You are here

Jay Graydon

Musician & Producer By Dave Lockwood
Published December 2001

JAY GRAYDON

The 'producers' producer', the 'guitarists' guitarist': Jay Graydon is back, with a new album and a renewed hunger for success. Always the consummate studio craftsman, when he talks technique, it's worth listening...

The sheer volume and diversity of this man's work defies belief. Over the last two decades, Jay Graydon has been connected with more hit records than I could list in an entire page of type set at this size! And when I say 'connected with', I don't mean a guitar overdub that got left out of the final mix and a gratuitous 'thanks' in the sleeve notes. A Jay Graydon credit often reads: producer, arranger, engineer, programmer, guitar, synthesizer... and sometimes vocals too. Oh, and did I mention that he will usually have written the song as well?

The California‑born, double Grammy winner's discography reads like a Who's Who of an entire sector of the US music scene: Steely Dan, Al Jarreau, Earth, Wind & Fire, George Benson, Kenny Rogers, Donna Summer, Dionne Warwick... I could go on, but you've probably got the picture. Or at least you've got a picture, for it is perhaps a distinction of Graydon's career that in many cases the only thing that ties together the artists he has worked with is him. Possessed of a highly distinctive style as a guitar player, his production work is more diverse. If there is one word to sum up what a Graydon production brings to a project, I think it would have to be musicality: irrespective of the style, and no matter whether the tracks are driven by programming or 'sessionsville's finest', the grooves will groove, the chord structure will be harmonically interesting, and the end result will be, somehow, just... right.

One of Graydon's many hits as a songwriter was Earth, Wind & Fire's 'After The Love Has Gone'.One of Graydon's many hits as a songwriter was Earth, Wind & Fire's 'After The Love Has Gone'.For many, their first taste of Jay Graydon came courtesy of his striking guitar solo on the track 'Peg' from eclectic art‑rock/jazz duo Steely Dan's 1977 Aja album. By the time of his acclaimed 'fusion‑turned‑up‑to‑11' dual‑guitar tour de force on Manhattan Transfer's 1979 hit single 'Twilight Zone' (produced and co‑written by Graydon), he was firmly established as the first‑call West Coast 'hot' soloist, and poised to become one of a handful of producers who would define the sound of the decade to come.

That year he and long‑term collaborator, keyboard player David Foster released a successful album of their own, under the name Airplay, and Graydon received the first of his two Grammys — a Best R&B Song award for 'After The Love Has Gone', co‑written with David Foster and Bill Champlin and a worldwide hit for Earth, Wind & Fire. In 1982, he was to receive another Grammy in the same category for the George Benson hit 'Turn Your Love Around', which he co‑wrote with Steve Lukather and Bill Champlin.

High Precision

JAY GRAYDON

For many, Graydon's work with crossover/jazz vocalist Al Jarreau on 1983's Jarreau and 1984's High Crime (left) still represents the pinnacle of the genre, but whilst some hear a tight, polished production with ultimate attention to detail, for others, adjectives like 'sterile' and 'over‑produced' come more readily to mind. Even now, however, when recorded performances are almost routinely dissected and reassembled using hard disk recorders, Graydon remains sensitive to the importance of capturing a real performance on record. "I have been accused of using the microscope over the years," he admits, "but this was in 'performance‑land' and not using Pro Tools. I might have been tough on guys regarding the tracking date and overdubs, but it was still a performance. Yes, there were flaws, but they were kept if the flaw 'felt good'.

Al Jarreau's 1983 hit album Jarreau was recorded and mixed at Jay Graydon's Garden Rake Studios.Al Jarreau's 1983 hit album Jarreau was recorded and mixed at Jay Graydon's Garden Rake Studios.

"When cutting tracks with a real band using Pro Tools and the like, the producer might think 'We will just make a few passes, and fix or tighten up later, and maybe edit takes together.' This is supposed to be a time‑saver, but it's not really in the long run. The time spent lining up kick and bass hits, snare hits to the click or anything along those lines, is more of a time‑waster than recording more track passes looking for that magical take. Furthermore, each edit fix you do causes you to need other fixes elsewhere."

