is no difference between listening to this production and actually being in the original space.

This is the basis of aurilzation in our context. To synthesize sounds so that, upon listening to them, you are virtually transported to a different place.

© Michael J. Wilson

Auralization Procedure

As part of our Icons of Sound effort, we produced several auralizations, virtually synthesizing performances of the Herubikon, Kontakion, Prokeimenon, Trisagion, and Psalm 140 in Hagia Sophia and Memorial Church. We used two approaches to achieve our auralizations. In the first, we started with a recorded performance of the Herubikon by the Byzantine Greek Choir of Lycourgos Angelopoulos in Fontevraud Abbey, and re-imagined it as if it were performed in Hagia Sophia. In the second approach we made two recordings: one of Cappella Romana and one of Stanford Student Konstantine Buhler, capturing the singers' voices without acoustics. The "dry" voices were then processed to place them in Hagia Sophia and Memorial Church.

Both approaches process recordings according to the acoustics of Hagia Sophia and Memorial Church derived from measurements of balloon pops performed in the spaces. In particular, balloon pop recordings are manipulated to derive impulse responses of the space measured between the balloon location and listener locations. When a balloon explodes, a short-duration pulse is emitted in all directions. As this pulse propagates through the space, it interacts with the architecture and objects in the space, carrying with it to the listener an imprint of the space. In our processing, we simply decomposed our recorded waveforms into a sequence of pulses and processed each pulse according to the desired impulse response. In this way, the recorded performance is virtually placed in Hagia Sophia or Memorial Church.

To produce our dry recordings of the Trisagion, Kontakion, Prokeimenon, and Psalm 140, we outfitted the performers of Cappella Romana with Countryman E6 headset directional microphones, which effectively record only the singer's voice---very little of the reflected energy of the room is captured. The difficulty is that the singers rely on the acoustics of the space they are singing in to determine their meter and phrasing and even, in subtle ways, their pitch. We provided the singers with ear-buds and gave them real-time feedback as if they were chanting in Hagia Sophia. While the singers had the impression that they were performing in Hagia Sophia and adjusted their tempo and phrasing accordingly, we were able to record only their "dry" unprocessed voices.

To create an auralization of the Cheroubikon as if it were performed in the Hagia Sophia, we transformed a recorded performance of the hymn made in the 46-meter long refectory of the Royal Abbey of Fontevraud, France. The recorded performance was analyzed to determine the frequency-dependent reverberation time of the space. We then synthesized a late-field "correction" impulse response having an evolving decay rate, fixed according to the Hagia Sophia late-field decay rates recorded by Bissera Pentcheva and compared with the data reported in.

Convolving the original recording with the late-field correction produces an output, which approximates---in a psychoacoustic sense--- the late-field response of Hagia Sophia. The abbey refectory is a large space with stone walls and a vaulted stone ceiling. Therefore, the early part of its response is expected to share features with the Hagia Sophia early response. Accordingly, the synthesized Hagia Sophia performance was created by adding the original recording to a delayed convolution of the original recording with the correction. In this way, the synthesized performance has the early response of the original recording and a late-field mimicking that of the Hagia Sophia. Owing to the very long reverberation time of the Hagia Sophia, the late-field reverberation is expected to dominate the psychoacoustic impression of the space, and the synthesized performance should convey an accurate sense of the Cherubic Hymn performed in Hagia Sophia.

© Jonathan Abel

Recording to Create Auralization Source Material

A 13-singer ensemble from the group Cappella Romana visited CCRMA in March 2011 to make a recording specifically for use in our Hagia Sophia auralization model.

Although the ensemble was recorded in a "live" context -- with everyone together in a small concert hall, singing simultaneously -- for this recording we used a process and technology that enable us to 1) capture the sound of each singer individually, into separate recorded files; 2) without the acoustic characteristics of the recording space; and 3) performing with the acoustic influence of Hagia Sophia.

The result is a multi-channel recording of musical selections appropriate to Hagia Sophia of Byzantium, in which each singer's voice is recorded in a separate sound file whose timing and articulation is synchronized with the entire ensemble. These individual voices are recorded "dry", that is, the sound is not imprinted with the acoustic characteristics of any space. Thus, the voices of any combination of singers can be used together to reproduce the music, and imprinted with the effects of any simulated acoustic context.

