Recording Morcheeba's 'Rome Wasn't Built In A Day'

Pete Norris

Published in SOS October 2000
Printer-friendly version

People : Artists/Engineers/Producers/Programmers

Morcheeba's last album Big Calm was a commercial and artistic breakthrough, and expectations were running high for their new album, Fragments Of Freedom. The first single from the album, 'Rome Wasn't Built In A Day', was an intriguing taster. Producer Pete Norris told Tom Flint how it was written and recorded.

By integrating blues, rap, trip-hop, dub, pop and many other styles, Morcheeba have defied musical pigeonholing ever since the 12-inch release of 'Trigger Hippie' first caught the public's attention way back in November 1995. Their first album Who Can You Trust? was a slick trip-hop record, similar in feel to the work of bands like Massive Attack and Portishead. The 1998 follow-up album Big Calm displayed many of the same musical influences as Who Can You Trust?, but was far broader in its scope. And their third album, Fragments of Freedom, released this July, has moved the musical goalposts yet again.

Morcheeba's continuing development has not endeared them to certain sections of the music press: "We've been slated in the press for our change of direction, that is absolutely nothing new," explains the band's long-term producer, programmer and studio engineer Pete Norris. "The only thing that annoys me about it is that now the journos are pretending they liked Big Calm. Well they didn't like Big Calm when that came out, and now they're saying 'Fragments is not nearly as good as their triumphant album Big Calm.' You buggers!

"When a band stops moving, it's over. Who Can You Trust? is suicidal, Big Calm is a little bit happier but still basically miserable. This one is cheerful. It's a progression because we all move on as people and we make records the way we feel. We feel happy."

The now-smiling Morcheeba, comprising brothers Paul and Ross Godfrey and singer Skye Edwards, begun work on Fragments Of Freedom in their Clapham studio during July 1999. Like many of the album's other tracks, the gospel-influenced 'Rome Wasn't Built In A Day' -- the first single from the album -- started life as a four-track demo put together by multi-instrumentalist Ross. Pete explains the importance of that early demo. "Ross played himself four minutes of drums onto a regular recording Walkman. He then put that tape into his four-track to work from. By the time he'd finished his demo, the vocal line and verses were already there. When we get to the recording stage we know we've got the song -- that part is done and dusted. Then we start embellishing that."

The song's basic composition having been established, a period of routining began, with the band's drummer Martin Carling playing the kit, string arranger Steve Bentley-Klein supplying guide keyboards, Ross roughing out guitar parts and Skye providing the basis of the vocal idea. A very temporary guide bass part was also included. As with the four-track demo, the jamming-to-tape method was used in preference to sequencing or digital audio editing, as a faster way of working out the structure and flow of the song. "A lot of things you think are going to work really well, usually the more complicated things, fall flat on their faces when you get a bunch of people playing. We've generally taken that as a reasonable yardstick."

Once the band were satisfied with the ideas developed through the jam sessions, recording of the drums commenced. Pete: "We try to keep it as simple as possible, so the first thing I do is pull off the kit anything which isn't going to be hit -- which, on this track, leaves kick drum, snare drum, some hi-hats and a cymbal. Stripping the kit to its basics makes it sound much better. For example, I removed the toms, which are otherwise fixed on to the bass drum, so there's no unnecessary ringing or rattling. It also makes it a lot easier to mic up, because things aren't getting in your way."

The kit was recorded using an Electrovoice RE20 for kick, Shure SM57 on snare and a pair of AKG C3000s for overheads with the bass roll-off switched in. The preamp section of a TL Audio C1 preamp/compressor was used along with a Valvetronics Gain Ryder 3 compressor for recording the kick and snare onto the studio's Otari MTR90 24-track machine. The overheads were recorded through the preamps in the studio's Mackie 32:8:2 desk and then into the Otari. "We try to get the fattest sound that we can. Since we got the 2-inch Otari machine, that's a lot easier. The sound of rock drums, for me, is the opening bars of 'Highway To Hell' by AC/DC, which is essentially over-recording onto 2-inch tape. In fact, everything's jammed onto two-inch analogue tape, which solves so many of your sonic problems at a stroke without you having to do anything. I don't usually bother compressing drums on recording, I just make sure I get a more than healthy level on tape. It's that in-your-face fat sound -- not horrible like in the '80s when every record sounded 20 miles away!"

Once several drum takes were recorded to 2-inch, Pete and Paul began choosing the best sections of the performances and sampling them into their Akai S3000 sampler. Editing of the sampled sections was done using Steinberg's Recycle running on an Apple Mac. "We were not going to use the whole drum track live; it would be lovely but it's too uncontrollable. It's not a question of quantising Martin's playing, which is really tight -- we use it more as a way to get our other machinery things to sit with his playing. If he feels it slightly faster or slower we'd rather have that reflected -- the chorus speeds up slightly towards the end of the song, for example."

