I came across Studio Melodium whilst Googling about studios in Paris and I later discovered that the owner, Nicolas Dufournet, had been an artist and record label guy before becoming a studio owner. Melodium is one of the more affordable studios in Paris, and has even developed its own recording philosophy, dubbed “Classic Sound Techniques“. I sat down with Nicolas to learn more about this approach, and the result was an extensive conversation about not just recording, but also his time as an artist, his tenure at Virgin Records, and his philosophy behind creating Melodium. We also broke down some 60s and 70s records to see how they were created.
– Hi Nicolas. Thanks for having me over. Can you tell me how you got your start in the music world?
Sure. I was introduced to music through the records my parents played at home. Those records motivated me to become a musician, although like many other thirteen year-olds I also wanted to impress the girls at school. I chose the bass because my friends told me a four-string instrument would be easier to play than a six-string, and the story behind my first bass guitar was kind of funny. My parents had hired a 30-year old painter to work on our house, and since I didn’t know of any music stores in town he said to me, “I’ll buy you a bass if you give me the money “. I naively trusted him to keep his word, but after one week the bass still hadn’t arrived. When my father found out, he went looking for the guy and told him, “Listen, you’re going to take my son to the store and come back with a bass guitar, or else “. And that’s what happened (laughs). That was in 1980, when I was fourteen. My father also gave me his Sony tape machine to record with, which lasted me a few years, but I replaced it with a Fostex four-track cassette recorder in 1983.
– And at some point you joined a band called Oui Oui, correct?
Yes, that’s right. The other members were older than me, but we became friends because we were pursuing the same girls at school. The band formed in 1981 thanks to Etienne Charry and Michel Gondry, and I joined a year later along with Gilles Chapat. They were looking for a bassist and I tried out for the spot even though I didn’t consider myself a good player. But I ended up getting it because I wasn’t as bad as I’d thought.
– Would you say that Oui Oui had a successful career?
Well, we weren’t interested in following the trends of the decade, which probably limited our success. Most other bands were trying to sound like Joy Division, but we preferred groups like Talking Heads, and people used to say that we sounded like Devo on guitars, although that never helped us achieve success. Back in the 80s, it was very hard to become a successful rock band in France unless you conformed to the norm, and everyone else was making “serious” music at the time, whilst we preferred things that were light-hearted. But some people in the music industry believed in our sound, which is why we ended up having a career, although we never received regular radio play. We sold around 17,000 copies of the first album and 3500 copies of the second one, which was very low considering the money spent by the label on recording, touring and marketing. But it was what it was.
– What were some of the differences between being a band in the 80s versus today?
One major difference was how slowly things moved back then. To even find the name of an A&R was really hard – most of the time you needed a friend to connect you with someone at a label. Secondly, we weren’t the best musicians, although Michel Gondry was a good drummer, but we needed time to rehearse our songs. That made things even slower, and we also assumed no other French bands were interested in our genre of music. There was no Internet to show us what was happening in a neighboring town, much less the rest of the country. So our mindset was different from what I see in today’s bands, who expect things to move quicker. Back then, working in a studio was probably five times slower than it is today – first you had to bring the tapes with you, which could have multiple versions that weighed a ton. Then you had to listen back to find which version you wanted, and if you’d brought the tapes from a different studio, the engineer might say “Sorry, our machine isn’t properly aligned to work with your tape because you recorded on a different model “. You can’t imagine how slow things could get at their worst.
Another difference between today and the 80s was the expense. Analog tape cost a lot of money and there were times we had to record three versions of the same track in order to justify using a reel – it would have been a waste of tape otherwise. A two-inch reel cost around €800 in today’s money, so we had to rehearse a lot before recording, which many bands don’t do now.
– So how did you transition from being in a band to working at a well-known label like Virgin France?
I was the only one in Oui Oui with a full-time job as a media planer, which I continued doing after the band split up. In 1987, I got a job at Publicis, but I later said to myself, “Maybe I can work at a record label. Companies like that always need advertising guys and being a musician is probably a plus “. I turned out to be right, and Virgin hired me in 1994 to do marketing for French artists like Alain Souchon, Axelle Red, Renaud and Jean-Louis Aubert from Téléphone.
– Can you expand on how you got the job at Virgin?
I’d worked in advertising for long enough and wanted to return to the music industry, but I wasn’t sure whether to apply for a label job or become an artist again. I’d stayed friends with Luca Minchillo, the A&R who originally signed Oui Oui, and he was working for a Virgin sub-label called Delabel. I asked if he could help me get a job at Virgin and he said, “Sure, just give me your CV and I’ll pass it on “. But I never heard back from him, so I decided to pursue my artist career instead. Then the phone rang a few months later and it was the boss of Virgin France, Emmanuel De Buretel. He said, “Hello Nicolas. I’ve read your CV and heard good things about you. Could you pay me a visit to discuss an opportunity I have for you? “. But I answered him, “I’m sorry, but it’s too late for me to take a label job. I’m working as a musician now “. Two weeks later he called back and said, “I don’t think you understood what I said. I’d like you to come and have a chat with me so I can explain what I’m offering “. So I realized he was pretty determined and I should probably go see him (laughs). He basically asked me to work for Virgin, and if I didn’t like it after one year I could resume my artist career. So I accepted his offer and stayed at there until 2002.
– Since you were at Virgin from the mid-90s onwards, you were present during the Daft Punk era. How was that?
I remember their first label meeting because they sat a few meters away from me, waiting for an hour until my boss to arrived. At the time they were just two shy guys in the hallway and one of them had a striking resemblance to Johnny Hallyday, which was obviously Guy-Manuel. I never worked on their projects, but I remember how excited Emmanuel was about “Homework“. In fact, he took all 140 Virgin staff members into a wooden church and locked us in to play the album at full volume. You’ve never heard music that loud in your life – it probably took 3000 watts to power the speakers, and it was so loud that it overwhelmed everyone. But that was how he sold us on the album, and the label got behind it after that. So he knew what he was doing.
