Introduction
This blog post is translated from our French podcast called Conversations. It documents the work of researchers and experts involved in preserving and promoting the recordings of the first Congress of Arabic Music held in Cairo in 1932.
Today, we meet Luc Verrier, sound engineer at the Bibliothèque nationale de France (BnF). He looks back on his 30-year career in sound engineering. He also explores his approach to digitising and processing the 334 pieces recorded in Cairo and saved from oblivion.
On 24 January 2024, he welcomes me into a cosy studio at the French National Library. The light is subdued, and we are surrounded by turntables. The conversation can start…
Interview
LV: I have been working at the National Library of France since 1994. I will soon be celebrating 30 years of working at this noble institution. I have training in this field, as I have always been passionate about sound. I did initial training to become a sound engineer, thinking I would be able to work in film mixing. That was the area I was most interested in, but I was able to study the different fields related to sound engineering.
Then I did my military service in the army cinema, where I was able to work with real professionals and became more interested in sound recording for reports and mixing interviews. After completing my military service, I was able to work for a company called Elison, which had been founded by a great sound engineer named Jean-Pierre Ruh, who has since passed away, and which did all kinds of work related to sound post-production, particularly for feature films.
My job was to transfer the rushes that arrived in the evening and had to be sent out again in the morning. I regularly transferred the Nagra tapes to perforated films, perforated magnetic tapes, so that they could be resynchronised with the image in the laboratory and the director could listen to his rushes the next day. I did that for two years, and that job allowed me to develop my ear and to use tools that enabled me to restore sound, or at least to filter the sound that came from direct sound recordings. So that was really the sound engineer [Jean-Pierre Ruh], who was behind the initiative for the recording of the “son direct “[1] in France.
More than 30 years ago, when we were filming, the cameras made noise, the lighting and the traffic made noise, so we tried to filter as much as possible to get the purest, most beautiful dialogue possible. I worked quite early in the restoration and filtering of these sounds. It happened that this company had developed sound restoration equipment, which was sold in mixing auditoriums, distributed by Nagra France (called Isma-12 or Isma-18). And I began to train my ear to detect problems in the signal that we wanted to reduce or eliminate, while simultaneously filtering to improve sound that wasn’t always very good due to a variety of parameters.
I was able to really start training my ear on that, and in the same company, we had to work with a colleague on purchasing digital equipment this time – because at that point, I wasn’t really working with digital yet – to buy digital equipment for sound restoration with tools called Sonic Solutions. It was an American company that developed the NoNoise software, which enabled very sophisticated sound restoration. That’s really what we wanted the company to buy so that we could set up a sound restoration unit.
As it turned out, the company didn’t buy it, but I met with the distributors at the time who were going to sell this equipment to the National Library. So, I sent my CV to the Library, which needed to recruit people to use this equipment, so they recruited me to set up that unit.
I was hired to assemble a whole bunch of devices, with other colleagues, to digitise an initial collection of sounds from the Phonothèque nationale that could be played at the opening of the Grande bibliothèque, 20 years ago. So we had to digitise a whole bunch of things so that we could play them to readers and the general public when they came. And that was very interesting because it allowed me to work on what were the most prestigious documents in the Phonothèque or the Audiovisual department at the time and at the same time to handle all types of sound media to digitise them.
The acquisition of digitisation equipment didn’t arrive straight away because of administrative red tape and other issues. So, little by little, I was able to familiarise myself with the collections, which at that point hadn’t yet been moved here because the buildings [of the National Library] didn’t exist here in Tolbiac. I was at the Richelieu site, on Rue de Louvois.
There, I worked with these historical collections and learned what a disc was, a 78-rpm disc, a cylinder, so I was really immersed myself in these old collections and benefited from the knowledge of all my colleagues, who knew these collections very well. Then, when we were able to purchase the equipment to digitise and restore the collections, I worked on restoring this corpus for the opening of the [new] Library.
So, I worked on cylinders as well as 78 rpm discs, magnetic tapes and magnetic wires; we worked on all types of media. What was quite interesting was that we highlighted the most prestigious documents, so we were working on really rare and interesting documents. I trained my ear on these and gradually became a bit more of an expert in everything to do with old media from our collections and thus acquired the expertise to read the discs correctly (i.e. the cartridge, the stylus, the turntable, the speed, the pre-fill).