Though Recycle on the Mac is used to chop up Morcheeba's drum loops, and a 32-track Pro Tools system is used to compile, edit and mix their tracks, the chopped-up drums sampled into the S3000 are still sequenced from an Atari running Cubase 3.1, as Pete explains: "All the MIDI stuff goes through the Atari which, to be honest, boils down to the MIDI parts of the drum track. We take the Atari out of sync with the Mac, which means Paul can fiddle with the sequencer while I'm moving audio around in Pro Tools. It's the Mac that runs the Recycle program, but I put the MIDI files it generates onto a floppy disk which I pass to Paul an

Delays, Reverbs And Real Room Sounds     Effects on Morcheeba records are used sparingly. Pete explains the philosophy behind their application. "It isn't effects that make records sound good, it's having a great song, a great arrangement and a standard of recording that allows you to hear those elements. Beyond that, I use two delay lines and a reverb, and most of that is used on vocals, keyboards and strings. It doesn't affect the rhythm track particularly and I don't find myself putting huge amounts of anything on Ross's guitars. We use a lot of the room sound; its various nooks and crannies produce different sounds which are always more convincing than getting it out the box. Having said that, we acquired a Lexicon 200 which we use for one of the plate settings and that gives a gorgeous sound. We use that and delay on vocals. Delay has a much more profound effect on how much space is around something than reverb does. On 'Rome' there's just a short delay with a handful of repeats."

d he loads into Cubase. The only trouble is that you need a supply of double-density disks and they have to be pre-formatted -- that's what lets you take it out of a Mac drive and put it into an Atari drive and have it recognise it. Then you have the Atari running the sampler."

With drums complete, the process of replacing the guide tracks and adding new overdubbed parts began. Pete explained the typical recording chain used for most overdubs: "For any single overdub, which includes vocals, I use an Amek 9098 mic preamp and an Empirical Labs Distressor compressor. I have the mic plugged into the mic pre section of the Amek, the signal out of the mic pre section goes into the Distressor, then out of there back into the Amek's equaliser section so I've got the compressor before the equaliser. I prefer compressing things before I EQ them. This setup also handles any DIs."

A week after the guide tracks were recorded, the bass part was replaced by Steve Gordon using a Warwick bass, recorded through the DI output from a valve head amp. Several tracks of bass were jammed through the 9098 and Distressor chain onto the Otari. From the Otari, the bass tracks were then transferred into Pro Tools for editing. "We sit down with pads of paper, looking for Mr Good Bar. We're looking for something which grooves with the drums to build the bed of the rhythm with. We take that and move on," explains Pete. On 'Rome', however, the bass part was later replaced by Steve towards the end of the recording to fit more comfortably with some of the other overdubs.

To further cement the basic rhythm track, an acoustic guitar was added to establish the basis of the melody. "It's a simple line played by Ross, using an old Ovation Balladeer which records nicely miked up. I used to use condenser mics on guitars, but the Ovation is quite bright and bass-light, so condenser mics tend to make it sound brittle. That was miked with the RE20 without bass roll-off, to compensate for the brightness."

After the guitar, a pad line was introduced to the composition from Morcheeba's one and only synthesizer module -- a Roland MKS80 Super Jupiter -- recorded as MIDI in Cubase. Pete: "Ross played simple triads. You don't want it to get too big because then you don't have any space for anything else. That line fills out the space at the back of the track, you're not really supposed to be consciously aware of it. It just does what pads have done for decades."

'Rome' has four recorded electric guitar tracks providing three different sounds, described by Pete as "a low one, high one and solo one." All were played by Ross using a Mosrite guitar, a Watkins amplifier and a variety of processing gear. "There's no power amp out on Ross's amp, so he gets banished to the live area with headphones. The fuzz line went through a Lovetone Big Cheese pedal and the filter section of our EMS Suitcase synth. The output of that went into the miked up guitar amp."

Strings were also used on the track, albeit in a rather subtle and simple arrangement. A violin, played by string arranger Steve Bentley-Klein, was tracked several times. The line comes in during the chorus under the vocals. "It's just another thing to lift the chorus and make it a little bit brighter," explains Pete. "Once again, the string lines were recorded using an AKG C3000, through the Amek 9098 and Distressor onto 2-inch tape."

Vintage keyboards played an important role in the creation of the sound of both 'Rome' and Fragments Of Freedom as a whole. "On the instrument front we pretty much hate using sampled anything," admits Pete. "We've got a Wurlitzer, Clavinet, Rhodes, a lovely Hammond organ and an upright piano, so we're not in the market for sample libraries. A lot of the time we plug the keyboard into a little Musicman amp and then run the speaker output down into the booth where there's a 4 x 12 Marshall bass cabinet with a mic in front of it. Then the Musicman is next to the player, the player's next to you and the noisy bit of it is in isolation. It's nice having people in the control room when they're playing."