Much of Graydon's '80s work is defined by its successful integration of programmed parts and real performances to produce musically convincing recordings. "I think a great song and great vocal performance is the key to this concept," says Graydon. "Machines can feel good so long as the samplers and synths are MIDI timing‑friendly and the programming is done by someone that understands how to get a good feel. I have fought MIDI‑land for years and the key is to get rid of as many data interrupts as possible. After the sequenced parts are ready to record as audio, I turn off everything that looks at the internal clock such as MIDI Thru, all filters, and so on. On the samplers and synths, again, everything gets turned off that causes interrupts.

Among Jay Graydon's favourite synths are his battered Minimoog and his equally well‑used Oberheim Xpanders.Among Jay Graydon's favourite synths are his battered Minimoog and his equally well‑used Oberheim Xpanders."I always record the sequencer stuff in one pass and I also make many passes so I can choose the best. There is always 'MIDI slop' and data interrupts so things will actually feel different from pass to pass. Locking up sequenced stuff to timecode and overdubbing to it will never feel good because the sequencer/synth/sampler data interrupts will never line up the same every time. We are dealing with machines and they are not listening to the other stuff recorded but simply clocking in 'slop‑land'. Also, when using sounds with oscillator detuning, such as a Minimoog bass, frequency cancellation will occur from time to time, so if the take has serious cancellation, I stop recording and make another pass."

There are no trade secrets in Graydon's basic working method, however: like most of us who work with programmed parts, he maps out the rhythmic and harmonic structure first. "I start with the basic drum groove. I then add the main keyboard part, typically a piano part. Next the bass part and if any other synth will be involved in a 'rhythm' part, I add that too. Then I go back to the drums and program them in full. I might also add some percussion at this point, if I feel it needs it. I then tweak stuff so as to make it feel as good as possible — I get rid of all possible data interrupts first and then experiment with tick offsets and the like. When that stuff is complete, I add little sounds and pads, if needed."

Vocal Recording & Comping

Graydon's approach to vocal recording is equally meticulous and structured. "I never record the vocal until the machine track is complete and recorded, and if there are guitars that support the track in a big way, they too get recorded before the vocal. I record around five vocal passes on five tracks. If the singer is incredible, I record from the beginning to the end of the song. I want 'feel' choices as every pass is an interpretation. I then comp the vocal and after comping, if a line or two needs fixing, we do so.

"There are only a few singers I have worked with that have exceptional pitch: Rich Page [of Pages and Mister Mister], Donny Osmond, El DeBarge, and Howard Hewitt — Bill Champlin and George Benson should be included as well. These guys can nail a vocal on one track with maybe a few little fixes. You may think Al Jarreau would fall into that category, but he is not even close. His pitch is all over the place, needing many passes and major combines. If the singer has fair to bad pitch, I work in sections. I will record the five or so tracks of the first verse and 'B' section, and grade each line on the left side of the lyric sheet using a vertical and horizontal line graph with track numbers above. During the performance, I grade on a scale from eight to 10: nines and 10s are 'keepers'. I then look at the graph, and if a line does not have mostly nines and 10s, I punch in fixes on the 'bad score' tracks, allowing the singer a 'ramp' of at least two lines in front. I pay serious attention to sonic matching and in order to get the singer to sing at a similar level, it's most important to put the singer's mic in 'input mode' on the recorder at least one line before the punch‑in. If not, the singer will sing louder at the punch point due to not hearing him/herself in the headphones before the punch point.

More of Graydon's vintage synths: an Oberheim OB8 and two Yamaha DX7s.More of Graydon's vintage synths: an Oberheim OB8 and two Yamaha DX7s."The reason I stay in such a small section is that the singer's voice will start to sound different over time and I want to be able to match. After I feel I have good stuff, we move on to the chorus, in the same fashion, and so on. Sometimes the lead vocal takes a few days. If the singer's voice gets tired, I know when to stop: working the singer past that point can cause harm and it may take a few days before the voice recovers.

"I have been 'bad rapped' for using such a tedious process but the proof is in the final product. My goal was, and is, to get a great‑feeling comped performance that is in tune and sounds like it was in one pass. I am so very careful when comping to keep a flow and not have things sound like they were punched or combined. I tell singers that even though many listeners do not know what is in tune or not, the brain likes in‑tune pitch."

