When Walter Becker passed away in September 2017, it marked the end of one of the most quixotic music partnerships in modern American popular music. The only constant members of Steely Dan, Becker and co-conspirator Donald Fagen created some of the most lyrical intelligent, musically intricate, and sonically impeccable albums of the past 40 years.
For both record buyers and audiophiles, none looms larger than 1977’s Aja, the subject of the third installment of “The Best Version Of…” (TBVO).
Together with longtime producer Gary Katz and engineers Roger Nichols and Elliot Scheiner, Becker and Fagen marshalled their (in)famous attention to detail to make Aja their best selling and best sounding release. Aja was Steely Dan’s first platinum album, moving over three million copies and spawning three top-25 singles (“Peg,” “Deacon Blues,” and “Josie”) in the year of Rumours, Saturday Night Fever, and Never Mind the Bollocks — no small feat for a set of intricately crafted, jazz-inflected meditations on nostalgia, mortality, and failure. Aja also quickly became a standard audiophile reference album for testing equipment. As Bowers & Wilkins’s Doug Henderson recently recalled, a common experience for anyone leading an equipment demo at a trade show became “anticipat[ing] the usual Steely Dan requests but want[ing] to avoid looking glum at the umpteenth playing of [Aja’s] ‘Josie’[.]”
The road to Aja began in 1967, when Becker and Fagen met at Bard College in New York’s Hudson Valley. As Fagen remembered, “We started writing nutty little tunes on an upright piano in a small sitting room in the lobby of Ward Manor, a mouldering old mansion on the Hudson River that the college used as a dorm.” Their experience at Bard also served to heighten Becker and Fagen’s already keenly-developed outsider orientation. “I noticed that everybody had a car, except me and Donald,” Becker said later. “I reluctantly had to conclude that not all socioeconomic groups were equally represented.”
Like many suburban white kids of their generation, Becker and Fagen admired the black music of the era. But unlike many of their peers this didn’t draw Becker and Fagen into the often-derivative white blues-rock milieu. Instead, Becker and Fagen blended their love of Howlin’ Wolf and Duke Ellington with influences as diverse as the Great American Songbook, Dylan, “Brother Ray,” and Leiber and Stoller to create a singular sonic amalgam that had as much in common with the hermetic studio creations of Todd Rundgren and Stevie Wonder, the cynical singer-songwriting of Randy Newman and Warren Zevon, or the glistening pop-funk of Earth, Wind & Fire (whom Dan aficionado Pharrell Williams once dubbed “the black Steely Dan”) as with their supposed “jazz rock” peers like Weather Report and Chicago. (“I’m not interested in a rock/jazz fusion,” Becker told Rolling Stone in 1974. “That kind of marriage has so far only come up with ponderous results. We play rock and roll, but we swing when we play. We want that ongoing flow, that lightness, that forward rush of jazz.”)
Becker and Fagen combined this unique musical mélange with a love of “W.C. Fields, the Marx Brothers, science fiction, Nabokov, Kurt Vonnegut, Thomas Berger, and Robert Altman,” as Fagen put it, to create songs that were as lyrically obtuse as they were musically complex. “I’ll tell you what I like about our group,” Fagen said in a mid-‘70s interview. “What I like about us, outside of our technical accomplishments, is that our music scares me more than anybody else’s. The combination of the words with the music — like a cheerful lyric and a sad or a menacing melody, or vice versa — I find that irony frightening…. Not music about doom and melodrama — that kind of stuff isn’t really frightening. What’s really frightening is mediocrity. The mediocrity of everyday life, the mediocrity we see around us. That frightens me.”
The “frightening” irony of Steely Dan’s lyrics wasn’t lost on those listening closely, even if it was missed by many casual fans. “Steely Dan [was] a group who served as the house band for every 1978 West Coast singles bar,” writer Chuck Klosterman quipped, “despite being more lyrically subversive than the Sex Pistols and the Clash combined.” As with Dylanologists, the elisions in and obscurantism of Becker and Fagen’s lyrics would send dedicated Dan-o-philes scurrying to decode their meaning. “We don’t construct them as puzzles,” Fagen claimed to Rolling Stone in 1977. “We try to tell a big story in a very short period of time. Naturally we have to exclude some information. We don’t discourage any speculation.” Nonetheless, Becker and Fagen were loath to explain their songs to interviewers, preferring to allow conjecture to run rampant.
Both musically and lyrically, Becker and Fagen’s songwriting partnership was a true collaboration. “I usually come up with germinal musical idea, and then we will arrange to meet…,” Fagen told the New York Times. “t is really a collaboration. It’s not one of us writing the music, the other lyrics. And it’s not like Lennon and McCartney, who as I understand it usually just wrote a song by themselves and then put both their names on it. It is a collaboration: we think very much the same musically. I can start songs and Walter can finish them.”
The songwriting duo took their first serious stab at making a career of it when they left Bard and moved to New York City, harboring hopes of becoming Brill Building-style songwriters. Save for placing a tune with Barbara Streisand that was “altered beyond the point where we would have to take responsibility for it,” according to Becker, Becker and Fagen’s efforts were largely fruitless. But their songwriting impressed Kenny Vance of Jay and the Americans (whom Becker and Fagen sometimes backed on the road) and Katz, who landed Becker and Fagen jobs as staff writers for ABC Records in Los Angeles.
