Producer and songwriter Fraser T Smith has made the unlikely journey from prog rock to grime. With Tinchy Stryder, his broad experience has paid off in the shape of three massive hit singles.
With American artists still dominating the hip‑hop/R&B genre, it's gratifying to see artists from different nationalities beating them at their own game. East Londoner Tinchy Stryder is one case in point. He's known as a grime vocalist, though the subtle distinctions between grime, UK garage, breakbeat, hip‑hop and other sub‑ and sub‑sub varieties of the urban genre will be lost on the masses. To the uninitiated, Stryder (birthname Kwasi Dunquah: he's of Ghanaian descent), simply makes urban‑sounding pop. The 22‑year old rapper's rise to fame this year has been dramatic, with three successive major UK hits in 'Take Me Back' and chart‑toppers 'Number One' and 'Never Leave You'. All three hits can be found on his second solo album, and first major label release, Catch 22, released last August.
A central role in the making of the album was played by Fraser T Smith, who co‑wrote, arranged, played, mixed and produced many of the songs, including the three above‑mentioned hits. Smith is currently making big waves in the UK urban scene, winning the 2008 Urban Music Award, but urban music was not always Smith's metier: in fact, he was previously best known as a prog‑rocker who played guitar with Rick Wakeman, most notably on 1999's Return To The Centre Of The Earth. For Fraser Thorneycroft‑Smith, as he was then known, this experience was as much a milestone as a millstone, which needed a few years, and a name abbreviation, to live down.
The now 38‑year‑old Fraser T Smith recalls how he took the long and winding road from old‑hippy prog rock to trendy urban youth music. "My first point of entry in the urban music scene was a remix I did in the mid‑'90s of Drizabone's 'Let It Out'. After playing with Rick I went on to work with Craig David — from 1999 to 2004 — and that really made my name in the urban genre. I played guitar for him and we produced one track together on his first album, Born To Do It , which sold millions. I also did bits and pieces in the studio, writing, co‑producing, and so on, on his albums Slicker Than The Average  and Trust Me . In addition I co‑wrote the track 'Broken Strings', by James Morrison and Nelly Furtado. But really, I'm not from an urban background, I'm a jack of all trades. So I was very pleasantly surprised to receive the 2008 Urban Music Award. I thought Mark Ronson was a shoe‑in.”
'Jack of all trades' is a modest description for someone who can play, write, arrange, program, engineer, mix and produce. Smith spent his teenage years studying the guitar, attended a guitar college for a year, and entered the London session scene at age 19. Encounters with the recording side of things sparked a fire, which went on to burn more and more brightly.
"Even when I was playing in bands I was always recording, and learning the craft of production and mixing. I grew up with a four‑track and a drum machine, and mixed things down on C60 cassette tape. When I did studio sessions as a guitarist I asked engineers endless questions, and after I did a few tours I realised that I loved the studio so much that I quit the whole session guitarist thing, and concentrated on writing and production. I had my first studio in a room at the Roundhouse in London, with a Logic setup and some analogue bits and pieces. The essence of what I work on has always been a drum machine, a guitar, a keyboard, and some way of recording it. That's still the same today.”
In Smith's MyAudioTonic Studio, where all the Tinchy Stryder material he was involved in was written, recorded and mixed, the drum machine is an Akai MPC4000, the keyboard is a Yamaha CX2, and the recording is done in Pro Tools. His workspace holds API 512, Chandler TG2 and Avalon U5 preamps, Neve 1073, API 550 and Manley Massive Passive EQs, Dbx 160, Urei 1176, Empirical Labs Distressor and Alan Smart C2 compressors, a Crane Song HEDD valve/tape emulator, Focusrite ISA430 and Avalon 737 voice channels, and Lexicon PCM70 and MPX1, Eventide H3000, Roland SD330 and Lexicon MPX1 effects processors. There's an Apogee Big Ben clock and a 32‑channel Chandler summing unit, and the monitors are Auratones, Yamaha NS10s and Genelec 8240As: the last are tuned to the room and Smith calls them "brilliant”. The musical instruments lying around include a Roland Juno 106 and SH101, a Korg Polysix and a Yamaha DX7. Perhaps surprisingly, he also owns an Otari MTR90 MkII 24‑track analogue tape recorder.