Mixing Techniques

Graydon adopts a similar approach to mixing too. "I work in sections. I want major track support and I want the vocal to ride the crest without being too loud. It is a give and take thing all the time. You bring up a level and then another instrument wants more attention. The give and take goes on for hours and you finally arrive at a happy medium. If things are not layering after many hours, it's time to start over, examining EQ and panning. Occasionally we are blessed with a mix that layers with little help, but there is only so much sonic space; If the track has a lot of information and is thick, EQ is the key to layering. If using pads (strings, or the like), roll out the bottom; if the bass is too thick, try adding a few dB around 800Hz and back down the level. It will speak and take up less room. Try panning things in areas, ie. instead of making all stereo information hard left and right, try 'panning in' slightly and use different spots for each stereo pair.

"You can get a great big drum sound and as soon as you add stuff, you are out of room. I put up the main elements like drums, bass, main keyboard and guitars and get a basic blend and then start tweaking EQ, panning and adding reverbs. I layer as I go along. If things get too thick, instead of trying to fight for room I tweak EQ. Negative EQ, especially in the low end, is a good friend if things get too thick. I explain this and other techniques in more detail in the recording techniques book I have written along with Craig Anderton, to be released soon."

Graydon takes a balanced view of the benefits of the increasingly sophisticated technology available in the studio today. "The key is still a good engineer, a quality recording chain, and quality performances. The gear of this era surely offers ways to do things that were not available in the past. The digital recorder allows so much flexibility, and things like multi‑band compressors allow buss compression to be slightly friendlier — before multi‑band compression, if the kick drum was forward in the mix, when it hit, all other frequencies were sucked back at the same time causing a pumping effect across the total mix.

"Mixing automation is very important but can be dangerous as well: smoothing out levels too much takes the life out of a mix, but so does buss compression. In this era, major compression is used quite often to get things as loud and punchy as possible on the master CD disc. Somewhere along the line, it seems a contest is in play to have the loudest CD. I prefer to keep the dynamics, so I never use compression when mastering. When mastering my new CD Bebop [see box] using the Alesis Masterlink, I used the 'look ahead' limiter mode on a few tunes to get the overall level to be in the same area as the other songs — just a dB or so. This type of limiter is very friendly as it sees the spike before it happens and smoothes it out instead of squashing it. It was only active for a few milliseconds per tune as the songs have major dynamics."

The Graydon Guitar Sound

One thing that is guaranteed to characterise a Jay Graydon production is quality guitar tone, from damped, clean, DI'ed rhythmic punctuations to his trademark rich, focused lead sound. He is, naturally, as meticulous here as with the rest of the recording process. "I typically use my Rivera signature amp series — the Rake amp head — and I go between 2x12 and 4x12 bottoms. I sometimes use a Marshall 4x12 bottom, depending upon the application. The amp head and the Rivera Jake combo are no longer in production but can be custom ordered [the URL is www.rivera.com]. I love these amps!

"I use a vintage Dan Armstrong Orange Squeezer compressor on the guitar, into an Ernie Ball volume pedal. The volume pedal is really not needed for most solos — typically, no volume swells — but I like to get rid of amp noise and finger position noise before I start playing. Sometimes, I fade out the last note of the solo. The volume pedal is always part of my setup. My typical solo setup includes an Eventide GTR4000 Harmonizer and a TC Electronic 2290 delay unit. These are applied at the console and never through the amp — adding pitch‑shifting through the amp causes pitch 'beating' which causes the amp gain stages and speakers to crap out. The same goes for the delay line to a certain degree.

The recording area at Graydon's studio, with a selection of his guitar amps. The Marshall speaker cabinets in the background are positioned, ready for recording, on top of a drum riser to decouple them from the floor.The recording area at Graydon's studio, with a selection of his guitar amps. The Marshall speaker cabinets in the background are positioned, ready for recording, on top of a drum riser to decouple them from the floor.

"Since I rarely use a room mic, the speaker bottom placement is not critical, but it is near the rear of my studio on the drum riser. The riser eliminates low‑end coupling with the floor. I am looking for a sweet mid‑range tone so as to not take up too much room in the track, meaning that I do not want low‑end information for solos. I use just one mic most of the time — occasionally, I will use a room mic, but only for major rock stuff. If you use more than one mic in close proximity and they are not the exact same distance from the speaker, comb filtering is in play, causing major frequency suckouts and dips. I also never use a condenser mic on a loud signal source: the key to using a condenser mic is to only use it on a sound source you would put your ear next to. I have tried so many dynamic mics over the years and always come back to the Shure SM57. I position the mic about one inch left of the edge of the speaker cone centre using a 22 degree angle and about an inch back from the grille cloth. This position allows for solid mid‑range as well as just a taste of the low end from the body of the speaker.