Eventually, Becker and Fagen’s continued inability to write tunes suitable for other artists (“We’re not particularly good popular-song writers,” Fagen explained to one interviewer) led to the realization that they were the only ones who could perform their songs. So, with Becker on bass and guitar, Fagen on keyboards and vocals, Denny Dias and Jeff Baxter on guitar, and Jim Hodder on drums, Steely Dan — a name borrowed from a dildo in William Burroughs Naked Lunch — was born.
Steely Dan’s debut, Can't Buy a Thrill, was released in November 1972 and peaked at number 17 on the Billboard pop album charts, buoyed by top-20 singles “Do It Again” and “Reelin’ In the Years.” Dan followed Can’t Buy a Thrill with Countdown to Ecstasy in July 1973, which peaked at number 35 on the charts, and Pretzel Logic in February 1974, which peaked at number eight. The former featured a minor hit about their time at Bard, “My Old School,” and latter spawned their biggest single, “Rikki Don't Lose That Number, which reached number four on the Billboard Hot 100.
Despite Steely Dan’s success, the bookish and mordant Becker and Fagen couldn’t have had less in common with the charismatic and photogenic duos (Page and Plant, Jagger and Richards) behind the early-‘70s other big acts. Indeed, the era’s music press seemed to take great joy in describing Becker and Fagen’s appearance in the least flattering terms possible.
Rolling Stone, 1974:
[Becker] wears smoked glasses, accenting a somewhat gnomish face marked by high cheekbones and an upswitched nose. He has the fool, slouched posture and crooked, potentially menacing smile of somebody who was probably a sarcastic outcast in junior high school. Donald Fagen…has [the] pinched rectangular face of his photos, which…in person turns out to be dominated by an alarmingly long nose, wide mouth and high forehead, giving him the aspect, perhaps, of a mad scientist.
The New York Times, 1977:
[Fagen] looks like Victor Mature reflected in a funhouse mirror that widens and elongates. He speaks in a slow, laconic New Jersey drawl (he was born in Passaic), and when his wide mouth smiles, the grin resembles a sneer…. Walter Becker is the fast talker and wisecracker. A small slender New Yorker, longhaired with a wispy moustache and beard, he looks like Peter Pan impersonating Fu Manchu.
When music journalist-turned-director Cameron Crowe — who also colorfully described Becker and Fagen’s looks in his ’74 Rolling Stone profile of the band — had one of the groupie characters in his 2000 film Almost Famous proclaim “Wow, they’re cute” while watching the Steely Dan play “Reelin’ in the Years” on “Midnight Special,” it was unambiguously understood as a punch line.
By the mid-1970s, Becker and Fagen’s evident discomfort in the spotlight was fracturing Steely Dan as a functional unit. With each succeeding album, Becker and Fagen’s desire to tour decreased and their desire to use studio musicians to augment (and, eventually, supplant) the band’s other members increased. “The situation just sort of evolved into having a band and making records with them…,” Fagen explained. “But after a couple of records, we decided that the situation was too limited for the kind of music that we were writing, so we arranged to have other musicians brought in. And that finally evolved into a situation where we could hire whomever we wanted to play individual songs.”
The band played its last live date in July 1974 and the original lineup dissolved, leaving Becker and Fagen as the only two full-time members of Steely Dan. By 1975’s Katy Lied, Steely Dan was Fagen and Becker augmented by Los Angeles’s top studio musicians, including Chuck Rainey (who’d already recorded with Yusef Lateef, Roberta Flack, Donny Hathaway, and Aretha Franklin) on bass and Jeff Porcaro (who would go on to record with Boz Scaggs, Hall and Oates, Jackson Browne, Pink Floyd, and Aretha Franklin) on drums. (Dias, who nominally remained Steely Dan’s third member, played on only one of Katy Lied’s tracks.)
Becker and Fagen’s increasing use of studio musicians in the pursuit technical virtuosity was matched by their growing desire for sonic perfection, aided by Katz and Nichols. “I think one of the best things about rock and roll as opposed to jazz is precision and a professional sound,” Fagen said in ‘77. “That’s what I like about popular music. We strive for that sort of slick sound.” A side effect of the band’s embrace of “slickness” was that Becker and Fagen became the poster boys for the “L.A. Sound,” even though they had little use for their adopted home city. (“I don’t love L.A. particularly,” Fagen lamented in 1974. “I mean, it’s comfortable to live here, but you know they’ve got it all set up so you never see any poor people? You never have to drive through Watts, say. When we moved out here, after a while we realized we hadn’t seen any black people — it really stuck out. ‘Where’d half the people go?’”)
By design, each succeeding Steely Dan album seemed to sound better than the last. “The strive for true hi-fi was common ground with Donald and Walter and Gary — we’re all perfectionists, especially Walter with his quad electrostatic speakers at home and the latest tone arm,” Nichols told Dan biographer Brian Sweet in 1993. When a mishap with the new DBX noise-reduction system damaged the Katy Lied tapes, the band undertook herculean efforts to fix the snafu and were devastated when the album’s original fidelity couldn’t be completely restored. (Katz: “I can’t listen to it. I hate to hear an album that we’re involved in that’s not up to our standards. It was the best-sounding thing I ever heard before it was ruined.” Becker: “If you had heard that album the way it originally went down on tape, you would have heard something else.”) But a sonic debacle by Steely Dan standards was still audiophile quality for its time. “It is a testimony to their studio prowess that the ‘flawed’ sound on Katy Lied is still much advanced compared to any of the competition,” Crowe mused in his ’77 piece.