Because this recording was made with the singers hearing their performance through the reverberation of Hagia Sophia -- via our measurement-based simulations -- the vocal timing, articulation, and intonation of the musical performance is customized for use in auralizations of Hagia Sophia. The transparent acoustics of the recorded tracks allows them to be used as neutral sound sources placed in any physical area of an acoustic model.

© Miriam Kolar

I. Kontakion, Romanos Melodos, On the Nativity I

St. Petersburg, MS Gr. 674, 13th cent., transcribed by Alexander Lingas.

Bibliography: Paul Mass and C. A. Trypanis, Sancti Romani Melodi Cantica. Cantica Genuina (Oxford: Clarendon Press, 1963).

R. J. Schork, Sacred Song from the Byzantine Pulpit. Romanos the Melodist (Miami: University Press of Florida, 1995).

Oliver Strunk, “Some Observations on the Muisc of the Kontakion,” in Essays on Byzantine Music (New York: Norton, 1977), pp. 157–64.

The thirteen chorists of Cappella Romana performed this piece all together (as psaltai and congregation), while the five drone singers held the ison. They chanted only the prologue and refrain. Possibly in its original sixth-century enactment, this kontakion would have been chanted by a soloist from the ambo, while the congregation would have joined in singing the refrain. It would have been performed on Christmas Day at orthros (matins). The melody comes from a thirteenth-century manuscript. Musical notation for the kontakia first appears in the tenth century. Yet, many original settings were orally transmitted. Given the prominence of this kontakion in the Byzantine liturgy for Christmas, it is possible that we have here a musical setting that hearkens back to the sixth-century prototype.

Romanos Melodos (born late fifth century, died after 555, feast day October 1) introduced the kontakion in the Constantinopolitan liturgy. It is a sermon in meter and set to music comprising 18-24 stanzas with a prologue (koukoulion) and a refrain introduced at the end of the prologue and repeated at the conclusion of each stanza. The name of the author appears in an acrostic composed of the first letter of each stanza. According to the legend narrated in his vita, Romanos Melodos received inspiration for writing kontakia directly from the Virgin Mary. On the eve of Christmas Romanos saw the Mother of God in his dream. She gave him a scroll to eat. On waking up, he ran to the ambo of the Church of the Virgin ta Kyrou from where he started chanting his sermon on the Nativity I. Romanos Melodos’ prolific oeuvre promoted the kontakion as a staple of the Constantinopolitan liturgy. Like the characteristic Sung Office (asmatike akolouthia) of the cathedral liturgy in the Byzantine capital, the kontakion also exhibits a predilection for chanted rather than read text and inclusion of the congregation through a sung refrain. The appearance of the musical sermon coincides with the construction of the Justinianic Hagia Sophia. Thus the kontakion from its inception interacted with the reverberant acoustics of Constantinople’s domed cathedral.

Ἡ παρθένος σήμερον    τὸν ὑπερούσιον τίκτει, 
    καὶ ἡ γῆ τὸ σπήλαιον    τῷ ἀπροσίτῳ προσάγει·
  ἄγγελοι μετὰ ποιμένων    δοξολογοῦσι,
    μάγοι δὲ μετὰ ἀστέρος    ὁδοιποροῦσι·
  δι’ ἡμᾶς γὰρ    ἐγεννήθη 
    παιδίον νέον,    ὁ πρὸ αἰώνων Θεός.

Greek from Paul Mass and C. A. Trypanis, Sancti Romani Melodi Cantica. Cantica Genuina (Oxford: Clarendon Press, 1963), p. 1.

Today the Virgin gives birth to the supersubstantial one.
Earth offers a cave to the unapproachable one
Angels and shepherds join in a hymn of glory,
as the Magi are guided on their trek by a star.
On our behalf there has been boor,
an infant now, yet God before all ages.

English tr. from R. J. Schork, Sacred Song from the Byzantine Pulpit. Romanos the Melodist (Miami: University Press of Florida, 1995), pp. 49–59, esp. pp. 50–51.