"I run the mic signal from the speaker bottom through a mic input on a module of my Neve V2 mixer. On the insert patch point output of this module, I patch into a George Massenburg EQ. The output of the GML EQ patches into a GML compressor and the output of the GML compressor returns into the Neve module's insert return. The mixer input channel is assigned to three busses — one for the recorder track, one for the Harmonizer, and one for the TC delay line. The Harmonizer and delay line return to two modules on the mixer which are assigned to the recorder track buss.

"The GTR4000 gets set ever so slightly sharp and the Harmonizer delay is set to 10 milliseconds (minimum). The delay line gets set to 45 milliseconds for a slight room mic effect simulation. Before I bring up the return levels for the Harmonizer and delay line, I set the EQ and compression for the dry guitar sound. Typically the compression is to just catch peaks — never more than a few dB. I then blend in the Harmonizer mixer module and then the delay line mixer module to taste. I hit an open string on the guitar and move around those faders until I get the blend I like. I then typically EQ the Harmonizer using a few dB of negative EQ on its mixer return module in the 10kHz area. For some reason, the Harmonizer seems to make the signal brighter even though on its own the source sounds the same. Go figure! Note that the Harmonizer level is typically at least 4dB lower than the source signal and delay level is a few dB less than the Harmonizer signal.

"I start with the monitoring level slightly higher than medium for dealing with EQ. This level will show if I have too much or (rarely) too little in the 2kHz 'pain' area. After I EQ in full, I monitor at a normal medium level. If I am having trouble getting inspired, I crank up the level and then as soon as I get into the groove, I back off to medium level again. If I am having pitch problems with the track, typically on inherited tracks I am paid big bucks to play solos over, I go to very low level and 'baby' the tuning. When the pitch seems to be OK, I go back to medium level. When songs have serious pitch‑shifting, I stay at a very low level and totally 'baby' the pitch using finger pressure on sustained notes to get in tune.

"I always record a tuning note in the front of the song. If the recording is with a real band, I ask the piano player to play A below middle C. Note that this A is slightly more than a cent flat with respect to A440. For rhythm playing, there's no problem regarding guitar tuning since the A below middle C on the piano is in the main piano chord range. When playing solos, I mentioned I use the Harmonizer slightly sharp blended into the sound. This usually makes up for piano tuning 'stretching' in the upper register. If not, I tune up a cent sharp — I always use my Peterson strobe tuner to tune to the track tuning note. All simple meter tuners are a joke regarding accuracy. When recording synth tracks, I tune all synths to A440 as most are not stretched, and record a 440Hz sine‑wave tuning note from a DX7.

"The bottom line on this issue is to always record a tuning note, for so many reasons: if the track is recorded on an analogue multitrack and the tape gets played on another machine for overdubbing, the odds are extremely good that the pitch will not be the same. I have two identical 24‑track analogue machines and one is slightly sharp in comparison. I have tested others and I have never found two to have exactly the same pitch."

Digital Or Analogue?

In the major league world of big‑budget productions, the digital versus analogue debate is still running, with large‑format analogue desks and multitracks often continuing to be favoured long after many had predicted their demise. Graydon works both with two‑inch 24‑tracks and ADATs, and sees benefits and drawbacks on both sides. "Analogue recorders have a big sound with warmth and if you want to hit the level hard, tape compression is musical‑sounding. In most cases, its a very reliable format too, with very few disasters. On the negative side, there's tape hiss and bouncing causes signal degradation, and there's a slight loss of high‑end information over time. In many cases, analogue multitracks are transferred to a digital format immediately after cutting the track so as to avoid the high‑end loss.

Jay Graydon has had a long association with Alesis as a beta tester. An SR16 drum machine rests on the Quadrasynth at the top of the rack; below it are two HR16b drum machines, for which Graydon supplied the samples, another SR16, a prototype HR16b (in standard HR16 grey casing), and a Yamaha DX7 synth.Jay Graydon has had a long association with Alesis as a beta tester. An SR16 drum machine rests on the Quadrasynth at the top of the rack; below it are two HR16b drum machines, for which Graydon supplied the samples, another SR16, a prototype HR16b (in standard HR16 grey casing), and a Yamaha DX7 synth.