"The philosophy of my studio is to be able to do everything here: writing, recording, mixing and production,” says Smith. "I therefore also have a live room, with piano, drums, and all my guitars. In the past I had a TL Audio desk, which was fantastic, and I love to be able to spread things out on a desk and I love twiddling knobs. It's a great feeling to sit in front of a large desk, but apart from the romantic image, I also remember sitting at 3am behind an SSL and the recall is not quite right, and the desk is not catching up with Pro Tools, and so on. Having a desk here does not work for me in any case, because for my business to work, I need to have a turnaround of quite a few tracks per year — the music industry is a volume economy these days. And on any one day I may be working on five, six, seven, eight different projects, so I need to have instant recall. If a desk came out that had perfect instant recall, that was almost like a DAW controller, and yet had the sound of an SSL, I might think again. I know that manufacturers are getting close to that.
"But my setup here at the moment really works for me. I get the sound that I want by doing everything in Pro Tools, and then using my outboard and the Chandler, which sounds great and gives me a sense of analogue warmth. So I'm not fully in the box, as my stuff always passes through analogue, whether it's the Manley, or Alan Smart C2, or the Chandler. I got the last because of a blind test on an American web site [www.vintageking.com] that had sound files comparing different summing boxes and Pro Tools fully in the box. I love Pro Tools, but if you can hear the difference in a blind test five times in a row, you have to do something about it. There was something in the stereo image and the width, and in the top and low end that made the Chandler files sound better.”
"With regards to analogue, having the Otari 24‑track is quite unusual these days. I think it's always good to have something that goes against the grain, and that can hopefully make things stand out a little bit, because everybody is using more or less the same gear these days. I don't record directly into the Otari, but use it more like an effect. So I record things in Pro Tools, then run low‑end tracks — anything below 500‑1000 Hz — out to the Otari, via the heads, and back into Pro Tools. I also sometimes send strings to the tape recorder, just to take a little bit off the edge. But while I get a warmer sound back when I send higher‑frequency stuff to the Otari, I then have to take a lot of that warmth out again with EQ to make it cut through in the track. I end up sort of chasing my tail with sounds above 1k, so I don't do that any more. The Crane Song HEDD also has a tape saturation feature on it, and all my mixes get printed through the Crane Song. Add all my analogue outboard, and ultimately, I have a lot of options for imparting an analogue feel to my work.
"As for the synths, the amazing thing, in the current climate in which everybody is buying virtual synthesizers, is that I can buy hardware synthesizers really cheaply! I recently bought my DX7 for £99! I am having a field day buying all these old modules and synths for next to nothing. Even if I get just five sounds from each module, it is worth it. In my view, their sound quality is so much better than of the VST Instruments, and you get loads of knobs to twiddle and patches to play around with. I also often put these synths through outboard gear, to strengthen the analogue effect. For instance, I put the Juno through the Avalon U5 preamp into an 1176, kind of like a bass guitar, and that sounds great. I also sometimes use the API preamps. My drums from the MPC4000 usually go through the Neve 1073, and then I compress them.
"There is not much going on in the box in terms of MIDI stuff. 'Number One' was all beats from the MPC, synth sounds from the Roland SH101, Juno 106 and Korg Polysix, and only the sampled strings were in the box, from [Digidesign] Structure and Access Virus. In addition, 'Never Leave You' had live drums, piano and strings. The live drums were recorded via API preamps, with a very standard mic setup: [AKG] D112 on the kick, [Shure] 57 on the snare, [Sennheiser] 421 on the tom‑toms, [Neumann] TLM170 on the overheads. I bounced the live drums down to a stereo pair, so they felt more like a stereo loop; I did not want to have a song for Tinchy that was all live drums, because his demographic is still very much into hearing beats. But I thought it would be nice to have something a little bit more organic underneath, and after 'Number One', to have a track that sounded a little bit more mainstream. I always like doing tracks that combine synthesizers and beats with more organic elements.”
