AH: Did you get your start recording classical music?
JK: No, just rock and pop mostly. Mostly punk rock and lots of jazz. I was in a band. I'm of this new breed of engineers: they were calling me for keyboard and drum programming. The Yamaha DX7 was my first, and it brought me the most jobs. No keyboard player knew how to change even the release time — you know, with six operators and the explanation deep in the manual. It was like, "Can you make the release on these strings longer?" And drum machines — we had at first the Roland TR-808, and then the Yamaha RX11. These were my first jobs. Then I just started playing around with the tape machine.
How did you get involved with the Filharmonija?
This one is the best orchestra in the country. I got the job because the National Radio usually records this session and there are 24 mics in the orchestra. You can mix and stuff, but the sound is just not the sound of an orchestra. It just gets muddied with too many mics — phasing problems. You have a lot of control — you can lift up cellos, whatever — but the sound is not the right one.
Is that a 63-piece orchestra you're working with today?
Yes, I believe so. With five percussion players for this piece.
I noticed that the space is fairly dry sounding.
Too short and too small for this size. This should be a 20 to 25-piece orchestra for this space.
Will you add artificial reverb to make it sound bigger?
Just longer. Not louder. Because the reverb itself is pretty strong — but short — like wooof wooof. Like one big kitchen. I'll be adding a tail — a couple of seconds long and very long pre-delay — not to change the perception of the hall, but just to make the breaks longer so they are not so dry.
How do you work with the fact that there's a short pre-delay already but not a lot of diffusion coming from the hall picked up by your mics, and then you also add additional reverb? It's like two different spaces.
We're happy that brains are not perfect and can be fooled easily. Especially if you come from the stance of physics — soundfields and theory — there are so many things we're doing wrong completely. But they work.
JB: Do you use a convolution reverb?
Last time I used a Lexicon 480 and a Quantec [QRS]. The Quantec is unbelievable. Also, there are two beautiful EMT 140 plates at Metro with vacuum tubes. In the end they will be really, really quiet — just slide them under.
Will you also master it at Metro?
I'll do all the mastering at Metro. It's the best-equipped studio in Ljubljana.
Do they have a separate mastering room?
No, in the same room.
It's just one process?
Yeah. It's not much mixing because it's stereo mics.
So you just add reverb? What about compression and limiting and EQ?
I do all parallel processing, like adding top end but not sending the whole signal through. Just record the EQs back to the workstation and add these tracks in slightly. At Metro it's mainly Manley Massive Passive. You just slam it — 10 kHz on up. Then a Summit Audio [EQP200A], which has a little bit less distortion than the Manley, and slam this, also like 10 kHz on up. And Night Technologies [EQ3D], which is the most precise. Then I mix together in parallel all three and record a mix of all three EQs back to the workstation. Then I add this layer to the original, so the original is untouched. The same with compressors — multiband or just mix compression — just to lift up the quiet parts.
Do you find that you hear any phasing from adding the Massive Passive to the original signal in parallel?
No, but you have to take care to use bell curves and not shelving. Shelving can become a bit funny with phasing. And filters — low-cut, high-cut — they're not really phase-linear. But bell curves are perfect. The Summit is also perfect. Completely phase-linear. Just high-cut gets a bit weird — some phase anomalies.
Is this something you do only with classical music?
No.
Do you tend to process everything you do in parallel? You always have the original signal untouched, and you add to it the processed signals?
Yes. For top end I don't just use one EQ, I use three. For full-range compression I use three different compressors in parallel.
You never put them in series?
No, not really. Sometimes multiband and full-range after it. It's always a question. Sometimes when you set a compressor on slow release, it's like wooof wooof [slowly]. It's cool. But also on fast release, it's like oomph oomph oom [quickly]. Also cool. But if there are no in-between settings you can use two compressors: one long release, one short release. That's the advantage. Metro is great, but in the end, when I'm doing this mastering and I process top end separately and midrange, like 200 cycles, separately, I end up with seven layers. To do this in real-time, I'd need something like eight Summit EQs and six Night Technologies EQs and five Massive Passives. But I only have one piece of each. It's fine economy-wise. You may only have one EQ and one compressor, but you use them ten times, over and over, each time recording back to the workstation.
What do you sum it all with?
Just in Nuendo. Digital summing.
You prefer doing your mixing and mastering in Nuendo, but you're recording today with Pro Tools.
Yeah, it's just convenience. It's "house" Pro Tools. Doesn't make me bring a different computer. They're all basically the same — it's just a question of which mindset.
JB: Have you done any experimentation with Altiverb or other sampling reverbs?