"Digital recorders have very low noise compared to analogue, and there's also auto punch‑in, track time‑shifting and digital bouncing and digital copies with no generation loss. The negative side with digital recorders is the possibility of crashes that cause loss of data! If you do not make backups quite often, you will eventually lose some recorded information.

"The newer analogue‑to‑digital and digital‑to‑analogue converters also sound much better than in the past and this is the key issue with the medium. If major dynamics are involved, such as in classical music, the bit rate can make a difference in very quiet passages, but we are dealing with pop music so this is not such a big deal. The key is the quality of the converters and in this era, cheap converters are very good. Digital sonics in general these days are very good, even in the mid‑price stuff.

"One major problem with most digital recorders, however, is that they do not allow an analogue throughput, thus causing a delay when monitoring the source signal. For example, when recording a vocal using an analogue mixer, the vocal mic routes through the mixer, or an outboard preamp, to the digital recorder. The output of the digital recorder routes back through the mixer for monitoring. The problem is that if the digital recorder does not allow an analogue bypass — ie. if the digital converters are still in the audio chain — there will be a delay of at least 600 microseconds. The better the converters, the longer the delay; in this era, at least one millisecond. If any other digital gear is in the chain, the delay increases, especially if the digital gear is used in the analogue chain. The vocalist will be using headphones to monitor and is most likely hearing the vocal in the headphone mix. Typically, the vocalist will also pull the headphones slightly off one or both ears and the induced delay will then cause a comb‑filtering effect, sounding like the vocal is in an empty fishbowl.

"You can get around this by monitoring the vocal from the mixer vocal input and not the recorder return, but when punching in, you will need to monitor the recorder return and switch the monitoring at the same point you punch in. If you have a 'black‑face' or M20 ADAT, no problem, as there is a true analogue throughput option — I made sure of that, as I was a consultant and beta tester on both projects — but it is time we told all the manufactures of digital recorders and digital mixers to allow an analogue throughput mode. If this does not happen, as the converters get better in the future, the delay will get worse and that will be a problem for all analogue source recording!"

Hungry For Success

In recent years, Jay Graydon has maintained a relatively low profile for someone who has spent almost the entirety of his career to date at the very top of his profession. In that time he has observed the established order of the record industry under threat from both new forms of piracy and alternative methods of distribution. "The on‑line MP3 thing has been a setback for the recording industry, as we all know. Even though that particular problem has been fixed to some extent, similar sites are still operating. Everyone loves the fact that they can download free music, and many people copy CDs for friends, but here is the catch; less sales have caused record companies to cut back on staff and artist signings. The guy that copies a CD for a friend or downloads stuff for free does not realise that the consequence is less music to choose from in the future!"

Graydon's enthusiasm for the music itself, however, remains undimmed. "What motivates me is hearing great music — I get excited and I want to create. When I first started playing guitar, almost all pop music excited me. As time passed, my standards changed as I grew musically, but the one thing that does not change is a quality song with a thinking lyric, a great singer and great players."

He may already have crammed several lifetimes' worth of achievement into his career, but Jay Graydon now seems to have a renewed hunger for more. "For the last 10 years or so, I have been having fun! It's human nature to stay in the stress and grind of working; we are afraid to let go and just have fun. Inevitably, as we age, living life in 'funland' enters into the equation, but the hunger for further success continues no matter what. I am at the point now where I want to get back into the grind!"

Recording The New Jay Graydon Album Bebop

JAY GRAYDON

"I have just completed my new album, Bebop. The band consists of Dave Weckl on drums, Dave Carpenter on bass, Brandon Fields on alto and tenor sax, Bill Cantos on piano, and yours truly on guitar. I recorded the album in my studio using Alesis M20 pro ADAT recorders. I was involved with specifying some of the features of the M20 and I often beta‑test for Alesis and gear designed by Marcus Ryle [the chief designer of the M20 who is now with Line 6]. Since I needed to beta‑test the recorders, I decided to use a real band and I was impressed with the sonics!