- Written by K Danquah, Fraser T Smith, D Contostavlos
- Produced by Fraser T Smith
Fraser T Smith: "Writing 'Number One' was quite a convoluted process, because of everybody else's schedule, but we wrote the basic material in half a day, then spent another day finishing it off, and then the guys left and I spent two or three days on the production, and finally a day on the mix. Everything was done here at my studio. I came up with the rough beat on my MPC4000, which is still the basis of my whole MIDI and beat setup. I'll work in the MIDI domain until the beats and the synth parts are right, and then I'll record everything as audio in Pro Tools. Tinchy and Dappy [from N‑Dubz, who guests on the track] wrote the lyrics, and we wrote the hook together. My engineer, Beatriz Artola, and I recorded both of them using the signal chain Soundelux E250 microphone, Avalon 727 preamp, Urei 1176 compressor, Crane Song HEDD, Pro Tools.
"While I was arranging the track and playing the parts, I was sending stuff to the record company and we were going backwards and forwards with it. After we agreed on a direction, I put the icing on the cake, arrangement‑wise, and mixed the track. Like everyone these days, I mix as I go along, but I have made a decision to deliberately define the point at which I start mixing. I can always add in more or change things while mixing, but I think it's helpful for my headspace to define the mix stage as such. I'll clear the decks, clean up the Session, consolidate all the tracks, and look at the Session purely from a mix point of view. I would say that I get the track up to 65 to 70 percent by the time I hit the actual mixing stage. Sometimes I'll hire in extra outboard gear for the mixing stage, like on 'Number One' I had a Manley Vari‑Mu compressor and a Neve 33609. One other thing I'll do before I begin a mix is to send various low‑end tracks out to the Otari tape machine. In the case of 'Number One' these were the kick, snare, bass and low pads. I'll print the returns in the session.
"I typically take about seven to eight hours to mix, and will start my mix with getting the drums rocking, adding in the bass, and then the vocals. I'll fit the rest of the elements in with these elements. The rhythm section and the vocals are the most important, so I prefer that the other elements find their way round them, rather than the other way round. After that it's obviously a lot of tweaking things, going backwards and forwards between different elements, working on dynamics and balance, adding more effects, taking out frequencies that are not necessary, tidying up loose ends, working on the automation, and so on. As you'll see on the screenshots, I have my Session well organised, with drums at the top, then bass, then synths and keys and vocals, and then my effects and VCAs and the master mix at the bottom. I find that because of the amount of recalls that I have to do, having a system in place that makes it possible for me to go back and run off a quick mix in five minutes is essential. I also have all the waveform images in the Edit Window running from the beginning to the end of the Session, because to me it feels like the Session is set in stone, and nothing is going to be moved any more. It's a visual/psychological thing. I also retain the track numbers, so I know how many I have, and if I get to the high 60s, I know I need to bounce some things. The 'Number One' session is 47 tracks, which is a medium‑sized mix for me.”
- Drums: Waves Renaissance EQ, Compressor, Reverb, Maxx Bass and C4, Lexicon MPX1, Digirack Delay, Neve 1073, Dbx 160a, McDSP CB2, Tech 21 Sansamp, Alan Smart C2, Crane Song HEDD.
Kick (tracks 1‑6): "I had six tracks just for the kick in the 'Number One' Session! It's a dance crossover track, and it also has to come through loud and proud on the radio, which it does! I ran them all via the tape machine, at 30ips, hitting them as hard as I could. Track 3, 'Big Kick In', is my main kick. Track 1, 'Hi Kick tape', is a higher‑sounding kick that's more percussive, and track 2 is a kick part that only occurs in verse two. I copied that from track 1 and it has lots of effects on it — I often duplicate a track if I want to effect it another way. On tracks 4 and 5 I have 'sub' tones that are triggered by the main kick on 3, using a gate, so they have exactly the same pattern. Track 4 has a sub tone of 31Hz and 5 one of 63Hz, and I printed them. It's the purest tone that you can add, and they're purely there to reinforce the low end. Track 6 is an auxiliary track to which I sent the three main kick parts, tracks 1‑3.