A bit, but not on this material. We sampled these plates at Metro and it's nearly there. But when you put up the real thing it's this slight difference. You can tell.
Are there a lot of people that come from outside of Slovenia to record here? Is the cost here lower or higher?
Much lower. Especially from a country like Norway. For example, Chris Eckman did a session in Oslo with strings. But it cost him more and the session was terrible, so they erased everything. They came here, and instead of a five-piece they used a twelve-piece orchestra. With studio time, flying and hotel it was cheaper than at Oslo. And much better results.
Do you think music is that much more part of the culture here, and that's why there are better musicians, or it's cheaper? Culturally, what's the difference?
Especially musicians are cheaper. That's the main advantage. Having a session in London, I mean you can get some students. They are rehearsing and stuff. But here, for the same price you can get top-level musicians. There aren't many, but some are very good. It's also the influence of the Balkans — you know, timing is different. With The Walkabouts, we did this André Heller song, and it was in 7/8. You know, Dah duh Dah duh Dah duh Dah Dah duh Dah duh Dah duh Dah [very quickly]. The Walkabouts' American keyboard player [Glenn Slater], he just couldn't get it. You know, where to get in. I called the clarinet guy, who is from here, and he just came in, and — prima vista — he just nailed it. You know, Wum bum bummmm... Wum bum. I think that sense of time changes most with culture.
Is this place known as a destination for music production?
We do have some tradition. Out of all the rest of Yugoslavia, Ljubljana is the center of Slovenia, the most western country. Years ago here it was like Italy — much more open. It influenced the technological tradition. Lots of big stars from Yugoslavia were recording in Ljubljana. It was the place to be.
Big stars in all forms of music?
Yeah. Especially pop and also rock. In Ljubljana was the first multitrack studio in the Balkans — more than 30 years ago. The owner died. He was like a real legend, because he was building tape machines by himself. [A collective "whoa" comes from the Tape Op crew] A really funny studio. I worked there for a month, but the owner always had some problems drinking and so on.
JB: What was the name of the studio?
Studio Akademik.
30 years ago, was it hard to purchase equipment?
Too expensive. He built a mixer by himself. There was no patchbay — even no plugs. Everything was soldered to the mixing desk — even the repro head, because he built the tape machine himself. You save a lot of money if you don't use proper Cannon connectors. Solder everything straight to the desk — the best sound for sure. It was impossible to overdub in the control room, because the mixing desk and everything was wired directly to the repro head. We always listened from the repro head. You never listened to what was going into the tape machine — always what was coming off the tape. There was a small mixer — separate — wired to the sync head and to the headphones wired in the studio. [laughter from Tape Op crew] Since you listened with like 200 ms latency, for punch-ins you had to really get a grip before... [more laughter]
There was no way to bring back a headphone feed from the studio to the control room?
Oh, we didn't mess up his studio. He was a funny guy. He had a whole rack of Audio Design Compex limiters — expander, limiter, compressor, whatever. A whole dozen or something. But only after a week of working did I realize that only the limiter was working. Only one was complete. He built the rest of them himself — the whole front plate and everything. But nothing was working but the limiter in all the machines because the limiter was the only good component from the first machine.
After coffee Janez returned to his orchestra recording, with a promise to continue our conversation at Studio Metro later in the day. We ran back to the hotel to grab our bags and head to the project studio of Gregor Zemlji?, a multi-genre producer, a mastering engineer, and a lecturer at SAE Institute (through Croatian National Radio & Television). Gregor has a modest room just across from the Ljubljana train station, packed full of gear and treated with DIY broadband absorbers. ATC SCM150ASL midfield monitors strike an imposing stance in front of the room. At the desk is an SPL MixDream analog summing system.
Tell us about your signal chain.
I use a lot of analog EQs and compressors. I don't use a lot of plug-ins, except for some basic stuff — notch filtering, side-chain compression and similar. When everything works in Logic, I group everything to stems, and I do the final stage of mixing on MixDream, which makes a big difference, especially for dance-related genres. It's a big change in dynamics. The majority of this kind of music goes to vinyl, and for vinyl you need a bigger dynamic range but a much more controlled dynamic range if you want to get punch. When I first mixed with an SPL MixDream, the difference was really subtle. But, when I started seriously working with MixDream, it was completely different; I heard music more precisely, and I started mixing differently. I have 16 channels of DA connected directly to MixDream , and on the first four stereo stems I can insert analog EQ followed by a compressor.
What stems do you typically mix to?
I always combine kick and bass in one stem. Then I group all the loud drum stuff and all the loops. Next stem is composed from all the drive synth sounds and other drive elements like guitar. It is usually followed by vocals. These four stems are processed with analog outboard. The other eight channels are not processed with analog during the final mixdown because I don't have that much analog yet.