"Regarding mics, I used a vintage AKG C414EB on the piano, Neumann KM84s on the acoustic bass, U47 on the sax, and C414 48s on the drum overheads — all of which I had modified by Klaus Heyne — along with a Shure SM57 on the snare, and many different mics for the kick, toms and hat. The guitar was recorded direct using a very rare direct box made by Eddy Reynolds many years ago. I did use a guitar amp on one song and the mic was an SM57. All of this ran through my Neve V2 mixer. I compressed the bass using a GML compressor and used GML EQ on the piano, snare, kick, sax, and guitar. Reverbs when mixing were very simple; a Lexicon 480 for the drums and a 480 for the guitar and sax, and I used an EMT plate for the piano.

"I did not play on the tracking sessions as I needed to concentrate on engineering and dealing with the beta version of the recorders. I was the only overdub so the project stayed in my room. For mastering, I used the Alesis Masterlink [CD recorder/editor]. A great tool!

"I am working on licensing deals for Bebop around the world at this time and have started releasing on‑line at www.cdstreet.com.

Jay will also be rereleasing Airplay For The Planet and the surf record Surfers Drive Woodies (by Rake And The Surftones) along with some previously unreleased material in the near future. Details of release dates and availability may be obtained via Jay's web site at www.jaygraydon.com.

Jay's Personal Studio

Graydon's own studio is based around a Neve VR desk, which he uses to record either to his ADATs or to analogue 24‑track.Graydon's own studio is based around a Neve VR desk, which he uses to record either to his ADATs or to analogue 24‑track.

Jay Graydon's own studio is based around a mid‑'80s Neve V2 large‑format analogue console, partnered with 64 channels (eight machines) of Alesis ADAT digital eight‑tracks and a pair of MCI two‑inch 24‑track analogue recorders. The console is hard‑wired to the recorders using Monster Cable to maximise sonic integrity by removing all unnecessary connections from the signal path. It also has automated cuts on the monitor section to minimise noise when using them as effects returns during mixing, and further tweaks to reduce the amount of electronics in the monitor amp path.

Favoured keyboards and modules include Yamaha DX7, ARP 2600, Minimoog and Oberheim Xpander synths, Linn 9000 drum machine, and the Alesis Quadrasynth and HR16b drum machines for which Graydon originated the samples. Principal outboard consists of Lexicon 480s for reverb, Eventide GTR4000 and H3000 Harmonizers and GML (George Massenburg) equalisers and compressors.

From Oberheim To Emagic: Graydon's Sequencers

"The first sequencer I used was the Oberheim DMX, in conjunction with the OB8 synth and DSX drum machine. The clock from the drum machine was recorded to tape as track 24 on the 24‑track MCI. After recording the clock to tape, I used a Y cord to clock the drum machine and sequencer at the same time. The first change was to eliminate the DSX and use a Linn Drum along with the DMX. The DMX is a 96 clock device and the Linn Drum is a 48 clock device. I asked my tech, Ian Eales, to build a box that would split the clock in half for the Linn Drum, and both units clocked together very well!

"The Linn 9000 was next — a sequencer and drum machine in one box. All setups have a 'feel'; this one had a slight problem in that, since the drum sounds were internal to the 9000, they triggered quicker than the MIDI output sent to the synths. You might think it would be a good idea to make three passes when recording to tape — one for recording the clock on tape, the next for drums, delaying the recorded clock on tape to around 20 milliseconds, and then experiment with a shorter delay for the synths to make up for the MIDI delay/synth processing time for the third pass. Well, that never felt good, as there are always data interrupts within the sequencer. No matter how many data interrupt functions you turn off, there will still be timing slop. I lived with the slight delay in the synths and always ended up recording the drums and synths in one pass. In this era of digital recorders allowing micro timing track offsets, the problem could be improved using an average.

"The next thing I used was [MOTU] Performer software, which meant another set of major data interrupts — more than before, as the hardware is not dedicated to the software. Performer allows tick offsets, but the offsets should be at least 10 times greater. As time went on, Performer kept getting new features adding more data interrupts. Again, I recorded all my drums and synths in one pass to keep the clocking common.

"I am now about to switch to Emagic's Logic software, as I have been told the data interrupts are less than the other sequencers. As usual, it'll be another time‑burner to learn the program's nuances and the workarounds for data interrupts."