"In terms of effects, on the main kick drum (3), I have a Renaissance EQ, boosting 5.8dB at 4555Hz, a Renaissance Compressor, and a 'Tiled Room' reverb [in Waves' Renaissance Reverb], which is an emulation of a preset reverb from the PCM70. I don't like the sound of reverb on drums in this kind of music, but if it's completely dry it can sound a bit demo‑ish, so the 'Tiled Room' gives a bit of space. It's like a bathroom reverb, and I'm not sending too much — the send is at ‑25. Track 3 also has 'Kickcompot', which routes it to an aux, where it triggers the McDSP compressor on the main bass track (14) as a side‑chain, so every time the kick hits it ducks the bass track 2‑3dB to give space for the kick. I do a similar thing with hi‑hats and some of the synths. In addition to giving space, it also gives the music a bit of a pulse.
"On track 1, I just have the 'Tiled Room', but on track 2 I have a whole bunch of effects: Renaissance EQ, adding low end and around 5k, a quarter‑note delay and an eighth‑note delay using Digidesign's Long Delay plug‑in, my outboard Lexicon MPX1 which was set to a three‑second reverb, plus a 1.5‑second Renaissance plate reverb. In verse two you only hear track 2 and the voice, hence all these effects. Finally, the kick aux track (6), has a Maxxbass setting that I copied from Dave Pensado, I found it in Sound on Sound [Jan 2007], adding some harmonics around 120Hz. There's a Waves C4, dipping 3dB around 125Hz, to give space. In turn, I also dipped the bass around 60Hz to give room for the kick, it's like complementary EQ. The kick aux track is also sent to my outboard Neve 1073 and Dbx 160A. I have marked that and the settings at the bottom of the track in the mix window. If I were to write my outboard settings in a book, I would lose them. For me it's much easier to have all these details in the session.”
Snare (7‑10): "Yes, four tracks of snare! I didn't use [track] 8 in the end, it was purely triggering a gated reverb, which was nudged slightly forward in time. The main snare track is 10, which has Renaissance EQ, the same plate reverb as on track 2, and the 'Tiled Room'. The 'mono' and 'stereo' snare tracks [7, 9] went to the tape recorder, and the former has the 'Tiled Room', and the latter my Lexicon MPX1, with the same effect as on the kick on track 2. Track 9 is the bigger‑sounding snare track. I used the Waves PAZ Analyzer on tracks 7 and 8 to see what was happening frequency‑wise.”
808 hat, FX, 'crush' (11‑13): "The 808 track was an 808 sample in my MPC. I put the McDSP CB2 compressor on it, which is also picked up by the side‑chain from the kick, with 3dB compression, to make the hat groove a little bit more. The Distortion comes from a Sansamp plugin, just to thicken up the sound, and there's some Renaissance EQ and 'Tiled Room'. The FX Bounce track has some cymbal splashes that I used some Renaissance EQ on, to lift the top end. Finally, track 13, Drums Crush, is all the drums together, sent out as a group to my Alan Smart C2 compressor and smashed to pieces: 10:1 ratio, really fast attack of 0.1, 12dB compression. I had that sitting underneath the main drum group. I don't like the sound of heavy compression on individual drum tracks, because I think it makes them sound smaller, but I do like to have some really heavy parallel compression underneath the sound. In that way you get the best of both worlds: the dynamics and openness of the uncompressed sound and the heavy compression underneath. By the way, the 'AES Input' is the signal that comes from the Crane Song HEDD, with a little bit of tape saturation.”