Sure. Since it was twenty five years ago, I can tell you a funny story about working with Air. I became the Marketing Director for Source in 1998, but three years prior to that I worked alongside the label’s Artist Manager, called Marc Teissier du Cros. The founder of Source, Philippe Ascoli, told Marc that he needed a replacement track for the Source Lab compilation because the one he was expecting from Roudoudou still hadn’t arrived. Marc was friends with Nicolas Godin from Air, and he said, “I suggested numerous times that we put Air’s track on the compilation, but you kept turning it down because you didn’t like it. So let’s just do it now “, and Philippe was like, “ Alright fine, let’s do it since we have a free slot, but I still don’t like it “. So when James Lavelle from Mo’ Wax heard the compilation, he reached out to Air and offered to release their “Modulor Mix“, since they didn’t have label deal yet. Air then called Virgin and said they’d been contacted by the most exciting indie label in the world and that we should probably start taking them more seriously. When my boss heard about that, he wasn’t happy. He was like, “Guys, if you don’t sign Air to a proper international deal right now, you’re all fired! “. So that’s what happened, and Air went from a totally anonymous band to being on twenty French magazine covers within two years of their album release, in addition to getting tons of press from abroad. For the first time, a French band was generating real interest from the American and UK pop industry, and French journalists found that remarkable. But there was a reoccurring theme in all the headlines: Air’s success came from being launched by the British music press thanks to Mo’ Wax, not Virgin France. Even Daft Punk had their first release of “Da Funk“on a Scottish label, Soma Recordings, not a French one. But nonetheless, it made Source into a respected label because we helped generate the kind of interest that the French music industry hadn’t achieved in over 50 years, and that’s why many artists wanted to be on the label.
– It’d be nice to hear some more stories about the artists on Source. Can you tell me anything interesting about the following:
Phoenix: We used to live in the same town, but we never hung out because of the age difference; they were seven years younger than me. But when we finally met, I thought it was cool that they were into American jazz-rock acts from the late 70s and early 80s, like Donald Fagen. Most people from our town didn’t listen to that in 1998, and when they started working with Philippe Zdar on “United“, I knew they’d be in good hands. It was clear that they wanted the album to showcase the diversity of American music, which is why it featured everything from country music to 80s-inspired hits, as well as stuff that reminded you of Paul Simon and The Beach Boys. It’s one of the best French rock albums I’ve ever heard, and it was also my first worldwide album release as Source’s Marketing Director.
– Did you work on “United” at all?
We were all in the studio together when Philippe Zdar asked me to play bass on a track, but I wasn’t sure if his request was a joke or not, so I didn’t take it seriously. It was 1999 and I hadn’t played bass in seven years, so I didn’t think it was a good idea to try. (laughs).
Sebastien Tellier : The first question he asked me when I met him was, “Do you know anyone who can make me a pair of diamond-studded shoes? “, and I responded, “I like your question, and I’d be happy to get you some diamond-studded shoes if you can give Source a diamond album ” (laughs). It was partially a test question, but I think he was also saying that his brand was unique and would catch people’s attention. He and I only had a few interactions. I think he felt the label didn’t understand his music too well, and he preferred using his manager to communicate with us. But I always thought the secret to his success was a careful combination of innovation and fashion, but with a very delicate balance. So he was always a step away from either being a genius or looking ridiculous, and it’s very hard for a record label to sell something like that. I was one of the only guys at Virgin who said, “This guy is very talented, but it’ll take a lot of time to present his ideas clearly to the public “.
David Axelrod: Source was the destination of choice for many indie labels that wanted their records licensed in France, like the Beastie Boys’ Grand Royal label and Mo’ Wax. Even Warp had a licence deal with Source, and they’d would send us new records by Aphex Twin, Boards of Canada and Prefuse73. So I think David Axelrod’s self-titled album came to Source from Mo’ Wax, and we did some interviews with him to promote it.
– How did you end up leaving Virgin in 2002 to create your studio? Didn’t your boss object to that?
I had to explain my motivations of course, and I told him, “Thank you for the job, but it’s time for me to leave. I learned a lot more at Virgin than I imagined, but I need to have some self-respect and return to what’s most meaningful to me, which is working in a room with musicians “. And so he let me go.
– What was the biggest challenges in setting up a studio in the early 2000s?
I had already seen how record labels had failed to answer the MP3 problem, and the coming decline of the CD was obvious. So any new studio would have to be built on a different model. I knew that artists would have less money to record than ever before, so the new model would have to charge only a third of the regular price. Big places like Studios Ferber were charging €1000 per day, and Studio Davout’s prices were around €1300. Smaller studios charged around €750, so I knew I had to build one that could survive on €250 per session, which was very risky. But I eventually succeeded.
– Did you have a specific vision for the type of studio you wanted?
I always preferred working in studios that had lots of instruments and effects because they help create excitement for the artist. Running a studio isn’t about using a tape machine or staring at a VU meter – it’s about helping the artist achieve their vision, and gear can sometimes distract from that process. That’s why I get upset if my equipment has problems, and I hate museum gear that just looks nice but doesn’t work. So instead of being precious about a compressor, I prefer to pay attention to the skills of the musician, the choice of instruments and the musical performance, which makes the recording process easier. Granted, mixing can sometimes be challenging, but let’s take The Beatles as an example: they’re music is classic, but even a ten year-old can mix a Beatles record because it was mostly done on four and eight-track machines. So a mix can be simplified based on the choices made during the recording stage.
– But if you wanted to recreate the sound of your favorite recording, then you’d have to follow certain steps, regardless if they were complicated or not.
Yes, recreating the sound of your favorite records can be challenging, and even if you know what gear was originally used, you may not know the order of each device in the chain. If you change the order, the result could end up very different.
– I had a similar conversation with Fred Falke about how the order of gear matters. He explained how Michael Jackson’s music went through far more transformers than just ones on the Harrison 32. But what if the process involved fifteen steps of equipment? You can’t reverse-engineer that.
That’s true. After passing through multiple input and output transformers the signal quality will have changed significantly due to added noise and coloration. But I wouldn’t overemphasize the role of transformers because not everyone can hear those subtleties. What I would emphasize is the difference between a kick and snare going “boom, clack”, as opposed to “thump, smack” – that matters. The way a drummer plays his kit is what makes the drums exciting, far more than tape machines or transformers. So you have to figure out what kind of performance you want, and only then can your engineer amplify that. A good engineer is someone who acts as an extension of the artist’s ideas, not just a button-pusher on a console. So if the idea is exciting, it makes the technical path much clearer.