Then in the digital field, the converter, the editing tools, the restoration tools, we naturally followed market developments, and it really became my core business, which led me to move from sound technician to technical manager for everything related to digitisation, because I trained other colleagues, who still work here, to be able to digitise our collections in the best possible way.
So that was my core business, and what I find quite interesting is that, since I’ve been here for over 30 years, I’ve been involved in all the preservation plan campaigns or the digitisation plans of the Audiovisual department (which is now called the Sound, video and multimedia department).
I was able to follow all these campaigns, and now I’m a bit of a survivor because all my other colleagues have retired. I remember that this or that collection was digitised in such and such a way, or that we took such and such a step, or that there was such and such a problem. This memory, which is a bit of an oral memory, although we have documents, reports, but in any case, I’m the one who still holds the secret.
Regarding the Cairo Congress recordings, as I was saying, I arrived in 1994 and, fairly quickly, the question of digitising this corpus arose and we started to work on it. The National Library had already done a reissue in 1988 with several recordings that had been digitised, restored and put into a box set, but the desire to do a complete edition was particularly present in the mind of Pascal Cordereix, who from very early on had wanted to do a complete edition, because we knew that everyone, all the researchers, were interested in this document, but it had to be exhaustive.
FG: Can you remind us of who Pascal Cordereix is?
LV: Pascal Cordereix is now retired, but he was for a certain period the head of the Sound section of the Audiovisual department (now the Sound, video and multimedia department). I certainly worked and collaborated with him a great deal, and it was a real pleasure to discuss with him the Archives de la Parole collections, the Cairo recordings and other collections as well.
In 1996 (I did some research in my personal archives) and that year, we decided to carry out an initial campaign to transfer the entire disc collection from the Cairo Congress. I don’t know why, but this first transfer campaign was not carried out by us, probably because we didn’t have the equipment yet, or perhaps we thought that by working with an external service provider, we would be able to achieve a better transfer.
In the end, we worked with a real expert named Lionel Risler, the director of Sofreson. It was at his place that we did this first digitisation transfer using what I believe was the best digitisation protocol available at the time. So we used the best cartridges, the best styluses, the best turntable settings, and the best converters we had available to achieve the best possible digitisation. The converters we used at the time were British, Prism Sound converters.
The styluses were Expert stylus, which are also made in England. Well, then, the cartridge had to be a Shure cartridge, but I think we also used UMT cartridges. And then there was a whole bunch of equipment that we used to try to make the best possible transfer. And it also happened that Lionel Riesler, in his company, had sound restoration tools. So we had him do the sound restoration work.
There was an initial sound restoration pass on the discs that had been digitised, which was done by Lionel Riesler’s company. But very early on, we wanted to have the most ethical restoration possible, which is our policy at the Library anyway. This means touching the documents as little as possible, altering them as little as possible, and in any case, preserving the useful signal so that as much information remains, even if we do restoration work.
So, we followed the protocol, which was simple: we did what we call a ‘straight’ digitisation (i.e. without any correction, clicking, filtering or anything else). So that’s really our reference master. From that straight version, we made a copy or a restored version. For that, we ran it through a whole bunch of devices, but mainly the Sonic Solution.
FG: How familiar were you with these sounds or this music from the Middle East and North Africa before you started working on it?
LV: Yes, thank you for your question because it’s very interesting. It really opened my ears to these sounds. I hadn’t had the opportunity to listen to or experience this music much. And for me, it was really the Cairo Congress that introduced me to Arab music, which allowed me to listen to it the most and get to know it better. And the more I hear it, the more I want to know about it. And while preparing for our interview today, I listened again to Jean Lambert’s lectures and listened again to certain records so that I could see what we were going to talk about today. And it’s true that it really made me want to go further.
Due to a lack of time, I really can’t go as far as I would like to. Because it’s true that when you work on this kind of background, there are two types of people. There were people like Pascal Cordereix or Jean Lambert, who are real music specialists, who know and can translate what people are singing, recognise the instruments and all that. And then there are the technicians, and we’re not specialists and, indeed, we often limit ourselves to the technical aspects of the transfer. We make sure that all the work is done very well, but we don’t necessarily have the time or the inclination to take an interest in these sound archives.