- Synths: API 550b, McDSP CB2 & Filterbank, Waves C4 & Renaissance EQ, Empirical Labs Distressor, Sonnox Oxford Transient Modulator, Lexicon MPX1, Neve 33609, Eventide H3000, Digirack Delay.
"The two bass tracks, 14‑15, came from the Juno 106. They are duplicates. The main bass track, 14, was bounced to tape, and then sent to my API 550b EQ with boosts at 300Hz, 1k and 5k, and compressed with the Renaissance Compressor. As described above, the McDSP CB2 is triggered by the side‑chain from the kick, and the Waves C4 ducks at 63Hz to make room for the kick. I then sent track 14 to a Distressor, set to 10:1, Dist 3, and that comes up on track 15. The two tracks of 'Punchy synth' (16‑17) are a stereo pair, with the same effects on both tracks. Plug‑ins are Renaissance EQ and the Sonnox Oxford Transient Modulator, which increases the transient value to add some punch. The two tracks also get sent to my Lexicon MPX1 and the Neve 33609 compressor. The 33609 went on all the synths, by the way.
'Low Indigo Pink' [track 18] was a low synth part that I sent to the Otari. 'Big Filter Pad' [19‑21] was a part that I played on the Juno 106, which went to the MPX1 and the Eventide H3000 on a micro‑pitch‑shift setting, ±16 cents, panned left and right, to make it wider. The Pad tracks also have four‑band Renaissance EQ and McDSP CB2. Track 21 is a duplicate of 20, EQ'ed slightly differently in the verse with a McDSP Filterbank plug‑in. 'Long Filter Notes'  is a high synth part from the 106, and 'Tinchy 1'  is a synth sound that I played with a quarter‑note delay using the Digidesign Extra Long Delay, on which I put an EQ, rolling off all the low frequencies and pushing up at 1.5k.”
- Strings: Sonnox Oxford EQ, Lexicon MPX1, Eventide H3000, Waves SSL Compressor, Sound Toys Filter Freak, Manley Vari‑Mu.
"I tried to get the strings to sound as real as I could, so I took a lot of time with each sound in the way I played and treated it, and worked with the dynamics. Most of the string tracks have the same effects, just in different amounts. 'Oxford 5BnEQ' is the Oxford five‑band EQ, 'MPX1' is my Lexicon hall, 'H3000' is the same effect as on the 'Big Filter Pad', 'PAZ' is again the Waves Analyzer. Track 29 is an aux track to which all the strings are sent. It's slightly compressed with the Waves SSL Compressor, and also has a bit of H3000 and MPX1, plus the Sound Toys Filter Freak, with an effect that modulates high and low filtering — it occurs in the middle eight where everything drops out and it's just Tinchy and the strings. I also sent track 29 to the Manley Vari‑Mu compressor, which sounds great on strings and really warms them up. I hit it with about 3dB of compression.”
- Vocals: Urei 1176, Sonnox Oxford EQ & Limiter, Waves C4, De‑esser, Renaissance Compressor & EQ, L1, Digirack Delay, Tech 21 Sansamp, Lexicon MPX1.
Tinchy (30‑33): "Track 30 is Tinchy's main vocal, which goes out to my 1176, hit quite hard at 8:1. It also goes through the Oxford five‑band EQ, high‑passed, taking out some low‑mids and boosting some high‑mids. Then there's the C4 with everything bypassed, apart from a slight cut around 1k. I really like the C4 on rap vocals, because it's dynamic, and when rappers get louder certain frequencies in their voices become more apparent. When Tinchy's voice opens up you can hear a lot more 1k, which I dipped out with the C4, with the threshold set so that it doesn't do that when he's singing more quietly. There's also a Waves De‑esser, which is cutting 3dB at 9k. The Renaissance Compressor is levelling things out before the signal goes to the Urei, which is set to a ratio of only 1.5:1, compressing by about 2dB. The Digidesign eighth‑note delay is in fact a 3/16th‑note delay. The Sansamp crunches the sound of his voice a bit.