– When I talked to Gerry Brown, I mentioned that Ronnie Laws made two albums: “Friends and Strangers” and “Solid Ground”. Yet the second one was considered to be inferior to the first, and Gerry agreed. He said it was partly due to newer Neve consoles and other technology changes. So sometimes gear can have a deciding impact on a record, regardless of the artist’s vision.
But it’s an engineer’s job to know his gear. If an album sounds bad because of the gear you have to question the engineer. We recorded all of Oui Oui’s music on Neve V series and an SSL 4000. Sometimes we thought our guitars sounded like crap and at other times we thought it sounded fantastic, but we had no excuses either way because you can achieve any sound you want on a Neve console, regardless of the model. Maybe if you’re stuck with a Behringer mixer it’s a different issue, but even a broken-down Neve is manageable. I’ve worked with worn-out Neves and I was able to figure out which channels were problematic, even though it took twice as much time. The issues tend to be intermittent and unpredictable, but I ended up isolating ten usable channels on a 36-channel console. So nothing is impossible. But be careful once you send your record to be mastered; your mastering engineer can mess things up by cutting too much low-end, boosting too much at 10 kHz or adding too much stereo spread. That happened to one of our albums – we lost weeks of work because of bad mastering. But if the engineer knows his gear, you should be fine during the recording stage.
– Let’s get back to your history. How did your first studio come about once you left Virgin?
I started off in a small space that used to be a basement for go-kart maintenance, so it conveniently had absorbent walls to isolate the sound of their motors. The owners had left by the time I took over and I was able to set up two live rooms there: one that was 22 m² and another that was 10 m².
Back then, the average French audio engineer worked only a few days per month, so I realized that if I spent every other day recording, I could definitely learn how to be an engineer on my own. So I became a studio rat for two years, even though things were slow in terms of clientele; not many people thought I could do a good job if I only charged €250 per session. So when I didn’t have clients, I’d take a Bee Gees or Motown album and try to recreate their sound. I made a lot of mistakes, and sometimes I’d achieve the sound without knowing exactly what I did, but I eventually came to understand the process. I also read a lot of books on recording, but it’s hard to hear what 3 db of gain reduction sounds like just because a book tells you how to set up a compressor. As a former musician, I wasn’t used to hearing that level of nuance – I came from recording loud rock music to 4-track machines in a small room where you can’t hear any subtleties, so it took a lot of time for me to become conscious of the four most important aspects of engineering: dynamics, EQ, space and modulation. I also had to learn how to organize those concepts in my head and to make the proper combinations of them.
– What kind of gear did your first studio have?
In order to attract clients, I wanted gear that no other studio had, so I switched out my TASCAM console for Ampex 351 tube pre-amps, which weighed around 10 kg each. They were designed to be mic pre-amps, so each one had seven tubes inside that were a pain to maintain. I bought eight of them on eBay from the US, so you can imagine how banged up an 80 kg shipment was when it arrived. I had to find a French technician who could help restore the dead transformers and tube sockets, and thankfully I found Patrick Noury, who runs a repair store called Musical Service in Pigalle. He used to be the Head of After-Sales Technical Service at Roland France and he showed me how to do things like use the Ampex with only four tubes instead of seven.
I couldn’t afford a Fairchild 660, so I got an Altec 438A because Abbey Road had the Altec 436B. I also had some vintage mics, like a U47, four Geffel UM 57s, four Neumann KM54s and a KM56. But a lot of those mics developed broken capsules because they were old, and Neumann suggested I buy new mics because they no longer serviced their vintage ones. But since I used to live in Germany, I was able to contact some old techs there who re-skinned the capsules for me.
Another type of gear I bought was things that looked visually interesting because I knew that aesthetically pleasing gear could produce interesting results, even if the sound was underwhelming. If you ask a musician to play a broken guitar that looks beautiful, you’ll sometimes get better results than if he played a great-sounding guitar that looks like crap. It’s a psychological thing, and I’ve seen guys who had a lot of fun with broken guitars, even when I told them the strings were rusty and would give their fingers tetanus if they played it too long (laughs).
– What were some of the costs and risks involved in setting up your first studio?
One of the risks was falling short on my €600 monthly rent payments when I only had three clients who I charged €150 – €250 per day. I also had maintenance costs for the eight Ampex pre-amps, as well as my Telefunken V72 and the Altec 438A. I was one of the first studio owners in Paris to have that much vintage tube gear in the mid-2000s, and there was very little expertise around for maintaining it. So I ended up selling all the gear because the maintenance became too much work. I remember selling the Ampex 351s in 2006 to Manfred Kovacic, who now runs Vega Studio, and the rest were sold to engineers in their 50s from places like Plus XXX and Studio Pathe Marconi who knew exactly how hard it was to find stuff like that in Paris.
– How did you find your current place, and why did you name it “Melodium” ?
In 2004, one of my friends from a band called Concorde recommended I check out this place. It used to be a hip-hop studio where the owners had been threatened by a gang to hand over the keys. They didn’t want any more drama with that, plus home studios were becoming popular and rappers were starting to record at home, which meant less business for the former owners. So they decided to sell their place to me.
Regarding the name “Melodium”, I was looking for interesting microphones and found a guy who was selling a Schoeps, which I thought might be a less expensive version of a Neumann. But when I tried the mic at his home, I wasn’t that impressed by the sound. So in his attempt to sell me something else, the guy brought out two Melodium 42Bs, which blew me away, and I bought them for €600. At the time, nobody except a few older engineers had heard about the Melodium 42B and there was nothing online about them, so I named the studio after the mics because brand was discontinued in 1966 and the name was free to use.
They’re very dark-sounding, with an impressive presence in the low-mids, around 200 Hz. They work great on kick drums, even though a Coles 4038 would do a similar job with less bass-response. I generally prefer the 4038s because they sound more balanced, but when I run out of ribbon mics I grab the 42Bs. But you can’t use them on a quiet source like an acoustic guitar – they’re too noisy for that.
(Below: Studio A at Melodium)
– Tell me about your “Classic Sound Techniques”, and how you’ve constructed Melodium around this idea.