One of the most distinctive parts on the single is the funky wah-wah sound which on first inspection sounds like a guitar line. In fact the effect was created by Ross using a Hohner D6 Clavinet played through a wah pedal. Two further keyboard parts were played by Dan Goldman using a Fender Rhodes electric piano and Hammond C3 organ. "Dan has a Rhodes with the big silver speaker underneath, so he did his duty in the live area, with the cab miked in stereo using two Shure Beta 56s. You start off with a fantastically wide sound that's lovely in solo, but you can't make it fit the track, so we narrowed the stereo tremolo. When you're placing stuff in the stereo field, something that's all-encompassing is not a huge amount of use. The Rhodes was mixed off to one side with something complementing it on the other, like the acoustic guitar.

"The Hammond C3 organ is layered with the Jupiter pad to complement that sound. The Jupiter is just creating a little bit of movement, but they're both fulfilling the pad function. The Hammond is stereo by virtue of the fact that it's got a Leslie. The top of the Leslie has two mics on either side, so when Dan flicks the switch you get that sensation of it speeding up and slowing down and whizzing round your head. I don't make the bass end stereo, that would create other problems in the mix because you already have a bass."

The song's brass parts were played by London session men Steve Sidwell (trumpet) and Chris White (tenor and baritone saxophone). "We use what, as far as I can tell, is the standard way of explaining brass

Pro Tools Track Listing     Setting the right DAT recording level: SOS January 1995. Noise and how to avoid it: SOS May 1995. A Concise Guide to Compression & Limiting: SOS April 1996. The Mysteries of Metering: SOS May 1996.Minimising Mixer and Effects Noise: SOS July 1996. 1 Beat L 2 Beat R 3 Ride 4 Shaker 5 Tambo 6 Bass 7 AcGtr 1 8 AcGtr 2 9 Lo MosGtr [Mosrite guitar] 10 Hi MosGtr 1 11 Hi MosGtr 2 12 Solo MosGtr 13 Hi String 14 Clav 15 Rhodes L 16 Rhodes R 17 Jupe Pad 18 Hammond Lo 19 Hammond L 20 Hammond R 21 Brass Comp L 22 Brass Comp R 23 End Hi Sax 24 OutBv [Outro backing vocal] 1 25 OutBv 2 26 OutBv3 27 ChBv [Chorus backing vocal] 2A 28 ChBv 2B 29 ChBv 2C 30 ChBv 1A 31 ChBv 1B 32 ChBv 1C 33 VsBv [Verse backing vocal] 1 34 VsBv 2 35 VsBv 3 36 LVox [Lead vocal] Comp 37 Libs Comp

parts, where it's us going 'Da da da daaaa' and they go 'We know what you mean'. We build that on the 2-inch machine as always. The brass in the chorus was built in harmonies and bounced together. Effectively I'm taking one saxophone and one trumpet part at a time. We continue overdubbing until we get the fatness and texture that it needs. The ridiculously loud bit of sax at the end we put in for a laugh. We asked Chris to do some mental saxophone as the fade comes, so he blasted at the end. I don't think wanted us to put that in -- he was offering to do it again."

The final parts to be tracked were the backing vocals and Skye's lead vocal. Regular collaborator Derek Green and his vocal group, comprising Paul Jason Fredericks, Joy Rose and Dee Lewis provided the backing. "We tried out harmonies to start with, then Derek assigned and distributed them to the four vocal members. I used only one C3000 mic. I tried a pair but they didn't produce the sound I was expecting and it didn't fit the track. They were grouped around the mic in the middle of the room, then we moved them either closer or further away from the mic as required. If you try to record them individually you're not going to get that gospel feel. If you get a bunch of singers all in a room together they'll enjoy themselves and you'll get a better performance."

Pete explains how the different chorus parts were arranged. "What makes things big or small is the nature of the musical arrangement. Generally, tracks start small and get larger. The first chorus is very regimented, because you're trying to establish the tune. Our job as producers is to make sure we deliver that message as clearly and succinctly as we can. After you've heard it once, we broaden it out. In the second you have the really high vocal and Derek doing the low vocal. Those parts are above and below the parts in the first. In the outro we cut them loose -- we said 'Imagine you're in a tin church in the southern states of America, it's dead hot, it's Sunday, no-one has to go to work and you're just having a good sing.' By the outro it really started to fly so we dropped Skye's vocal for the first of the BVs, and that's where they start stretching out and singing little loops of their own. You hear their ad libbing because you're not focusing on the lead voice, and that lets the gospel vibe shine through. Then Skye comes back in.