But the more I listen to them, the more I want to go further. I’m not necessarily looking for the same details. Therefore, I find it very interesting when we do sound restoration campaigns on certain collections or even sound recordings (In parallel, I’ve done a bunch of classical music recordings). So, I’ve worked with musicians and I’ve realised that it’s quite interesting to have high-level musicians, but also high-level technicians. This combination meant that everyone took care of their own area, but that one could contribute something to the other. Eventually, it was good to have two different perspectives on these different aspects.
Typically for sound restoration, it’s good to have experts in ethnomusicology on one side and technicians on the other, and that this kind of work proceeds from this kind of dialogue. If the result of this collaboration [on the Cairo recordings] was so interesting, perhaps it is because of that…
I limited myself to making the restoration as transparent as possible. But of course, it’s never transparent. From the moment you start restoring, you’re removing something. So we did affect the useable signal a little. But in any case, I endeavored to only remove noises that were distracting. We discussed this issue with Pascal Cordereix. I showed him what I was going to do. We agreed on the choice together and I started to restore it. It was a long work: it took me three to four months to restore everything. In any case, I like this dialogue between the musician and the ethnomusicologist and the technician. Some people can do both at the same time. But I really appreciate this dialogue…
FG: I’d like to go a little further in understanding this dialogue, or this triangular conversation between you, Lambert and Cordereix, were there either guidelines on their part, or proposals, or was it based on open discussions and interactions between you, where you provided arbitration? I’d like to understand how these discussions and decisions impacted the aesthetics of the sound that emerged.
LV: Frankly, since we had been working together for a long time and they had seen the work we had already done, I made some suggestions and they accepted them, but for some restoration work, there was a rather interesting dialogue between the other musicologists and us about how far to go or not to go in the restoration. We could remove more, we could remove less, and they allowed us to set the bar a little higher or lower to know where we should go or not. Nevertheless, we always have an idea in mind: we were trying to provide a reference version. It means that if we have to make a choice, it’s always a little bit restrained.
When it comes to sound restoration, very often for 78-rpm records, some people who do restoration will try to mask and eliminate all surface noise. And we’ll hear more than the instruments, but almost nothing in terms of surface noise. In any case, we insist, and have always insisted, on maintaining this authentication that we are working on an archival document. So, the person listening to the record knows that it is a 78-rpm record that has been restored. It’s not something that has been completely smoothed out. We can understand and see that it is an archival document.
This is especially true regarding the Cairo Congress’s recordings because we were very unobtrusive in our restoration work. Basically, I de-clicked the recordings, filtered them slightly and then I removed some noises that I found annoying. That’s basically it.
We left as much music as possible. Especially knowing that ethnomusicologists and specialists would be working on it, we wanted the bandwidth of the instruments to be affected as little as possible. We tried to transcribe all the richness of the instruments as best we could, while maintaining sufficient listening comfort so that our ears are not distracted, processed by something that prevents us from listening to the useful signal, which is still the basis of the recording.
FG: To sum up, both Cordereix and Lambert gave you carte blanche in terms of sound processing, based on your ideal of sound ethics, if I understand correctly…
LV: That’s right. The whole issue with digitising archives is to be as faithful as possible to the original sound that was captured and then recorded on a medium. My goal is to ensure that the listener can hear as much as possible of what was recorded and just written onto the medium.
However, the medium, particularly discs, has its own flaws, such as clicks and surface noise. I allow myself to remove them as long as they have very little impact on the useful signal, the source that was intended to be recorded. That’s why we are always very careful not to correct or equalise. Or if we do, it’s only in very specific contexts, which means that when you listen to the Cairo Congress [CD recordings] and the original [78 rpm] discs, it’s normally the same thing.
Afterwards, of course, the restoration process and a little bit of filtering help to bring listening comfort, allowing musicians to focus on the music, on the useful signal. Listening to surface noise is boring, but having all that richness in the bass and treble, that’s very interesting.
Sometimes we also try to remove alterations caused by time, because some records may have been played a lot or played on Gramophones and therefore damaged. Of course, that’s not the case here, because the records were in good condition. For some fortepianos, for some instruments playing extremely high notes, if we can try to remove certain distortion defects, I took the liberty of doing so, but it’s really to bring listening comfort with the idea behind it, not to betray the archive.
I also want to allow researchers to have access to all the information. That’s why, in the booklet, we have described exactly what the restoration process and approach were. I have mentioned that we filtered out certain things or removed certain noises, but only because they have nothing to do with the recording of the musician or the instrument.