A large part of Studio Melodium is about using vintage gear to do new and interesting things, which is what many 70s punk bands did when they played cheap instruments. It’s why Captain Sensible from The Damned played a Höfner violin bass, and even modern groups like Les Rita Mitsouko had one of those; so did I in 1987. So I didn’t want Melodium to only be a “vintage” studio, and “Classic Sound Techniques” are the ways I draw inspiration from the past, but with a modern perspective that lets my clients achieve a balance of old and new.
– On the Melodium website, you’ve contrasted two types of recorded sounds: the “big” sound and the “warm” sound. Can expand on that?
Sure. The “big” sound is about having multiple microphones at different distances from the source so you can capture different aspects of the instrument. The closest mic captures the attack and transients, whilst the second one captures the body and the third one gets the tail of the sound and room reflections. So if you balance those three mics adequately, it allows you to achieve the “biggest” sound possible.
– Does that mean “more is more”? Why not used seven mics instead of just three?
I’ve never had a big enough room to use that many mics (laughs). Also, three mics alone would result in a crazy track count if you used them on every source, so I don’t know if seven is necessarily better. Back in the day, most bands didn’t play in huge studios like Abbey Road. They mostly recorded in 35 m² studios, which is the room size you hear on many 60s and 70s records. So a three-mic setup might be positioned ten centimeters, one meter and four meters away from an instrument. If you put another mic six meters away, it would pretty much become a reverb mic. But beyond that, I don’t know if adding more mics helps.
– You’re saying that a “big” sound is about placing mics at a lateral distance away from the source. But what about height? On a drum kit, overheads are usually placed above the kit, not further away from them laterally.
Drum overheads are placed above the kit because the skins project sound upwards when you hit them. So having mics placed above is logical when only the kick directs sound laterally. But putting mics far away can be useful regardless of their position because distant room mics have random phase responses, which can useful for your mix.
The “big” sound isn’t only about distance – it’s also about musicians playing together. When a band is tight with their playing and have microphones around them at different distances, it creates a “big” sound because of the bleed and combined accents of all the players. It’s simple to understand: if you tell three people to punch someone at the same time, it has more impact than just one person throwing the punch. So musicians that hit their notes simultaneously will sound more forceful. But keep in mind that the “big” sound is ultimately about creating illusions. If you look at a picture that only has large people in it, they won’t seem large. You need small and large people standing side by side to see the difference. So “big” sounds need “small” sounds too.
– And what about the “warm” sound?
The “warm” sound is about understanding music culture, and how it’s different across the world. For example, many people in Asian countries consider the “warm” sound to be a boost at 10 kHz, whereas for Europeans and Americans it’s a boost at 200 Hz – 300 Hz.
– How can 10 kHz be considered “warm”? Isn’t that “bright”?
You say that because you come from a European culture, but “bright” means “warm” in many Asian cultures. If you’re mixing a record for a Japanese guy, he’s going to ask you to boost at 10 kHz and call that “warm”. Let me ask you something: why is it that high frequencies in microphones and speakers have increased in every decade since the 70s? Is it because most of our gear is made in China? That’s not necessarily the answer, but it’s a theory. When you compare a German-made mic to a Chinese-made mic, the difference in treble is very clear, and it only gets more pronounced over the decades. Also, if you look at music from the 1940s, there’s a frequency roll-off under 150 Hz and after 5 kHz, which also lessened over the years. Sure, audio equipment got better, but it’s also a cultural thing because it didn’t happen at the same rate everywhere.
My father never went to a rave that had tons of bass and treble blaring from the speakers, so he’s never heard that kind of thing in his life and it doesn’t interest him. But I had my first rave at Rennes Festival in 1995, where artists like Daft Punk and St Germain have played. That was my first time hearing 60 Hz of bass and a 10 kHz hi-hat at that volume, and I loved it. So these are cultural differences that matter.
“Warm” means low-end for the Western mind, like when people say, “I want to hear a warm reverb “. They probably mean something like a dark plate, rather than a bright spring reverb. It’s like comparing the “cold” sound of Pro Tools to the “warm ” sound of tape machines: one is dead and the other is alive, but not “dead” as in “lifeless”. “Dead” just means “precision”, where you get the most accurate representation possible. But if you already recorded to Pro Tools, you might not want any more accuracy, which is why we use summing boxes and tape machines to glue our mixes together. At other times you may have recorded through a console and tape machine, and so you don’t need anymore glue, in which case a Pro Tools mix works better. But again, your culture will determine which one you prefer.
– Let’s talk about engineering for a bit. In your experience, what’s the most challenging instrument to record?
If you can record a drum kit well, you can record anything. And when your drums start sound like those on a Led Zeppelin or Motown album, you’re already a competent engineer. Drums are the most difficult thing to record because they have everything: short and long sounds, dark and bright sounds, loud and quiet sounds, all of which are very unbalanced between each other. The kick and snare get played at different times with cymbals and toms popping in without warning, and you have to construct a specific sound out of all that with a limited number of mics. Also, a drum kit is more like a synth than a guitar. Guitars have a specific sound, created by the body, strings, and the pickup. Beyond that, you need a pedal-board to change the sound of a guitar. But a drum kit can change its sound through tuning, through dampeners on the kick or snare, and even by altering one’s playing style. So a drum is not just a “drum” – it’s very versatile, like a synth.
– When I talked to Philippe Weiss, he contrasted being a mixer versus being a recording engineer. He basically said the job of a recording engineer doesn’t compare to the importance of what a mixer does. What are your thoughts on that?
That’s a classic debate (laughs). But the answer depends on which era you’re talking about. In the 60s, a recording engineer was a technician who waited for orders from the producer. His job was to know the properties of the mics and the gear, and he wasn’t thinking so much about “art”, but rather on the coherence between music and the technology. Mixers are different – they have more responsibility because their work is about making the music tell a story, which transforms the final outcome in a very deciding way. So they focus more on art, which is more difficult than the technical things that anyone can learn, and so they get paid more. But what’s a recording engineer today? He’s the guy who does everything because he’s taken on the role of a producer. He chooses one track over another, decides when the track is done, picks the studio to work in, helps pick band members and instruments, etc. That used to be the job of the producer! Dedicated engineers don’t exist anymore, other than at big studios like Studios Ferber or La Seine. So the salary of an engineer could vary nowadays, depending on what he’s willing to do.