Despite the sonic advances of Katy Lied and 1976’s The Royal Scam (which took more than three months and over $100,000 to record), Steely Dan’s commercial prospects had dimmed since they stopped touring. Both albums charted lower than Pretzel Logic, as did each album’s most successful single. (“Black Friday” from Katy Lied reached number 37, and “The Fez” from The Royal Scam peaked at 59.)
Aja would change all that.
Becker and Fagen began recording Aja in Los Angeles in January 1977. It would mark the apotheosis of their quest for studio perfection. “I thought Aja itself was dangerously ambitious,” Becker told Musician magazine in 1981. “I really did.”
Aja required a budget that made The Royal Scam’s look like “chicken feed” and took longer to record than any previous Steely Dan album, even though with only seven cuts — “Black Cow,” “Aja,” “Deacon Blues,” “Peg,” “Home at Last,” “I Got the News,” and “Josie” — it contain the fewest songs of any Dan album. (It didn’t help that only one of the cuts clocked in at less than four minute and two clocked in at over seven.) Most of that time was spent on endless attempts at the same parts. When Crowe asked Becker and Fagen how they spent their days, Becker quipped: “Overdubbing. We overdubbed a lot of the overdubs over.” “Yeah, we did,” Fagen added.
“We just kept adjusting our standards higher and higher,” Becker explained to GQ in 2014. “So many days we’d make guys do 30 or 40 takes and never listen to any of them again, because we knew none of them were any good. But we just kept hoping that somehow it was just going to miraculously get good.”
“Every track, every overdub, had to be the perfect overdub,” Scheiner told Newsweek in 2017. “They didn’t settle for anything. They were always looking for the perfect.
As a result of their snowballing perfectionism and unlimited access to L.A.’s top players, Becker and Fagen’s own musical role on Aja diminished. “It wouldn’t bother me at all not to play on my own album,” Becker told Crowe with utter seriousness.
“Around the time we made Aja we figured out what it was we sort of wanted to do, you know, musically,” Fagen explained in the Aja episode of the Classic Albums series. “We realized we needed session musicians who had a larger palette of things they could do.”
“Donald and I had more of an idea that comes from an East Coast Brill Building tradition, of an almighty producer, when you had a Leiber and Stoller, or at its extreme a Phil Spector, who knew exactly what they wanted,” Becker elaborated in 2000. “What Burt Bacharach did with Dionne Warwick. He was looking for a diva to front his outfit and found her singing in a gospel group in New Jersey. His music was very difficult, and he needed someone who could execute what he was looking for.”
Becker and Fagen could be tough taskmasters in pursuit of their idea of perfection. “They are the most demanding group of people in the industry that I’ve worked for,” guitarist Larry Carlton, whom Becker and Fagen trusted to write many of their arrangements, told the Times in ’77. “Nothing goes with flutter on it. If three of the guys are cutting the part great and one doesn’t feel just right, they’ll call in a whole new band and redo the whole thing.”
But according to Katz, there was a method to their madness. “I was always amazed that they pretty much heard in their heads what it was gonna be like completely…,” Katz said in the Classic Albums episode. “[A]ll through the project, they would know, ‘Nope, that’s not it. That’s not working. This what I want.’ And it was amazing that, when the thing got done, finally I could see what everything was gonna be like. But they knew from the very beginning.”
On Aja, Becker and Fagen integrated their diverse musical influences more effortlessly than ever before. As Winston Cook-Wilson gushed in his Spin piece on Aja’s 40th anniversary:
Guitars provided auxiliary punctuation and effects-less solos rather than the brunt of the song; a stew of acoustic piano and electric keyboards, reminiscent of Miles Davis’ In a Silent Way, were at the warm center of the mix. Aja’s sound was a direct offshoot from jazz and fusion, steeped in its harmonic language, as well as that of turn-of-the-century modernist classical music (Debussy and Stravinsky, especially). The particular musical syntax on Aja was in many ways uniquely Dan’s, however, the misbegotten result of Becker and Fagen’s own self-taught musical education. Their chordal sense was central to the issue: The complex changes left the average rock listeners’ ear out in the cold, pointing toward whole new keys for choruses and away from easy resolution.
The musical complexity of Aja spawned academic studies of Steely Dan’s chord changes and prompted a Berklee College of Music songwriting course analyzing Becker and Fagen’s oeuvre. (For those interested in a slightly less academic, but no less detailed, analysis of the musical and lyrical intricacies of Aja, pianist Don Breithaupt’s short book on the album for Bloomsbury’s 33 1/3 series is highly recommended.)
Aja works so well together as a whole that singling out individual songs seems almost perverse. However, “Deacon Blues,” “Aja,” and “Peg” are not only among the album’s highlights, but also illustrate some of the musical and lyrical elements that make Aja so special.
For two lyricists who shied away from songs that allowed for easy interpretation, “Deacon Blues” was surprisingly clear and direct. “Deacon Blues is about as close to autobiography as our tunes get,” Fagen admitted in the Classic Albums episode.