When listening back to the Cairo Congress [original] recordings yesterday, I can see that there is a lot of humming due to filtering problems with the motors or equipment used, so we have a frequency that is typically around 192 Hz. It’s a 200 Hz hum that is quite present and has nothing to do with the music, so I removed it, or at least, I tried to make it less present so that the listeners are not distracted.
There’s another case that I listened to yesterday. I don’t know what’s going on, but it’s like people dropping objects on the floor or maybe doors slamming. These huge booms are present during certain recordings. So, I removed them. Since we now have tools – unavailable before – that allow us to see a sonogram, so we can see the sound with its fundamentals and harmonics. Then we can restore the image, which will be translated into sound again. This software allows us to take the sound signal and turn it into an image called a spectrogram. We are able to make corrections on this image if there’s a loud noise in a specific place, at a certain pitch. We can copy it, paste it, reduce it or patch it and then return to the sound.
This means we know exactly where we are working. It indeed takes a long time, but it’s almost a surgical restoration. The problem with these tools is that they take a lot of time because many things are still done manually. Fortunately, some things are automatic, but others are manual, and if we want to do it properly, it takes time.
FG: On the same note, I’d like to know what difficulties you encountered in the process of both digitisation at the outset and sound processing.
LV: I’ll finish with the digitisation process. So, as I told you, in 1996, we did this first campaign where we read the discs in the best possible way. We made a straight copy, which is our master, and then from that straight copy, we made a very slightly restored copy. When the reissue project with Abu Dhabi came back to us in 2015, I reworked this restored copy based on the recordings we had made on digital audio tape (DAT) and then transferred it again to our digital files archive.
I started again from these files and took the already restored version. What I did was a second round of restoration with tools that were almost 20 years older. This allowed me to use these new tools and technologies, to look at what had been done and sometimes go further, or else I would stop there.
Regarding the restoration this time, since I had all the files and had been given the time and permission, I was able to listen to each ensemble in its entirety. So, what I tried to do with the information Jean Lambert had given me was to identify the ensembles to see if there were any flaws affecting each ensemble. And of course there were, because these ensembles were recorded on such and such a day, and then we moved on to another ensemble.
In fact, each ensemble has its own sound recording flaws, bloating, saturation, noise, things like that. So, in fact, the filtering was the same for the different sides that corresponded to a particular type of orchestra. So, I was able to look at the different problems on each disc, group them together and deal with them more efficiently. Then there was a very long process of simply taking each disc, looking at everything and if there were any specific manual restorations to be done, I did them.
It really involved listening to everything in the right order, ensemble by ensemble, and to all the recordings. I feel strongly about that too. In fact, it’s part of our ethics. When we digitise a collection of discs, we try to ensure that things are processed at the same time and that they are in the right order. And since each mark generally has its own colour, the pressing has this or that defect. Recording machines are generally pretty much the same, as are pressing machines. So, this usually allows us to identify defects that we can correct and then apply restoration settings to the entire collection.
In fact, this means that I was able to apply nearly the same treatment to a particular category of ensemble. So, in fact, when you listen to an ensemble, it’s no different from one side to another or from one disc to another. In fact, I tried to make the treatments consistent, at least in terms of the different types of ensembles that were recorded. That’s quite interesting. You can clearly see that there are moments when there is no humming and other moments when there is humming all the time. So, adjustment, relocation, did they change rooms? Did they make a recording in another place? I don’t know.
This is something that happens quite often with the low frequencies. When I revealed the lower part of the bass spectrum of the recordings in the spectrogram, I realised that sometimes, there were a lot of lorries passing by. So, we have these spikes in the bass frequencies, which I tried to remove. That way, when you listen, you hear the oud and not the lorry and the oud.
When I’m working on restoring and transferring sound archives, I’m usually very keen on the sound recording that was made. Because I’m a sound engineer and I do sound recording for classical music. But above all, this whole business of recording equipment is something that fascinates me. I was just blown away by the quality of the recordings, which are really good, despite the significant surface noises. Indeed, 1932 was the era of 78-rpm discs with electrical engraving and the discs will already spin.
We also have a microphone, so with that, we already have the high fidelity of 78-rpm. On that side, it’s perfect. Additionally, as we play these 78-rpm records with modern hi-fi equipment, with cartridges that go up to tens of thousands of Hertz, in fact, we can get extreme highs and extreme lows.