– You once said that 80% of what a record sounds like comes from the artist and the instrument set-up. But how can that be the case when the artist relies on the engineer to capture the recording? A bad recording goes nowhere, regardless of how good the musicians are.
Wait a minute, have you heard the drums on Stevie Wonder’s “Superstition“? If you delivered that drum sound to an artist today, you’d be fired as their engineer, no matter how much you tried to convince them it was cool. You can barely hear the kick at all in that song, and today’s EDM-driven industry wouldn’t respond well to that. It was possible in 1972 because the landscape was different, and the role of the kick is actually being played by the Minimoog anyway. But the drumming isn’t on the grid at all, and I can’t image how the horn players overdubbed to that. So Stevie and his producer were probably alone in the studio with no-one else around to mic the drums except an intern, because it doesn’t sound like a real engineer recorded that track. So Stevie’s drumming was technically terrible, but overall result was a masterpiece, and the record still became a hit.
I can give you another example: if you listen to the first version of The Beatles’ “A Day In The Life“, they started with percussion and an acoustic guitar that was out of tune, and it sounded horrible. But the drums and vocals sounded great, and the overall song works because the composition was excellent. Also, the heavy compression of the Fairchild 660 helped the drum sound. So these are two examples where the starting point of two classic records was a terrible sound, yet no-one cares. So for me, engineers don’t make artists, although great artists can make great engineers. Marten Hannette didn’t teach Peter Hook how to play bass for Joy Division or New Order. Peter invented that style himself and is almost alone in that kind of aesthetic. So the artist and his vision comes first.
– So who’s more valuable to a project? The artist who might lack a sense of direction, or the knowledgeable producer with tons of experience?
The artist of course.
I think the rehearsals he did with his vocal coach, Seth Riggs, sound like magic. If you achieve magic in a rehearsal but can’t recreate it on a record, then yes, the artist is questionable. But Michael Jackson was able to achieve the same magic on his records, and Quincy Jones helped elevate him by advising him to write his own songs. Have you listened to the rehearsal demos he did with the Latin percussionists before recording “Off the Wall“? That’s when he put together “Don’t Stop Til You Get Enough“, which was the second song he ever wrote. So I think you should give Michael more credit. But I agree with you that producers matter, and it’s a bad idea to go in a studio without one. And don’t ask the engineer to act as the producer because it takes away from his focus. We’re living in a time where everyone thinks they can produce themselves because they have a home studio, but that means knowing which direction to go artistically. When people book time at Melodium the day after making demo at home, I get tempted to say, “You might need two months to make a good song, not two days “.
– I get what you mean. Let’s talk about recording and gear. What would you say is the most important part of the signal chain?
I would say it’s the microphone because it starts off the chain. But the thing is, if you use a great mic you’re obliged to also use a suitable pre-amp. For example, an expensive ribbon mic with a low output and high noise-floor won’t sound good with a cheap pre-amp. You’re won’t do yourself any favors with that combination.
– But what about high-output mics? When I talked to the founder of Roswell Pro Audio, he said that high-output mics don’t require expensive pre-amps because you don’t have to crank the input gain as much.
Well, a high-output mic on drums would require only about 10 dB of input gain, so a cheap pre-amp would be fine for that. But one difference between cheap and expensive pre-amps is the noise level. Neve pre-amps are generally transparent from 20 dB – 65 dB, but a €50 pre-amp will probably produce noise after 40 dB. So using a Neve pre-amp on drums isn’t as glamorous as people think; it’s actually the EQs that makes the Neve channel-strip useful.
– The role of dynamic and condenser mics are well-known, but the benefits of ribbon mics are not as well explained. Can you talk about that?
Ribbon mics are considered to be natural-sounding, and there are a few reasons for that. One of them is that sound loses energy in the high frequencies first due to absorption by furniture, carpets and people. Many ribbon mics reflect that by having a roll-off after 7 kHz. Secondly, most of them have a figure-eight pattern, with two thirds of the recording being the source and one third being the room behind the mic. That helps to create an image in line with what humans hear because of how sound reflects around a space. So when you combine the high frequency roll-off with the room pick-up, you get a mic that sounds pretty natural.
– I’ve always been interested by how certain gear can glue the sounds in a recording together. What’s the easiest way to achieve that?
Glue between sounds happens when you cut transients. Tape machines can do that, but so can plugins. There’s one called SPL Transient Designer which does that well. Consoles can help too. Hitting the transformers hard in consoles like the Trident 65 or a Helios can achieve a type of compression that can get dirty but pleasant. Amy Winehouse recorded on a Trident 65 and they only cost around €4000. I used to have one, and so did Nigel Godrich; he also suggested to Air that they get one. It’s a nice little desk that cuts transients and glues things together in a way that a Neve won’t. A Neve 1073 can cut transients a bit but they’re generally transparent, and once you max out the 27 dB of headroom, it sounds horrible. But keep in mind that you have to ride the input gain of your console with precision if you want to glue things together, and 10% – 20% of the recording might end up too hot if you’re not careful.
Tape machines can also be used to glue things together, and the ones from the 80s do it more effectively than a console. It was different in the 70s though; you had to watch the VU needle and ride your fader properly if you want to hit the sweet spot of 3 dB above 0. But it became easier in the 80s when tape machines had less hiss.
Secondly, glue is about composition; if a song has too much silence, it sucks energy out of the music and requires effort to regain momentum. So “glue” also means avoiding the loss of energy, mainly in the rhythm section. For example, a short snare may need reverb to fill the space around it.
– Let’s say you’re going to mix a record in another country, and your client has two consoles: one has only 500-series EQs and other has only 500-series compressors. Which one would you use?
For mixing, I’d choose the console with the compressors because today’s trends demand that your record is loud. But my personal preference would be to chose the one with the EQs. However, if I knew the record would be mastered, then I’d definitely go with the EQs because it lets me shape the emotion behind the music with more versatility than a compressor.
Look, there’s hardly anything inside an SSL pre-amp except Integrated Circuits. As for API, 90% of their pre-amps consist of less than thirty parts that you can buy for €100 and construct yourself. Their 2520 op amp is nothing to talk about, and the rest of the components are just resistors, capacitors and a few transistors. The only thing that’s expensive is the transformer, which costs around €40. Overall, API and SSL pre-amps are nothing like a Neve 1073, which is four times more expensive and has a more complicated design.