The song’s protagonist is man of presumably advancing age (Becker called him “a broken man living a broken life,” though Fagen has sometimes referred to him as a “kid”) who looks at his staid suburban life and longs to throw it all away in order to play saxophone in dive bars, even if the ultimate price is death. “You call me a fool, You say it's a crazy scheme / This one’s for real, I already bought the dream,” he argues with his skeptical partner (or, perhaps, his own better judgment). “So useless to ask me why, Throw a kiss and say goodbye / I'll make it this time, I'm ready to cross that fine line / Learn to work the saxophone, I play just what I feel / Drink Scotch whiskey all night long, And die behind the wheel.”
It’s a pipe dream, and the protagonist knows it. He can’t even play the saxophone. “The protagonist is not a musician,” Becker explained. “He just sort of imagines that would be one of the mythic forms of loserdom to which he might aspire. And you know, who’s to say that he’s not right?”
With “Deacon Blues,” Becker and Fagen narrated not only their own pre-Steely Dan lives (and what may have remained their lives had Katz never gotten them their ABC gig), but also the lives of many of their fans, for whom Steely Dan’s music provided the same dreams of escape that the songs on late-night jazz and R&B radio stations once provided to Becker and Fagen. “You know we were both kids who grew up in the suburbs,” Fagen said. “We both felt fairly alienated. Like a lot of kids in the ‘50s we were looking for some kind of alternative culture — some kind of escape, really — from where we found ourselves.”
For the protagonist of “Deacon Blues,” as for Becker and Fagen, that alienation isn’t something to be lamented; it’s something to be embraced. “I think that a lot of kids our age were very alienated,” Becker explained of “Deacon Blues” in a revealing 2008 interview. “To this day when I read some text that somebody writes about alienation, I always think to myself, Gee, they make it sound like it’s a bad thing!” For Becker and Fagen, alienation was the fulcrum around which their discovery of music revolved. “[The protagonist] turns to jazz and hip culture as something to grab on to,” Fagen told Sweet. “And the basic idea is that there’s a kind of culture of losers that he’d rather be part of than the general way of life in America.”
Ultimately, “Deacon Blues” is a tribute to the other working- and middle-class kids like Becker and Fagen who, through music and the entire “alterative culture” that comes with it, learned to reject the façade of the American Dream and embrace their outsider status, even if they remained “losers” by the standards of the kids who drove “Porsches and Jaguars” around the college campus. “You know, they’ve got a name for the winners in the world, and the losers should have some sort of franchise as well,” Fagen has explained. “And the name that he has chosen that conveys a certain power is ‘Deacon Blues.’” (Becker and Fagen landed on the name for the winners, “Crimson Tide,” by asking former tour manager Warren Wallace for the names of successful college football teams and choosing the University of Alabama’s moniker.)
Musically, “Deacon Blues” is marked by Carlton’s supple rhythm guitar, saxophonist Tom Scott’s hypnotic horn arrangement, and Pete Christlieb’s improvised sax solo. “They just told me to play what I felt…,” Christlieb told the Wall Street Journal. “[So I] recorded my first solo. We listened back and they said it was great. I recorded a second take and that’s the one they used. I was gone in a half-hour. The next thing I know I’m hearing myself in every airport bathroom in the world.”
Christlieb also solos over the song’s slow fadeout, which stretches over 30 seconds. “The song’s fade-out at the end was intentional,” Fagen said in 2015. “We used it to make the end feel like a dream fading off into the night.”
Becker and Fagen refused to edit out Cristlieb’s solo to shorten the seven-plus minute cut to more suitable single length. ABC released it anyhow, and it became a top-20 hit despite its length.
“‘Deacon Blues’ was special for me,” Becker marveled years later. “It’s the only time I remember mixing a record all day and, when the mix was done, feeling like I wanted to hear it over and over again. It was the comprehensive sound of the thing: the song itself, its character, the way the instruments sounded and the way Tom Scott’s tight horn arrangement fit in.”
While the meaning of “Deacon Blues” is fairly straightforward, the cryptic “Aja” has been meticulously dissected by writers and fans. In the Classic Album episode, Fagen calls “Aja” “a journey in space and time…[about] the sort of tranquility that can come of a quiet relationship with a beautiful woman.”
Musically, “Aja” is an example of how Becker and Fagen’s pairing of studio musicians could result in spontaneous magic. The song’s highlight is the interlocking solo by saxophonist Wayne Shorter — who had played on innumerable classics by Miles Davis and Art Blake but who usually eschewed pop music — and New York session drummer Steve Gadd.
“The title song I like…,” Becker told Sounds magazine in 1977. “We’d gotten this drummer we didn’t know but had heard a lot about, named Steve Gadd…. We had a chart for the tune, and it was like eight pages long — three music stands in front of every musician. What’s on the charts is very specific for some of the players…but very open for others…particularly the drummer — he really had to outdo himself on that one.”
“There was a little mark on the chart for Steve Gadd to ad-lib through a certain part and add a couple of different parts that we figured we’d talk about [after the initial takes] and so on,” Fagen explained to Sweet, “but [Gadd] just ripped right through it on the first take and we kept it.”