"I applied a lot of volume automation, which I often do on vocals, to make sure you can hear every single word. If I want half‑note or longer delays I usually copy the track to an adjacent track and just delay the actual audio. I did this with the 'Radio Delay' track . I added a telephone effect on the Renaissance EQ, high‑passing and pushing 1k and then compressing with the Renaissance Compressor. Distortion is again the Sansamp. Tinchy's overdubs [32‑33] have similar de‑essing to the main vocal. With the EQ, I'm not pushing the high‑mids as much, and instead adding low‑mids, so it does not clash too much with the main vocal. The rest is a little bit of delay, and widening, and a plate to give the vocal some space.”
Dappy (34‑48): "Track 34 is Dappy's lead vocal, with an effect chain that's similar to Tinchy's vocal. I felt that it would be good if the vocals sounded like they came from the same place. So it's sent out to the 1176 as well. I took out everything under 100Hz, dipped at 1.5k, added some Renaissance EQ taking out around 500Hz, and added Waves De‑esser and Renaissance Compressor, just compressing a couple of dB at a 1.5:1 ratio. The 'Radio Delay' track  was set up in a similar way as with Tinchy. I was constantly struggling to make the vocals sound bigger, so I duplicated track 34 twice to tracks 36 and 37, panned them left and right, added similar effects, and sent them to a bus. All Dappy's backing vocals [38‑47] have their own aux track . They're each treated a bit differently, with a little bit of MPX1 hall, quarter‑note and eighth‑note delays, and so on. On track 40 I used the Oxford Limiter, and on 42‑43 there's the Sansamp, crunching the vocals. I applied quite a bit of L1 on track 48, to bring the vocals out slightly. In general you can compress vocals and backing vocals quite hard without artifacts becoming noticeable. You want to hear every word of the vocals, and when you have a pumping track, you sometimes need to compress the vocals hard to make sure they sit above what's going on.”
"When you get to the end of a mix, you want to have control over larger sections of the mix at the same time. For this reason I set up two VCA tracks, 49 for all the vocals and 50 for all the music tracks. VCA mixing is fantastic, because it allows me to bring all the vocals up at the same time and balance that against the music — it's like having a VCA fader on an SSL. Tracks 51‑58 are effect tracks, where you can see many of the effects we discussed; 59 is my reference track, on which I can bring up other songs for comparison and reference; and 60 is my final mix. The whole mix was sent though 32 channels of my Chandler, then summed, and the Chander's stereo outputs went through my Manley EQ, and then the Alan Smart C2 compressor, which is not really doing a lot — the attack is 30ms, ratio 1.5:1, and a couple of dB compression — and then to the Crane Song HEDD, which adds a little bit of tape saturation. I then printed this back in the session. The session and mix were at 96k. I normally work at 44.1, but in this case I wanted to experiment with 96 to see how it sounded. If you get into large track counts, 96 tracks becomes a problem, but it worked really well in this case, and sounded great.”
Originally a musician, Fraser T Smith is self‑taught as an engineer and mixer. Fewer and fewer engineers are coming through the ranks in large studios, and there's widespread concern about the demise of a studio culture in which a great deal of know‑how is passed on. Smith recognises the issue, which is one reason why he is employing the services of engineer Beatriz Artola. He explains that he has acquired his technical knowledge from many different sources.
"In some ways, I wish I had come through the studio ranks, because it'd be great to have incredible knowledge of microphone techniques and so on. But my role is that of producer, writer, musician, arranger and mixer, and for me it's most important to have a very all‑round knowledge of the whole writing and production process. I've been learning about engineering and mixing as I go along, and one great source of knowledge for me has been the mix series in Sound On Sound. I've read every single one of them, and I've literally studied each in depth. The articles have been a huge help in my learning process, in terms of mixing and production. Normally you don't know what goes on in the studio, and you may hear something on the radio and you think, 'wow, that's amazing,' and you wonder how it was done. Lo and behold, the next month I'm reading how it was mixed, which is like absolute gold dust to me.” He said it, not us.