– So then why is SSL held up as a relevant audio brand?
Simple. It’s not because of the sound of their consoles, but rather the routing capabilities and summing buss compressor. 40 years ago, SSL and AMS Neve chose to invest all their money into routing, and people in the 80s thought you were a good sound engineer if you worked on an SSL because of how complex the design was. Presumably if you could master a big desk like that, then your impact on the recording would be significant because of your technical skills. SSL and AMS Neve emphasized routing a signal as creatively as possible with inserts, groups and auxes. But the thing is, you lose sound quality each time your signal passes from one component to another unless the desk is built right. So the sound suffered and no-one wanted to admit it because of how exciting the consoles had become. As long as you could send your signal to an Eventide or Lexicon unit and get it back into the console, no-one cared about anything else. That kind of routing made mixing engineers very powerful, so none of them wanted to speak out against it. But if you put a signal through Neve 1073 and send it straight to your converter, it sounds better than an SSL 4000 that routes things like a labyrinth.
The best way to build a console is to use input and output transformers at every stage, which those companies never did. Only Neumann did that for the N20 desk, which hardly anyone knows about. I’m talking about transformers before and after each pre-amp, EQ, insert, buss, pan pot and fader. As a result, one channel strip alone weighs around 12 kg and the whole desk is four times as heavy as an SSL.
– Can you expand more on the role of transformers? How much do they affect things?
Most consoles already have input transformers, so the difference in effect depends mainly on the outputs. If you have transformers on both ins and outs, your console generally has a better signal-to-noise ratio and therefore better definition. Lundahl, Jensen, Cinemag and Haufe are considered to be in the upper range of transformers, whilst API ones were known to be fast and boosted transients in the hi-mids, which caused the sound to jump forward. Neve 1073s have the most transparent ones which are the closest to what you get from Pro Tools summing. I have some Neumann PV76s, which are pretty colored with an emphasis at 200 Hz. They sound slow and warm with a roll-off from 5 kHz onwards, which makes for easier listening.
– Why don’t people talk about Helios and Harrison the way they do Neve and SSL?
Probably because they sold less units. Also, they had less after-sale service than other brands. For example, AMS Neve knew they couldn’t provide maintenance for all their sold consoles in the 2000s, so they asked SHEP to build their modules and handle maintenance for the VR series. But people forget how good other consoles sounded. Lafont was a French audio designer who only sold consoles in the US, and if you ask older American engineers about it, they know a Lafont is comparable to a Neve. But here in France, Lafont is hardly known.
– Let’s move on to tape machines. You said earlier that you prioritize gear that works over gear that just looks nice. But the music of the 60s and 70s was made on “unreliable” tape machines, yet it sounded fine and I don’t hear many people that complain about the recording process.
Are you sure people didn’t complain? I have first-hand experience with using that gear as a former artist and I heard lots of complaints. Oui Oui had a budget that gave us access to nice studios for a week to record ten tracks, and then we’d go to cheaper places to do overdubs. The cheaper studio had different speakers, consoles and tape machines from the first one, and suddenly our music sounded different. The hi-hat that used to sound like a “chss”, now went “chaa”. I still don’t know if it was because of the new speakers or the tape machine, and that’s the issue with all-analog recording chains: the sound changes when the gear changes. Young people today are products of the Pro Tools era where the output signal of a DAW matches the input exactly. You can’t imagine how lost people were in the analog era when they’d take their tapes into a new studio and the sound became different. So the 70s had tons of problems, but the musicians made great music in spite of that due to great compositions, instruments, engineers and rooms.
– What were some other issues you had with tape machines back then?
Well, you had to hit them at the sweet spot of their gain stage if you wanted the best sound possible. If you went over the sweet spot, the sound distorted. If you were under it, you’d get too much noise. So it was a fine line where you had to constantly tweak the gain and use gates. You also couldn’t hit the compressor too hard because excessive gain reduction meant increased noise on a device like the 1176. So the settings had to be perfect if you wanted the best result. Setting a compressor in the 80s was a real job because it impacted the final noise levels, which is why some albums have lots of hiss. As someone educated on a four-track cassette recorder, hiss is something I wanted to avoid, although I notice that many people under 25 seem to have no problem with it today.
– Do you know which decade the highest quality tape machines were made in?
The best-sounding tape machines were made in the late 70s because the manufacturers were willing to invest in good line pre-amps. Similar to a console, high-quality pre-amps cost a lot, but tape machines in the 80s became like SSL desks: they substituted their quality pre-amps with improved VU meters and overall functionality. But a good tape machine is primarily about pre-amps followed by good mechanics because the start and end of the tape reel don’t have the same weight nor tension. So the motors have to be able to balance all the moving parts, otherwise the speed won’t be stable and the sound suffers.
– In one of our previous conversations, you said that running a mix to tape after having recorded and mixed entirely in a DAW is a weak move. Can you expand on that?
I don’t want to be dogmatic, but it just seems logical to me: if you like the sound of tape, why wait to use it at the end of your record? All the decisions you made during the recording and mixing stages were based on the sound of a DAW, so why change all of that at the end if you already like how it sounds? To me, it seems like people do that to reassure themselves that they’ve achieved a “vintage” sound just because they used a tape machine. If you like tape, why not use it at the start of your record, or at least use it on high-transient instruments like drums? I could even understand if someone with no budget used tape emulation plugins throughout the recording and ran his mix to a two-inch machine at the end. But to use no tape at all and then print to a two-inch machine at the end? That’s weak in my eyes; it’s just a way to reassure yourself that you’re cool. None of the decisions you made over however many weeks involved tape, but now you’re sure that printing your master to a two-inch will make things sound better? I’m not sure about that. But I know one thing for sure: it’ll definitely change your in-the-box mix to something different.
– But that’s like saying, “All the mix decisions you made didn’t take into account the mastering. So why bother mastering a record? Your mix should have been perfect on its own “.