And since we make flat copies, we can really reveal the full richness of the record. What’s really interesting is that we can hear things that the technicians of the time couldn’t hear. They might have heard some of these low-frequency noises, but not much. When we listen to it on an active monitoring system with good settings and a subwoofer (that can reproduce the bass), when there’s bass and a lorry goes by, we can hear it clearly.
It’s the same for very high frequencies, for example. Resonances that may be present in the recording due to engraving problems, where there is resonance from the engraving tool, which causes this resonance that is usually recorded at the end of the disc, where we have frequencies around 15 kHz down to 8 or 7 or even 6 kHz. So it will go down in a glissando. Typically, these are things that will become much more apparent because we have devices that allow us to transcribe this information so accurately that we almost highlight them. So, these are flaws that we almost highlight.
And what interested me a lot in the Cairo recordings was seeing that the extreme low and extreme high frequencies had been recorded. That was very moving. Concerning the instruments themselves, what was fascinating was seeing their capabilities. I can’t recall the name of the instrument, but there was a very large flute that could produce very low sounds that were very interesting to hear.
I also remember that there were some rather rhythmic, metallic instruments. I don’t think it was a triangle, but if my memory serves me correctly, it was in the Coptic chants, where there was a kind of thing that was tapped, perhaps to give a little rhythm, like a small bell. It was great to say that at that time, we already had access to such high-fidelity recording and that the instruments were perfectly captured and reproduced thanks to the 78-rpm. So that was amazing.
There is one body of work that touched me deeply. That’s because it’s part of my beliefs, everything that is Coptic chant, the whole Coptic mass, that interested me a lot. I also liked everything that was called Nouba, that are a bit more danceable, where you feel like you’re in the Maghreb and you enjoy it a bit. Of course, not knowing the language and not having the translation of the songs, whereas Jean Lambert had shown me some of them, so we could see that it was really poetry, with all that that entails. It’s true that if I had made the transfer with the translation, I might have appreciated it differently. Then there are…
So, listening back to Jean [Lambert]’s lecture, I get the impression that he was saying that the artists were of different levels, that they weren’t necessarily always great artists. And we can indeed see that too. (…) So, there are different levels of interpretation, which are more or less rich and which touched me differently, but I appreciated this diversity.
But again, it’s true that my work was really focused on transcription and restoration, to make the signal as authentic as possible. And I didn’t find the time to take much interest in the musical aspect of things. I would have liked to, but frankly, we didn’t have the time. And I’m really focused on that. But on the other hand, it was precisely by listening again and preparing for this interview that I really wanted to go further.
Very quickly, my curiosity about recording technologies, who recorded it, how it was done, which company, when it was done, and so on, piqued my interest. So, last night I did some research on the sound engineer who made the recording for the Gramophone. I tried to find photos showing the microphones that were used. That’s where my passion really lies. I leave the musical side to the ethnomusicologists.
FG: Yes, but at the same time, I was almost convinced that a sound engineer’s perspective could enrich the perception of a musicologist or a listener.
LV: That’s true. I completely understand that. As I said earlier, it’s very important to have this dialogue between musicians and technicians, because one can shed light on certain things that the other doesn’t really know about.
Regarding the Cairo Congress, I was really surprised by the quality of the recordings and the engravings that were made. I consider it to be a monument to Oriental music or Arabic music, I don’t know what to call it (it used to be called Oriental and then it became Arabic, perhaps partly because of the Cairo Congress).
It’s a fascinating world unto itself, but it’s true that the language barrier already takes away that slightly poetic aspect that would perhaps help us to understand it better. What I find interesting is the good selection that represents well the context and the state of mind of the performers at the time. You can really see that they are all passionate musicians and singers who put their whole hearts into it. You can really feel that in the recording. These aren’t just recordings for the sake of it. It’s as much a technical achievement as it is a musical one. That’s how I felt about it, anyway.
FG: So perhaps as a last additional clue to this possible growing connection between technicians, musicologists and musicians, you said you were particularly struck by the very high and low sounds in the Cairo recordings. What does that say about this corpus or these music genres?
LV: It means that in these areas [i.e. Arabic music], which one might think are rather poor in terms of sound richness; when they are well recorded, we can see that there are a whole range of nuances that are very present (…) So, we know this about European classical music [where] it’s clear when a piano sound is very low or very high (…), but perhaps it’s true for Arabic music…
We might consider it as limited to a tambourine or a few violins, but it’s not much richer than that. Whereas if the sound recording is well done and the musicians are really good performers, then we can see that there are real technique and musical quality, a quality of interpretation and an interesting instrumental richness. Typically, in a specific [Cairo] recording, some drums go down to extremely low pitches. If they are well recorded, we are immediately transported to another context. It’s completely different.