"For Skye's vocal I also used a C3000. They work so well on her voice. There isn't a de facto best microphone, there's just what sounds best on the thing you're trying to record. I've made a pop shield out of old tights wrapped around a loop of stiff quarter-of-an-inch thick earth wire, so there are two membranes on each side of the wire. I place that two to three inches away from the mic so Skye can get as close to the pop shield as she likes.

"The rest of the chain is the 9098 and Distressor. I've always used that combination on Skye's voice and I can get the sound I want very quickly. I don't normally squash lead vocals too hard on recording, I use a little bit to even it out so Skye feels a bit more comfortable pushing it without having to draw away from the mic. I don't want her to do that, because the tone changes.

"She sings half a dozen takes. Around three or four takes she's peaking and that's usually where you get all your good stuff. It only looks like a complete track because I've bounced it down."

Mixing of 'Rome' began in February 2000, once all the album tracks were ready. Pete outlines the approach: "We try to sort the mix out while recording. If you can get a decent sound to start with, you'll enjoy the rest of the process because you won't be having to fiddle endlessly to make it sit. If there is something you're questioning yourself about, you should be making the decision then, because at that point the issues are fresh in your mind. You don't want to be in that situation where you have an awful lot of decisions to make at the end. With ours, if you push the faders up, all the sounds and arrangements are there. There isn't muting going on because we've already cut the unwanted parts of tracks."

Although the key to the Morcheeba mix is the quality of the initial recording, there is a place for some processing at the mixing stage. "One plug-in we use is Digidesign's Lo-fi. It will grunge things up. If you go too far it gets very distorted. As far as I can tell, it seems to be adding even-order harmonic distortion in the low end of the scale, which is like valve and class A gear. It produces that magic effect of being warm and bright at the same time. We used it on the Clavinet and on the drum beat. Other things are just small amounts of EQ to get the other instruments sitting tightly together. We're not trying to fit disparate elements or trying to make radical changes to the sounds, all we're doing is chamfering the edge of one part so it butts up against another.

"Our mixes take two days. Everyone takes home a mix on Minidisc to see whether we've left anything out or to see if we have a revelation with the arrangement. You can end up with a situation where you're expecting listeners to take hold of two ideas at once and that doesn't usually work. If we want something to be heard we won't clutter it up with other things going on at the same time. For example, there's no Clavinet when you have the guitar solo. After the first day, we snip those bits out.

"I make vocals as loud as I possibly can. To me, that is pop music. If you hear an early Beatles record in a café, all you can hear is vocals, maybe a little bit of guitar and the top end of the snare. So whenever a lead vocal is going on it's the boss, and everything else is subservient to the delivery of the song and singing.

"The final mix is on the hard disk at 24-bit as a Pro Tools Sound Designer II file. To master the album,

Choosing Pro Tools     In their last interview with Sound On Sound in the December 1997 issue, Paul and Pete made no bones about their hatred of digital recording and preference for analogue tape. So it came as something of a surprise that they are now working mainly in Pro Tools. "In the first Sound On Sound interview we were slagging hard disk recording to the nines, but although we've 'succumbed to the dark side' we can defend ourselves. We had listened to Pro Tools 3 because we were quite aware that we were sampling addicts and it would make so much sense, but we couldn't get on with the sound of the 16-bit systems. On any individual thing it's OK, but when you start mixing them all together you really start noticing the limited bit resolution, and it sounds like you've put everything in the sampler. It wasn't until I heard a 24-bit system that I thought 'Maybe the rules have changed a little bit now.' Thirty-two voices is good, plus the fact that if you get into a tight spot they're dynamically allocated, so if something's not playing in the middle you can use that voice for something else."

we hired in an Ampex half-inch tape machine. Ampex half-inch sounds lovely and takes that digital edge off it. There are several level-maximising plug-ins which are fantastic for demos, because there's no way you will get anything approaching the audio level of a commercial CD without that. I don't use mix compression, though, because we have our records mastered by a chap called Mike Marsh of the Exchange -- I don't want to burn any of his bridges for him. Recently, mastering engineers have been complaining that people are giving them stuff that's been absolutely flattened to a dynamic range of a couple of dBs. We presented him with sonically the best thing that we could, and that was the 24-bit output of the Pro Tools onto half-inch tape."

As Pete's story, and the full Pro Tools track listing (see box, left) demonstrate, 'Rome' ended up becoming far more involved than a typical Morcheeba song. Were all these elements really necessary, or had they simply become carried away by the process? Pete: "We got what we intended. It wasn't ever another type of song. There's way more stuff on 'Rome' than we normally do, but it was required to execute the song properly, although I don't think any of us could have predicted Skye would have turned into Aretha Franklin at the end! Sometimes these things just come out."