But the best mastering in the world involves hardly making any changes to a track. The ideal situation is for your mix to be a masterpiece that achieves everything you want it to. Today’s mix engineers already know how to mix loud, which is why many mastering studios will eventually downsize or close, and the ones who remain will have the best-sounding rooms and good brains, not loads of gear. Guys like Chab are just that: a great room and sharp brains. Before meeting him, I assumed he had a long gear chain, but then he told me, “No, it’s all about converters “, and I was like, “What? “. But it’s true. He has a bunch of different-sounding converters that cost €7000 – €10,000 just for a two-track. So when Chab wants more treble on the master, he runs the track through a particular converter, rather than an EQ. So you need to choose the right gear for your ears and build the right room. That means having a heavily isolated and treated space where simply shuffling your feet creates a disturbance. I couldn’t even stand in Chab’s room with my shoes on because it kept bugging him when took a small step forward (laughs).
– Our conversation thus far has me wondering something: if you have all this expertise in Classic Recording Techniques, plus a vintage studio of your own, why don’t I see any records on your résumé that are on the same level as the classics from the 60s and 70s?
That’s a fair question, and I’ve asked myself many times why I built a place like Melodium when most of today’s artists prefer the convenience of modern tools. But just so you know, about 3% of the projects I’ve done at Melodium did achieve a sound that was on par with the 60s and 70s. But the reason for the low percentage is simple: virtually no-one wants that sound anymore. They say they do, but in reality most artists want a mix of the past and the present, not just the past. So when my clients tell me they want a “vintage” sound, I tell them, “Sure, we can do that if you really want it, but you might end up retreading someone else’s sound from the past. Is that really what you want? “. If you look at music history, the first artist to sell millions of albums by imitating a sound from the past was Lenny Kravitz, with “Let Love Rule“. It was first time the industry took notice and wondered how an artist could draw so much inspiration from the 60s yet still have good songs and commercial success. But I don’t remember anyone saying, “Hey Kravitz, stop doing that! That’s old music and we don’t want it! “. Actually, most people said it was cool. I was able to achieve something similar with a French band called The Cavalier. They told me they wanted a 60s sound, so I put up two Melodium 42Bs as room mics and they achieved 50% of the mix. The guitar player had a Fender Reverb Unit, the bassist had a Dan Electro ’64 with flat-wound strings and the drummer used a kit tuned to sound like one from 1964. So they ended up sounding like a ’64 surf-rock band. I have about five other projects where we achieved that kind of sound, and to be totally honest, it’s not hard. There’s no magic involved: if you give me the right composition, lyrics, instruments and musicians, we can get it done. But like I said, most artists tell me that they want a vintage sound plus something new, so I put a lot of attention on balancing the two.
(Below: Nicolas Dufournet)
– Let me ask some last questions about the studio. Tell me about the staff you work with at Melodium.
Felix Beguin and Scott Bricklin are two of my partners that have been here almost since the beginning. They’re both musicians and sound engineers, and Felix was an intern in my previous studio at the go-kart basement. Scott’s an amazing musician who sings and plays drums, bass, guitar and synths – and he does it well. He plays bass as well as he plays guitars and drums, which is rare. When you have a guy like that who can cut a track in three hours by playing multiple instruments, but also understands what the artist and producer wants, you’re lucky to have him on your side.
I also have other engineers that I trust to handle sessions because I trained them myself, like Maxime Kosinetz who is doing very well right now. I earn less money when I hire them, but I know the sessions will be done well, and it’s also a way to stay connected with good engineers that have their own clients.
I currently have a technician who works a lot at the studio, whose name is Vincent Henon, and we have an agreement where I give him studio time in exchange for repairing my synths. He was recently able to accomplish what two of the best techs in France couldn’t do; they couldn’t figure out what was wrong with my Jupiter 4, but Vincent fixed it after twelve hours of work. The Jupiter 4 is a nightmare for techs because it was created at a point when adding digital components like RAM to analog synths became an industrial effort, which created all kinds of problems. So I’m happy Vincent was able to fix it.
I also have a community manager, Robin Delattre, who does the content creation for my newsletter.
Even though they’re not my staff, I’d like to mention Yann Arnaud and Adrien Durand; They’re producers who played an important part of the studio’s history. Bertrand Burgalat is another friend who trusted me right from the start and gave me a lot of responsibility.
– And what do you charge for clients who wish to record here?
I have two types of prices at Melodium. There’s one for self-produced indie artists, which is €425 for the main studio and myself as engineer, although I may lower that price if the project is simple or my bookings are low. The second price is for record labels; for an indie label I charge 10% more, and for a major label it’s 20% more.
– It’s been a long talk, but let’s end the interview by analyzing some 60s and 70s music. I’d like you to explain what you hear in the following tracks. First up is “Hihache” by the Lafayette Afro Rock Band, released in 1973.
Drums: The drums are heavily compressed and not well-muted; I hear lots of harmonics in the snare and it rings outs even when the kick hits, so the engineer didn’t care much about the drum tuning. Maybe they didn’t hear it whilst recording, but once they added the heavy compression, the ringing becomes audible. The kick sounds like many others from that era: it has no subs or highs, but lots of low-mids. The batter head was probably muted with a pillow that deadens the sound a few milliseconds after the kick hits. But since you can hear the kick sustain, they probably left the resonant head unmuted. Another trick for capturing more sustain is using a figure-eight mic inside the kick drum. That way you’ll get two-thirds of the beater and one third of the room and resonant skin.
The drum tuning is higher than what you’d hear nowadays. Bass and high-end in music have increased in the last twenty years, but most people didn’t strive for that in the 70s, so the tuning was different. You can tell that this record was made in the early 70s because of the drum sound. A drum kit is the most demanding instrument to record, so the more console channels you have, the more of them get used for the drums. That’s why drum sounds change so noticeably from decade to decade.
The toms sound more muted than the snare, and aren’t as compressed. So there was probably no tom mic, which means you’re hearing the toms in the overheads. But the snare is clearly compressed because it rings out, so there was a close mic on that. The cowbell was probably done as an overdub, maybe with a ribbon mic. If you want make your mix easier, pay attention to how you record your transients. Ribbon mics are useful for harsher sounds, whilst condensers work well on softer ones.
Bass guitar: Neither a cabinet nor a microphone are 100% consistent in their sound, so the bass was probably recorded through an amp because I can hear distortion and inconsistencies in sound. The mic is also close to the speaker because I can’t hear any of the room. But the bass also has a strange sound, as if they used a phaser pedal on it with a slow rate. It’s mostly mid-rangy, and the instrument is probably a P-Bass played with fingers; you can’t get that kind of sound with a pick. It also sounds like roundwood strings were used.