Gadd’s tempestuous solo has become legendary among drummers, who’ve dissected its every nuance (including a beloved stick click near the five-minute mark that has generated much debate — intentional or accidental?). Shorter’s biographer Michelle Mercer also credits Gadd for helping to inspire Shorter’s equally legendary solo, which she calls “majestic and stately, tracing a mountainous arc with cleverly displaced references back to the vocal melody.”
If the solo on “Aja” illustrated the magic of first takes, the solo on “Peg” demonstrated how Becker and Fagen’s willingness to overdub (and overdub and overdub) with different musicians could also yield transcendent results.
“Peg” was the last tune recorded for Aja, cut in New York while the rest of the record was being mixed. The tight, infectious groove laid down by Rainey and drummer Rick Marotta was almost automatic. “We had done stuff with them before so we knew what to expect,” Marotta said. “Chuck and I had played together so much that we got into a groove… [Y]ou could have hung a coat up on the groove.” (In the process of laying down that groove, Rainey smartly defied Becker and Fagen’s instructions not to play slap bass by erecting a studio partition so that they couldn’t see that he was slapping.)
The guitar solo on what Becker and Fagen called a “pantonal 13-bar blues with chorus” didn’t come nearly as easily. “[‘Peg’] I think is infamous among studio players in that we hired a couple guitar players to play the solo,” Fagen said in the Classic Albums episode, “and it wasn’t quite what we were looking for until we got through three or four, five...six or seven, eight players.”
They tried Elliott Randall, who’d played the “Reelin’ in the Years” solo, along with former McCoy’s frontman Rick Derringer, who’d played on Countdown to Ecstasy and Katy Lied. But Becker and Fagen weren’t happy with the results.
“Rick Derringer was there for about three or four hours,” remembered Scheiner. “We got something out of him. The minute he left, Walter looked at me and said, ‘Erase it.’ I said, ‘OK.’ You never questioned it. You didn’t say, ‘Come on, really?’ It was over.
“We were embarrassed for them and for us,” Fagen admitted to Breithaupt. “We felt silly spending all this money for this one brief blues solo.”
They finally brought in L.A. session guitarist Jay Graydon, who’d played on albums by Marvin Gaye, Joe Cocker, and Wayne Shorter.
“I found out I was the seventh guy. For about an hour and a half, I’m playing my hip, melodic kind of jazz style. Then Donald says to me, ‘Naw, man. Try to play the blues.’… [So] I play bluesy for a while. I get melodic for a while. I get bluesy again. Then I get melodic and bluesy,” Graydon told Newsweek in 2017. “The whole thing probably took about four, five hours…. When I walked out of the studio at the end of the night, I didn't know it was a keeper. I turned the radio on one day, and there it is.” (As Aja’s first single, “Peg” was all over the radio in 1977, reaching number 11 on the Billboard charts during its 19-week run.)
While “Deacon Blues,” “Aja,” and “Peg” are some of my favorites, deep cuts like “Home at Last” (which Fagen called “a little blues about Ulysses”) are just as good. And if you read Breithaupt’s book or watch the Classic Albums documentary, it becomes clear that every song on Aja has a compelling story, because every note and every word was intentional.
“An album like Aja matters not just because it contributes to civilization a handful of date-stamped audio treasures, but because it puts into sharp relief the dreck that surrounds it,” according to Breithaupt. “How else to know the true banality of ‘Come Sail Away,’ ‘Hot Legs,’ and ‘Three Times a Lady’ (all of which shared chart space with songs from Aja)?”
With the fanatically pursued perfection of Aja, Becker and Fagen exposed the “mediocrity” that so terrified Fagen. That’s why it’s worth it for true Dan-o-philes to pursue the perfect digital version or Aja — or as close to it as we can get.
But the basic story is that there have been (at least)1 seven digital masterings of Aja: 1) a 1984 CD mastered by Steve Hoffman, 2) a 1984 CD mastered by Nichols, 3) a 1984 Japanese CD with uncertain mastering credits, 4) a 1988 Mobile Fidelity Sound Lab CD remaster, 5) a 1993 remaster by Glenn Meadows found on the Citizen Steely Dan CD box set, 6) a 1999 CD remaster by Nichols, and 7) a 2010 “flat transfe[r] from Japan[ese] original analogue master tapes” by Hitoshi Takiguchi at Tokyo’s Universal Music Studios, used for both a 2010 SACD and several subsequent CDs.
Aja’s path to digital began years before the first CD release of the album.
With the help and encouragement of Nichols, Becker and Fagen had been early adopters of digital technology in the studio. (They deployed “Wendel,” Nichols’s pioneering digital sampler for drums on 1980’s Gaucho.) Nichols was just as enthusiastic about the possibilities of digital for home audio. As Nichols wrote 1991:
I originally got involved in recording music because I hated clicks and pops on record. I figured that the only way that I was going to get good quality recordings to play was to record them myself. I could then bring home two-track 15 ips copies to play on my stereo…. When the Compact Disc became a reality, I was beside myself. I was also close by the side of any record company exec who could get me any discs to play on my new found CD player. Since CDs preserved all the characteristics of the original master tape, I could now enjoy music without the drawbacks of black vinyl.
Not surprisingly, Nichols was proactive in preparing the Steely Dan catalog for CD release.