Working With Donald And Walter

Steely Dan's 1977 Aja album is now considered by many to be the pinnacle of the band's career. Compared to the guitar‑driven excitement of its predecessor, The Royal Scam, which had seen the introduction of session star Larry Carlton to the studio band, Aja was gentler, more introspective, built around acoustic piano and, for the first time, Walter Becker's simple clean‑toned guitar lines. Graydon's 'outside' solo in the track 'Peg', therefore, stands out all the more in contrast. Like all great solos, the piece is a mini‑composition in its own right, and pulls off the curious feat of being both angular, almost disjointed, and yet always intimately connected to the track, taking the listener effortlessly into dissonance and back out again as if it were perfectly obvious that that's how it should go.

JAY GRAYDON

"The solo took about four hours of playing in full, including the fade (with many breaks). I was the seventh cat to play the solo as they could not get a solo they liked from other great players. Donald and Walter wanted to stay on one track [of the tape] as they were holding other solos and were most likely out of tracks. Ideas were lost and I thought some were worth keeping, but around hour three they guided me into the key to the thinking: Donald said something like 'think blues over the chord changes'. The chord changes did not outline blues, as Cmaj7 to G2/B are 'pretty' chords. OK, the G2 can deal with blues notes, so I went into my double‑stop note‑bending and made sure I landed on the fifth and dominant seventh of that chord. The rest of the solo, and fade blowing, fell into place from then on.

"My style of playing is to hang on the edge and take chances in sheer terror! I just go for it and hope I connect ideas across the chord changes. In some passes, I totally suck and others seem to work out in the melodic flow. With that in mind, a solo may come in the first pass or I may take hours and need to combine.

"For the 'Peg' solo I used my red '63 Gibson 335 that I used for many years, set on the bridge pickup, with an Orange Squeezer to an Ernie Ball Volume pedal, to a Fender Deluxe highly modified by Paul Rivera; the volume was on 5, but I don't remember the other settings. I am fairly certain the mic was a Shure SM57. I am always in the control room when I overdub solos — you have to listen over the monitor speakers as headphones make things sound too good — so I had a 50 foot guitar cord from the volume pedal to the amp."

Absolute Phase Is Important

"When recording real drums, all mics that hear the sound source from above are reversed in phase. Practically every record you have ever heard has this problem. For example, when you use a single mic on a snare drum, you mike it from above. When the drum stick hits the snare head, it pushes the head away from the mic, thus causing the air to move away from the mic diaphragm. This equals a negative waveform at attack. Does this really matter, as it is just one cycle of the waveform? Yes, as the attack is not pushing air forward through the monitor speakers. This leads to less punch and less bottom end. It is slight, but does make a difference. If you are using two mics on the snare, miking on the bottom and top of the head, engineers typically reverse the phase on the bottom mic. This is backwards, as the bottom mic is seeing air pushed towards the mic diaphragm. It is the top mic that should be phase‑reversed on the mixer.

"You should always note down on the track sheet whether you flipped the phase to the recorder or need to do it later. The kick drum is the only mic that does not need to be phase‑reversed in this way. The studio wiring path, of course, is critical — pin 2 or pin 3 hot has been a problem since the '70s as, until recently, America went against the rest of the world in using pin 3 hot. You must check your studio for phase with a phase tester — a boring gig, but very necessary!"

The Trouble With Auto‑Tune...

"Auto‑Tune is a great tool. There are singers that really need this tool — you know, the good‑looking artist the record company signs knowing that the pitch corrector will make the record sound believable. Live, the singer will sound terrible but that is not the producer's problem. Hey, maybe they use the pitch corrector live too. The bottom line on this issue is that it is a common problem for tools to get over‑used in the beginning of their existence. Eventually people get it in perspective and it starts getting used tastefully.

"There is a problem with real‑time pitch correctors, though: there is a delay of no less than around seven milliseconds and possibly up to 15 milliseconds — I have performed tests. The big problem here is that it is a random delay — if the delay time was fixed, a track time offset could be used during or after the bounce in regard to the pitch‑corrected new track. I've mentioned this to a manufacturer and it's time I checked back with them to find out if they added such a mode — the fix would be to have a mode that forces the delay to maximum at all times, noting the exact delay time for each bit and sample rate. This is a 'must' fix so as to keep the vocal feel intact when bouncing."