Keys: I hear distortion on the Wurlitzer, so it was probably recorded mono through an amp.
Guitar: That sounds close mic’d through an amp with a little drive. Or it could have been done with a small 10 Watt amp that was cranked.
Brass: The brass is roomy and bright, so it sounds like an ensemble of three or four people playing in a 30 m² room, about three meters from the mic. It also sounds like they added a lot of reverb, and I think they used EQ to boost the treble because the ensemble sounds bright even though the mic is pretty far away, which wouldn’t make sense. But the brass solo sounds like it was recorded with a dynamic mic at close distance.
– Next is “H.K. Theme” by Hansson & Karlsson, released in 1967.
Organ: The organ probably has two close mics that are 30 cm – 50 cm away from the leslie, with two room mics about two meters away. Leslies from the late 60s distorted easily because of their tubes, but I don’t hear any distortion here so the organ probably wasn’t too loud. So the big sound organ comes from combining the close mics and room mics.
It sounds like he’s playing the Hammond P-100. People seem to think the only good Hammond organs were the B3s. Hardly anyone knows they made at least 50 other models, many of which were good too. The P-100 was one of the smallest and sounded similar to the B3, but only cost around €500. But the flip side is that you’ll be spend hundreds on maintenance. I know because I used to have one (laughs).
Drums: The drum kit has at least one overhead, with another mic near the snare and hi-hat because you can hear the definition on them. There’s probably no mic on the kick for the opposite reason: it lacks definition. Kick mics weren’t common until the late 60s anyway.
Double bass: The double bass has no definition nor treble, which points to a lack of close micing. I find that odd because it’s obvious that you’d need a close mic on a double bass that’s six times quieter than a drum set. So maybe the double bass mic was placed too close to the drums and picked up too much bleed from them. That would explain why they couldn’t push up the volume of the double bass in the mix, and maybe had to cut the treble because of excess cymbal bleed.
Master buss: To make the recording upfront they would have used compression. Maybe they used it on the master, but it could have also been done on the room mics, which adds a lot of energy by enlarging all the instruments in the room. It sounds like they recorded in a good live room because things sounds balanced, whereas bad-sounding live rooms have audibly asymmetrical frequencies.
– Let’s look at Manzel’s “Midnight Theme”, released in 1979. What can you tell me about that?
Drums: The drums are very processed and you can hear that it was deliberately done. They may have used an outboard gate to cut the sustain between the hits. You can achieve a similar effect by dampening your drum kit, but the gaps in sound are too sudden for that in this case.
The snare has a lot of harmonics, so they may have used one with a steel-body. It’s tuned relatively high, since higher tuning means more harmonics. The snare and hat are also close mic’d because you can clearly hear their definition.
The kick and snare sound heavily gated and compressed, but the transients aren’t being boosted so I’m guessing they used a fast attack with a slow release. A slow release would normally boost the room sound, which is absent here, so I think they recorded with overheads and no room mics.
The hi-hat is loud in the mix, but it’s not harsh at all, so maybe it was recorded with a ribbon mic. It sounds very smooth because of how gently the drummer played; he doesn’t hit it loudly at all. Drummers nowadays are overly focused on following the metronome, so they smash the hi-hat to reassure themselves that they’re on-grid. I see that a lot of that in my studio and it’s a major mistake. The hat is just a link between the kick and snare, so you don’t need to overemphasize it. Also, it’s pointless to give your drummer a click-track unless they’re relaxed. Otherwise they listen more to the click than the music itself, which leads to a drum performance that doesn’t complement the rest of the instruments.
Bass: The bass is reminiscent of James Jamerson’s, where you have a lot going on but it still sounds simple. It’s very controlled, with no unnecessary notes. It sounds like the musicians on this track made very concrete decisions about how they wanted to record, which is reminiscent of Motown. Bass is the hardest frequency to isolate, and you need a lot of foam to do it well. That’s why Motown didn’t use bass amps and built a special DI box that gave their basses strong transients and presence. But the bass on this track sounds like it was amped.
Glue: A lot of the glue on this track is due to the wonderful musicians, especially the drummer and bassist. Like I said before, when a band plays together in the same room, you get bleed in all the mics and that creates glue between the sounds. In the 50s and 60s, everyone played together with few overdubs, and you had to achieve the most you could with four to eight tracks and a live band. So labels like Motown would only call the best musicians because the recording had to start all over again if someone made a mistake. And by the way, being the “best” just meant you played music five hours a day or more.
– The next one is Blue Öyster Cult’s “(Don’t Fear) The Reaper”, released in 1976.
Guitars: I hear stereo guitars that are double tracked. They might have also been chorused together, along with some gated reverb.
Drums: The drums sound ahead of their time for something made in 1976. The hi-hat sounds very thin, which means it had a close mic that was EQ’d, with reverb added after. The snare probably has a wooden body with a damper on the head. It also sounds compressed and gated to make the sound jump forward, all of which means it was close mic’d. This kind of production was typical of big studios that had the privilege of many mics and channels-strips, so it was probably done on a 24-track console.
I sounds like the drums, bass and maybe one guitar were tracked live and the rest was overdubbed. The feeling I get when all the instruments are recorded in the same room is absent here. I don’t hear much bleed and things sound a bit isolated from each other. But they probably thought the approach was cutting-edge in 1976. It was a time when engineers felt they had power to do unconventional things, like sending one cymbal to a delay, putting reverb on the toms and gating the snare.
– We’ll end things off with Gil Scott Heron’s “We Almost Lost Detroit”released in 1977.
Reverb: It’s clear that the artist wanted to achieve an intimate sound on this record, so it’s less about effects and more about the performance. But you can hear a small two-second reverb on the snare that glues it to the vocal, which also has a long plate reverb on it.
Vocals: The vocal was close-mic’d and compressed so it doesn’t move around, though you’d normally ride it before compressing to avoid hearing the compressor working, which makes it sound smooth. Also, the Moog and guitar aren’t panned, which would distract from the vocal.
Making this kind of music requires doing a lot of mall things right, rather than make big moves. Complex things that look effortless is the result of many small, gentle choices.