According to Nichols and ICE magazine, in either 1981 or 1982, Nichols transferred all of the Steely Dan master tapes to Scotch 3M digital tape, with the intention that these digital flat transfers, not the original analog masters (which Nichols later said “were in terrible shape, due to improper storage, and had poor fidelity”), would be used for subsequent CD releases. (Notably, the master tape for the b-side of Aja could not be found, and a tape copy was used for the digital transfer.)
However, the source of the '80s CD version(s) of Aja are shrouded in mystery.
According to mastering engineer Steve Hoffman, when it came time to master Aja for its first CD release, Hoffman mastered it at Bruce Botnick’s Digital Magnetics studio in Hollywood from the original analog tape, not Nichols’s digital transfer. According to Hoffman, the Aja master tape “sounded quite dull” and had a “midrange suckout problem,” which he attributed to it being mixed on horn speakers that exaggerated the midrange. So, Hoffman’s mastering “fill[ed] in the midrange hole” by boosting the mids a bit, but was otherwise a flat transfer of the analog master.
In Hoffman's telling, MCA pressed around 5,000 CDs with Hoffman’s mastering, but as many as 4,000 of them were destroyed when the band decided they wanted Nichols to have control of their CD releases. While Hoffman’s mastering wasn’t supposed to be released, he says that some nonetheless eventually found their way into stores.
Mastering engineer Glenn Meadows, who worked with Nichols and the band on subsequent CD releases, told me that the '80s CDs produced by the band also came from the analog masters. These masters were transferred to digital tape on Sony PCM 1600 system, which was the preferred format for sending to CD pressing plants. According to Meadows, it was only after these masters — which Meadows attributes to both Nichols and famed mastering engineer Bob Ludwig — were prepared that Nichols made the flat transfers of the original master tapes to the 3M digital tape format, and it's those 3M tapes that both Meadows and Nichols used for their respective '90s CD masterings. (For more on this confusing story, see my October 2021 update below — an update that calls into question the existence of both an '80s Nichols mastering and a Hoffman mastering.)
How do what are commonly referred to as Nichols's '80s mastering and Hoffman's '80s mastering compare? On a few tracks, Nichols’s mastering differs little from Hoffman’s. But on most tracks, Nichols’s ’84 mastering has more upper-midrange and treble than Hoffman’s, leading to a brighter sound, as the GIF below comparing Nichols’s (light blue) and Hoffman’s (purple) masterings of “Aja,” “Deacon Blues,” “Home at Last,” and “Peg” using Har-Bal’s “average power” graphs demonstrates:
(click on image to see the animated comparison)
Despite the significant differences in sound, it’s difficult to visually distinguish Hoffman-mastered ‘84 CDs from Nichols-mastered ‘84 CDs, since the liner notes don’t list the mastering engineer. But online sleuths have gone to great lengths to figure out which CDs contain the Nichols mastering and which contain the Hoffman mastering based on catalog and matrix numbers.
Making matters more complicated, there’s also a mastering on some ’84 Japanese CDs that’s neither Hoffman’s nor Nichols’s. (Thankfully, it’s easy to identify by its catalog number.) This Japanese mastering is overall more similar to Hoffman’s mastering than the Nichols’s, but it’s much more midrange-focused than Hoffman’s mastering, as the GIF below comparing Hoffman’s mastering (purple) to the Japanese mastering (pink) shows:
(click on image to see the animated comparison)
The next digital release of Aja came in 1988, when Mobile Fidelity Sound Lab released a new mastering of the album sourced from the original analog masters (a fact that irked Nichols). As the GIF below shows, the MFSL mastering (orange) has more high-end and low-end, but less midrange, than Hoffman’s mastering (purple):
(click on image to see the animated comparison)
In 1993, Steely Dan released the Citizen Steely Dan box set, which featured new masterings of the band’s entire catalog by Glenn Meadows. Meadows’s remasters were sourced from Nichols’s 1982 digital tapes. According to Meadows, he was instructed to “pretend you’ve never heard the albums before and do what you feel appropriate” and given an almost open-ended time frame from Nichols and the band to put together the box.
“Processing included all digital EQ/compression signal processing,” Meadows told me via email. “Also, several functions of CEDAR was used. On the older albums, a light processing of De-Noise was applied. This was used in conjunction with the EQ processing so that when the noise was removed, there was also no loss of hi-end (a typical false accusation about CEDAR — the reality is an overly aggressive application of the process). The final CEDAR process was their ‘Phase Correction’…. This was a subtle improvement, but was obvious when taken in and out of the signal path.”
According to Meadows, his Citizen Steely Dan mastering was approved by Nichols and the band with no changes.