II. Psalm 140 (Vulgate 141): 1–7


Transcription and edition of the musical setting Ioannis Arvanitis


Alexander Lingas, “Festal Cathedral Vespers in Late Byzantium,” Orientalia Christiana Periodica 63 (1997): 421-59

The first half of Psalm 140 from the cathedral vespers (hesperinos) tradition includes a solo intonation (domestikos: John Michael Boyer), choral verses sung by the psaltai Alexander Lingas, Mark Powell, and John Michael Boyer. The drone (ison) chanted by Daniel Burnett, Daniel Buchanan, Aaron Cain, Patrick McDonough, and David Krueger. The entire Cappella Romana ensemble sings the refrain intended for the congregation (all the above plus Stephanie Kramer, Catherine J. van der Salm, LeaAnne DenBeste, Wendy Steele, and Kerry McCarthy).

The Late Antique character of the Sung office of Hagia Sophia would require the soloist known as domestikos to ascend the ambo and sing from there the opening and the refrain. The professional choir, psaltai, standing under the ambo then begins to chant the psalm antiphonally, meaning they divide into two groups and sing in alternating fashion verse by verse. When one group chants the poetic line, the other holds the drone (ison) and vice versa. At the end of each verse the congregation joins in with the refrain.

Psalm 140 was likely performed after the congregation led by the clergy has entered together the nave of Hagia Sophia. The second verse makes the connection between prayer and evening sacrifice of incense; the lifted hands rise like the smoke of thymiama: “Let my prayer be set forth before thee as incense; the lifting up of my hands [as] an evening sacrifice.” It is this line that recommended Psalm 140 for the vespers service. Its poetry resonates both with the evening service and the ceremony of censing the congregation.

The Byzantine vespers intercalates the verses of Psalm 140 with refrain “We glorify your saving resurrection, Lover of humankind,” thereby conveying the idea how human salvation has become possible through Christ’s sacrifice at the Crucifixion. The prayer addresses the faithful to God. It is a request to guard him/her from the iniquities of the ones who do not follow the right way, to preserve the soul of the believer from death and dispersal.

The way the refrain completes each verse conveys circularity. Moreover, in Greek the word choros denotes both a circle and a choir. In Byzantine theology choros also designates the circular movement that elicits divine response and leads to renewal. Thus for a Greek mind the word “choir” simultaneously emerges as a visual and acoustic manifestation of enclosing, reverberant shape.

The interlacing of text, incense, and chant enables the sensual experience of simultaneous sound and smell. In addition, the spacious domed and marble interior of Hagia Sophia slows down the tempo, endowing the chant with great solemnity. The air becomes dense, filled with reverberant sound and fragrant incense. It is at this point that the priest announces: “Sophia, stand upright!” Sophia – wisdom – becomes manifest in the multisensory experience of acoustic and olfactory choros.

καὶ εὐθὺς ὁ Δομέστικος τὸ κεκραγάριον
εἰσάκουσόν μου. Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.

Οἱ χοροὶ ἐναλλάξ
1 κύριε ἐκέκραξα πρὸς σέ εἰσάκουσόν μου πρόσχες τῇ φωνῇ τῆς δεήσεώς μου ἐν τῷ κεκραγέναι με πρὸς σέ. 
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
2 κατευθυνθήτω ἡ προσευχή μου ὡς θυμίαμα ἐνώπιόν σου ἔπαρσις τῶν χειρῶν μου θυσία ἑσπερινή. 
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
3 θοῦ κύριε φυλακὴν τῷ στόματί μου καὶ θύραν περιοχῆς περὶ τὰ χείλη μου. 
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
4 μὴ ἐκκλίνῃς τὴν καρδίαν μου εἰς λόγους πονηρίας τοῦ προφασίζεσθαι προφάσεις ἐν ἁμαρτίαις.
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
σὺν ἀνθρώποις ἐργαζομένοις ἀνομίαν καὶ οὐ μὴ συνδυάσω μετὰ τῶν ἐκλεκτῶν αὐτῶν.
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
5 παιδεύσει με δίκαιος ἐν ἐλέει καὶ ἐλέγξει με ἔλαιον δὲ ἁμαρτωλοῦ μὴ λιπανάτω τὴν κεφαλήν μου.
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε. 
ὅτι ἔτι καὶ ἡ προσευχή μου ἐν ταῖς εὐδοκίαις αὐτῶν. 
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
6 κατεπόθησαν ἐχόμενα πέτρας οἱ κριταὶ αὐτῶν.
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.
7 ἀκούσονται τὰ ῥήματά μου ὅτι ἡδύνθησαν. ὡσεὶ πάχος γῆς διερράγη ἐπὶ τῆς γῆς διεσκορπίσθη τὰ ὀστᾶ ἡμῶν παρὰ τὸν ᾅδην. 
Τὴν σωτήριόν σου ἔγερσιν δοξάζομεν, φιλάνθρωπε.