Do you process them with plug-ins?
Sometimes. I usually process tracks beforehand. I go out of the box and use various combinations of EQs and compressors in the rack. I have a collection of vintage pedals, Space Echo delays, outboard and effects. I have an EMT 250 reverb downstairs. Stuff like that. I used to have a mixing desk here, but few years ago I decided, that it is impractical for my way of work. My idea was to build a modular mixer with sophisticated, open patch bay to which all outboard devices are connected and all routing, automation and sends are controlled from Logic. I have 16 inputs and 24 outputs. So, for each project I can "build" a unique mixing environment.
What was the first project you completed working this way?
Magnifico [Export-Import], which was released on Sony BMG worldwide. It was quite successful.
Is that electronica?
It's kind of ethno-electronica. The genre is called "Balkan Beat". They gave me really good arrangements and really good songwriting to work on, but they had weak sounds and drum programming. With the equipment I had it was impossible to do the complicated mix I wanted to do. I processed each sound through an analog chain, added the reverbs and delays that it needed, and bounced it back to the workstation. In the end I had a flat mix in Nuendo with no plug-ins. I did the final summing and automation in Nuendo.
How in your mind do you think of the big picture as you're working on one track, two tracks, or maybe even six tracks at a time, as you're processing each track or set of tracks separately.
When I started doing it, it was like, "Okay I'm doing a big experiment." We didn't have a mixing desk anymore and this was the only way to do it. I just went along, and because the album had dance elements I knew if I would get it right with the bass spectrum — kick drum, bass, and the loops — then everything should work. This is the area I had the most experience. If I did anything wrong in the process I just went back and bounced again. But it came out great and I didn't have to go back many times. The biggest advantage of a system like this is that you can do live mixing, live dubbing, live automation. When I was working on dub delay effects, I could do a take tweaking the feedback and every other parameter available and record it. If I was not satisfied with the result I just repeated the procedure. It sounded much more alive and better sounding at the end than if I had a static mix with just an open send. [Gregor plays some tracks from the album]
I noticed that most of the drive elements are right there in the center, but you managed to have the vocal closer in soundstage. What did you do to process that depth?
The whole magic in his vocals I suppose is an old Neumann KM 56 I had. It was through a Neve, completely dry. It just worked. For the depth I used really short delay times, which were then sent to the Lexicon PCM60 plate preset. I don't use reverbs a lot, I use short delays instead, but they're all unique. For example, RE-201, RE-301 and RE-501 are completely different Space Echo delays. If you want a dirty sound, you use a 201. But in a clean mix like this it's kind of useless because it's noisy and has an overdrive that comes out too much, so I might use a 501, which is much cleaner. On this whole album it's all Boss DM-300, an old analog, bucket-brigade delay. Everything I used is a combination of DM-300, PCM60, and Mu-Tron Bi-Phase. I put a lot of stuff through the Bi-Phase with LFO off so it doesn't sweep, but it stays in a fixed position. You can get a really warm color. A lot of reverbs I use, like spring reverbs and the PCM60 (still my favorite reverb — it's so simple, it's got a few presets and you can do a lot of stuff), I just usually put through the Bi-Phase and do a really, really mild phasing, so it kind of evolves over time. But on this album I think it's the mic. On the new album we tried many, many mics, but the only one that approaches the right color is the Brauner Valvet — but at the same time it's too precise for his vocal. But this is the commercial part of the stuff I do as a producer or mixer or co-arranger. Usually I do much more experimental electronic music. My main project, Random Logic, is basically the first thing that started my whole career. I started as an underground electronic musician, and we were releasing a lot of 12'' vinyl, even in the United States — about 40 12'' EPs under the Random Logic name. We sold 70,000 vinyl.
You're a musician and a composer, but you were also engineering and producing. Can we hear some Random Logic?
This is really experimental. This is the new album and all sequencing is made without a computer. Everything was done with analog sequencers: Doepfer, Roland, Korg, and so on. No presets are used. We used the computer just as a tape machine and an edit machine. The idea is to play the song — we want a live sound. When we compose we set up a kind of big modular system with all the synths and we just create one big patch, like a kick drum, hi-hat and snare drum patch. Then we make a groove and we jam and record. We tweak relevant parameters for the complete period of the song. Then we reset everything and we do the bass and lead, and we program another live jam on top of what we have already. This is similar to the original approach to electronic music, like Kraftwerk and so on. They recorded on a tape machine. If you make a mistake you can't recall anything. It's much more alive. [We listen to Random Logic and hear lots of interesting textures on top of strong rhythms and beats]