Ultimately, Meadows’s mastering of Aja (which is unfortunately split across two CDs on the box) differed little from Nichols’s ’84 CDs, aside from less tape hiss and a little more treble on a few songs, as can be seen in the GIF below comparing Meadows’s Citizen mastering (dark green) with Nichols’s ’84 mastering (light blue):
(click on image to see the animated comparison)
Ever frustrated with how the Steely Dan catalog sounded on CD, Nichols began remastering the entire catalog yet again in the late-1990s with the intention of creating the “definitive” Steely Dan CD releases. These CDs were once again sourced from his 1982 digital transfers. Interestingly, however, Nichols’s new mastering of Aja (dark blue), which was released in 1999, was nearly identical to Meadows’s Citizen mastering (green), aside from a slight volume difference:
(click on image to see the animated comparison)
Nichols’s 1999 remaster of Aja remained the last word on the album until 2010, when a new Japanese SACD featuring a “flat transfe[r] from Japan[ese] original analogue master tapes” by Hitoshi Takiguchi was released. From the release’s description, it’s not clear whether the tape used for this release was the U.S. master tape, a Japanese copy of the master tape, or even an EQ’d “cutting master.” But Takiguchi’s mastering (red) differs from all previous masterings, being slightly less bass-centric than Hoffman’s (purple), but not nearly as bright as the Nichols or Meadows masterings, as midrange focused as the previous Japanese mastering, or as “scooped” in the midrange as the MFSL mastering:
(click on image to see the animated comparison)
Evaluating the different Aja masterings is easily the most difficult task I’ve tackled in a TBVO column so far.
None can be easily dismissed based on a lack of dynamics. Whether measured by crest factor DR score or R128 dynamic range, all of the masterings have very similar dynamics, registering a DR score between 13 and 15 and an R128 between 6.5 and 7.9 dB. The 1999 Nichols remaster is ever so slightly less dynamic according to DR score (12), but slightly more dynamic according to R128 (8.0 dB). Likewise, the Takiguchi mastering is somewhat more dynamic than the 1999 Nichols remaster (and in line with the previous masterings) according to DR score (13), but is the least dynamic according to R128 (5.0 dB).
Ultimately, none of the masterings can be considered overly compressed or “brickwalled.” A comparison of the waveforms of the Hoffman mastering (purple) and the Takiguchi mastering (red) shows that the latter is slightly louder and has ever so slightly shorter peaks, but is far from compressed2:
(click on image to see the animated comparison)
There’s a better case for throwing out some of the masterings due to technical flaws.
Overall fidelity is not the issue, despite the constant kerfuffle over whether the analog masters or Nichols’s digital transfers should be used, debates about noise reduction, and uncertainty about what tape was used for the Japanese remaster. While the Citizen Steely Dan box and Nichols’s 1999 remaster have less tape noise on songs like “Home at Last” (and Meadows is correct that the noise reduction is mostly tasteful), the noise level isn’t an issue on any of the CDs sourced from the analog tapes. With very close listening on the most resolving gear, there are small differences in clarity between the various versions that seem to go beyond fact that each mastering’s EQ emphasizes different details.3 I’ll get to these differences later, but overall the level of clarity on each CD clears the bar to keep them in contention.
Some masterings, however, possess notable glitches. Masterings sourced from Nichols’s 1982 digital transfer — including Nichols’s original ‘84 mastering, the Citizen Steely Dan box, and Nichols’s ‘99 remaster — have several bursts of noise in the left channel near the 2:18 mark of “I Got the News.” There’s also a click in the left channel near the 1:03 mark of “Black Cow.” These don’t appear in any of the versions sourced from analog tapes, including the Hoffman, ’84 Japan, MoFi, and Takiguchi masterings. The Mobile Fidelity CD has its own glitch, too. Near the 2:10 mark of the title track, the sound stage shifts slightly. This is more noticeable on headphones than speakers, and it might just be a case of one of the many edits in the song being more apparent on the MFSL than on other masterings.
In my view, the “I Got the News” error is reason enough to remove both Nichols masterings and the Citizen Steely Dan mastering from the running. But there are also other reasons to do so that raise fundamental questions about what makes for a definitive mastering, questions that thus far haven’t been explicitly addressed in TBVO: Is the best mastering the one that (subjectively) sounds best to the listener, the one that reflects the sound of the master tape, the one that captures the intent of the artist, or the one that sounds like what listeners heard when it was first released?
I would argue that, by necessity, the first question trumps the others, but doesn’t render them obsolete. We rarely know the sound of the master tape or the intent of the artist, and it’s sometimes the case that what listeners heard when an album was first released was compromised in one way or another, either due to technical limitations or errors. But when information about the master tape or artist intent is knowable and when the album’s initial release wasn’t marred by poor production or mastering, it’s important to all of those factors take those factors into account, even if we filter them through the somewhat subjective lens of what sounds best to our ears today.
In the case of Aja, the album’s initial release was regarded as an audiophile masterpiece, and we know a great deal about the band’s intent and a fair amount about the sound of the master tape.
If one’s ideal mastering is a pure reflection of the master tape, then it’s likely that the Mobile Fidelity Sound Lab mastering of Aja is the best choice. The MFSL’s slight midrange dip comports with Hoffman’s description of the master tape and Mobile Fidelity’s reputation for putting out (mostly) flat transfers of the analog master.
However, Hoffman is right that the album sounds more natural and balanced with some of that missing midrange added back. Moreover, we know that the MFSL’s presentation of Aja, even if it matches the master tape, isn’t how Aja sounded to listeners in 1977 — a sound that made Aja a go-to audiophile test record.
The original ’77 release of Aja was mastered by Bernie Grundman. I reached out to Grundman, and he explained that Becker and Fagen didn’t simply sign off on his mastering. Rather, Nichols, Becker, and Fagen actively participated “in the mastering session with [Grundman] and [Grundman, Nichols, Becker, and Fagen] did the album together.”
Given that we know that the initial mastering of Aja both reflected the band’s artistic intent and made Aja an audiophile landmark, it’s worth investigating how the digital masterings compare to the original vinyl Aja.