Then again, I won’t hide what needs to be said. Certain sounds hurt my ears, too. In any case, I was glad to move on to something else.I don’t have any clearer memories, but some songs are a bit monotonous or maybe certain pieces where the violin was always playing the same thing. Finally, you get tired of it!
Indeed, we’re really used to richness in musical and instrumental composition and it makes us feel good when we find it. Therefore, I really appreciated moving from one universe to another and, to have understood it better a few days ago, from one country to another, from Syria to Iraq, for example. We can clearly see that there are completely different sound signatures, let’s put it that way, due to the ensemble present, the voice. Tunisia is not Algeria, which it is not Morocco. Each time, we go on a journey.
FG: The ultimate question. These Cairo recordings were published in 2015. Now, are there digitisation or sound processing tools or approaches that have emerged since then that could have helped you at the time? Or that you would have done differently thanks to new technologies and new approaches to sound?
That’s a really big question. As I told you, the digitisation was done in 1996 and the recording medium at the time, to collect this digital information, was DAT, then CDR. The resolution used was 16 bits and the digitisation format, the sampling frequency, was 44.1. Nowadays, everything we digitise is in 96 kHz and 24 bits. We’ve gone from CD quality to a digital HD version. So, we could say that, if we had the time, we could re-digitise everything in high definition.
But is it worth the effort? I’m not sure from an ethnomusicological point of view. And that’s something I realised while working on 78-rpm archives and the Cairo Congress in particular. The technologies that were developed 5-10 years ago for sound digitisation were optimised for CD quality at the time. Now, that has changed somewhat. We may have tools that are better in terms of algorithms, performance and so on, and that could allow us to go a little further. But it’s so much work that I don’t know if, in the end, the results will be worth it.
Nevertheless, for the conservation and preservation of our heritage, it would be a good idea to re-digitise everything in high definition anyway, so that we have a master copy, a ‘straight’ copy from the originals. That’s something we need to do at some point. We’ll do it, that’s for sure, because we’ve moved on to a slightly better digitisation protocol.
Concerning the tools, they are evolving a little, but not that much. Anyway, sometimes it’s a matter of small details that add up and mean we gain a little something. But given the number of hours and the work that went into it, I think we can be proud of the work that has already been done.
For me, it’s the Coptic repertoire which affected me a lot, because I’m a Christian. I realised later that they were Coptic songs because I hadn’t understood that at first. I was taking care of my transfer, the catering and all that, when Jean Lambert told me, ‘be aware, those are Coptic recordings, and it’s a mass’. I understood that in the Arab world, Christians were given a place, and this touched me deeply because Arabs [meaning Muslim Arabs] and Christians don’t always get along. In any case, now we are less than we could be, less than friends, at least.
But in those times, we knew how to get along. And then, what also touched me a lot, even though I’m less familiar with that music, was hearing Jean Lambert saying that most of the musicians were Jewish. So that’s also true of certain ensembles (FG:particularly in Iraq…). And I thought to myself, it’s incredible. These are people who are fighting each other. But there you go, we have the best performers of a certain type of Arabic music, who are Jews. So that’s the fraternal side of it. That’s what really impressed me. It’s seeing that in this music, at that time, we were able to be brothers.
[1] “Son direct” in French refers to the sound recorded on a film set. A finished film is a blend of direct and indirect sound, added in post-production. In France, Jean-Pierre Ruh used to record “direct sound” for documentaries and TV reports. He first used it in cinema with the director Eric Rohmer in the movie Ma nuit chez Maude (1969).
In his article below, published after his death (2006), Jean Pierre Ruh wrote: Contrary to what has been said, the Nouvelle Vague and Godard were not at all taken with direct sound. It was Rohmer who grasped its potential, with all its inherent imperfections, its breaths, its background noise, the acoustics of the location. It was a true revelation that a director would embrace the “son direct” to such an extent, with such intelligence, sensitivity, and thoughtfulness, and a tremendous opening for other filmmakers, like Resnais! Ultimately, my entire career has revolved around this pivotal film – since I am and remain a staunch advocate of the “son direct.”
FIND OUT MORE