And immediately the Choir Leader the refrain of Psalm 140
Hear me. We glorify your saving resurrection, Lover of humankind.

The Choirs alternately.
1 O Lord, I have cried to thee; hear me: attend to the voice of my supplication, when I cry to thee.
We glorify your saving resurrection, Lover of humankind.
2 Let my prayer be set forth before thee as incense; the lifting up of my hands [as] an evening sacrifice.
We glorify your saving resurrection, Lover of humankind.
3 Set a watch, O Lord, on my mouth, and a strong door about by lips.
We glorify your saving resurrection, Lover of humankind.
4 Incline not my heart to evil things, to employ pretexts for sins.
We glorify your saving resurrection, Lover of humankind.
With those who work iniquity: let me not unite with their elect.
We glorify your saving resurrection, Lover of humankind.
5 The righteous shall chasten me with mercy, and reprove me: but let not the oil of the sinner anoint my head.
We glorify your saving resurrection, Lover of humankind.
For yet my prayer shall be in their pleasures.
6  Their judges have been swallowed up near the rock.
We glorify your saving resurrection, Lover of humankind.
They will hear my words for they are sweet.
7 As a lump of earth is crushed upon the ground, our bones have been scattered by the [mouth of] the Hell.
We glorify your saving resurrection, Lover of humankind.

Greek from The Septuagint with the Apocrypha: Greek and English, ed. Sir L. Brenton (Peabody, Mass., 1986) with the additions of refrain and ceremony rubrics by Alexander Lingas, Vespers. According to the Rite of the Great Church of Hagia Sophia, Constantinople for the Feast of the Vigil of the Feast of the Fathers of the 1st Ecumenical Council of Nicaea. The Rt. Rev. Dr. Kallistos Ware Presiding and the Greek Byzantine Choir of Lycourgos Angelopoulos performed in the Chapel of St. Peter’s College, Oxford, Saturday May 26, 2001.

III. Prokeimenon for the Feast of St. Basil celebrated on January 1


Transcription and edition Ioannis Arvanitis

Bibliography MISSING

Psalm 48 (49): 3 
Ὁ στόμα μου λαλήσει σοφίαν 
Psalm 48 (49): 1 
ἀκούσατε ταῦτα πάντα τὰ ἔθνη 
Psalm 48 (49): 2 
οἵ τε γηγενεῖς καὶ οἱ υἱοὶ τῶν ἀνθρώπων

Ps. 48: 3
My mouth shall speak Sophia (wisdom)
Ps. 48:1
Hear these words, all ye nations
Ps. 48:2
And those born sons of men

The prokeimenon bears resemblance to the Latin gradual. It was chanted before the office of the saint. The domestikos most likely sang it from the top of the ambo. Here John Michael Boyer chants the solo parts and is joined by the other two psaltai (Alexander Lingas and Mark Powell), while the five drone singers hold the ison. The women did not perform.

The opening lines dramatically introduce the descent of wisdom/Sophia: “my mouth will speak Sophia,” Ps. 48:3. Then the soloist addresses the faithful: “Hear these words, all ye nations.” With this carefully selected excerpts from Ps 48 (49) the space and congregation of Hagia Sophia are both engaged. As the interior fills with the chanting breath speaking ‘sophia,’ reverberating against the marble floor and walls and golden dome, all the nations gathered inside slowly consume this very pneuma.

Characteristic of the Byzantine liturgical performance is the attempt to make present what is absent: empty space that becomes dense with people, smoke of incense, and most aesthetically – with the air pressure of sound waves. The chanted office of the Constantinopolitan Hagia Sophia makes sentient the spherical empty space of its naos, its kallichoros (the beautiful space/chora), filling it up with a choros.

© Bissera V. Pentcheva