I was able track down three high resolution digital transfers of original Aja pressings attributed to Grudman. Two of the three — an ABC AB 1006 Santa Maria pressing and a Japanese YX-8114-AB pressing — featured EQ that was remarkably similar according to Har-Bal. They also sounded virtually identical, despite the inherent variation in vinyl pressings and the number of variables that can skew how vinyl sounds, especially in the process of transferring it to digital. Those pressings also happen to be two of vinyl audiophiles’ favorites.
Comparing both of Nichols’s masterings and the Citizen Steely Dan mastering — all of which, as noted above, feature very similar EQing — to the vinyl, it’s clear that they’re much brighter sounding than the original mastering that Grundman worked on with Nichols, Becker, and Fagen, as this GIF of the ’84 Nichols mastering (light blue) versus the vinyl (grey) illustrates:
(click on image to see the animated comparison)
The ’84 Japanese mastering (pink) is much closer to the vinyl (grey), though it still displays a more midrange-focused, bass-light tonality that’s even more apparent with close listening:
(click on image to see the animated comparison)
Hoffman’s mastering is extremely close to the vinyl on several tracks and more bass-heavy on others:
(click on image to see the animated comparison)
Finally, the 2010 Takiguchi mastering (red) is almost identical to the vinyl (grey), further hinting at the possibility that the Japanese tape is a vinyl cutting master:
(click on image to see the animated comparison)
Taking both the Har-Bal analysis and close listening into account, the Takiguchi and Hoffman CDs are closest to the original vinyl’s EQ, with the Takiguchi being slightly, but notably, closer on several tracks. One of the most significant differences, for example, comes on “Deacon Blues.” Hoffman’s master has more low-end, which tends to make it sound slightly bloated on some gear, relative to the original vinyl or the Takiguchi mastering.
However, the slight, but undeniable difference in fidelity mentioned earlier favors the Hoffman mastering.4 Likely due to Hoffman’s use of the original U.S. master tape, as opposed to the Japanese tape copy, the Hoffman mastering sounds clearer and more three-dimensional when played on resolving gear. Gadd’s floor tom strikes in the right channel of “Aja,” to cite one example, sound more resonant and lifelike on the Hoffman mastering. The same is true for Rainey’s bass on “Peg,” despite the fact that the song’s equalization on the Hoffman and Takiguchi CDs is nearly identical.
The crown in the third installment of TBVO, then, is a near-tie between the Hoffman mastering of Aja and the Takiguchi mastering of Aja. A slight EQ edge goes to Takiguchi, but the fidelity edge goes to Hoffman, giving the Hoffman the overall win by a small margin.
If you want the absolute best digital mastering of Aja, try to get your hands on a Hoffman CD (or one of its “clones”)5. But the recent Japanese CDs and SACDs with the Takiguchi mastering are also an excellent choice.
Whichever version of Aja you pick up, our grotesquely cartoonish historical moment is the perfect time to listen to some Steely Dan (preferably on a nice set of speakers or pair of headphones). As Dan Moffett put it in his meditation on how he became a Dan-o-phile, “The more absurd life gets, the more sense Steely Dan makes sense.”
UPDATE FEBRUARY 25, 2019
Thanks to @EmmettM , who mailed the disc to me, I was able to hear the 2018 MQA CD of Aja. While I remain dubious about MQA as a format, I’m happy to report that the mastering on the disc is excellent.
The MQA CD is billed as a “flat transfe[r] from US original analogue master tapes by Eli Brown at Universal Music Studios, LA, in 2018.” Its fidelity seems to bear out this claim. In terms of detail, depth, and noise floor, the MQA disc is closer to the Hoffman mastering than the Takiguchi mastering.
EQ-wise, the MQA CD is very slightly bass-shy relative to the Hoffman and Takiguchi masterings on a few tracks. But, overall, it’s very close to both of those masterings, as well as the original vinyl, as the GIF below — which rotates through several tracks on the MQA CD (faint pink), the Hoffman CD (purple), the Takiguchi CD (red), and the original vinyl (grey) — shows:
All things considered, I’d now move the MQA CD into second place behind the Hoffman CD but slightly ahead of the Takiguchi mastering, thanks to the MQA CD’s slightly superior fidelity.
UPDATE OCTOBER 23, 2021
As if the story of Steely Dan’s Aja on CD couldn’t get more complicated, some eagle-eyed fans – “Paul P” and “strippies” – over at the Steve Hoffman Forums (SHF) have sussed out some new wrinkles in the saga. (Start reading here if you’d like to follow the conversation yourself.)
Before we dive into the new information, let’s back up.
According to the sources cited in my TBVO, both Roger Nichols’s first mastering of Aja and Steve Hoffman’s only mastering of Aja made their way onto CDs in the 1980s. Since this information came from credible contemporary sources and has been widely repeated in the years since, I took it for granted in writing my TBVO.
However, the aforementioned SHF members have raised an intriguing third possibility: That what I’ve identified in my TBVO as the ‘80s Nichols mastering is, in fact, an early-‘90s mastering by Glenn Meadows. This supposition is based on something Meadows wrote that I overlooked. Let’s review some of what Meadows said:
The story actually starts at the original transfer of the analog masters to CD. At that time, the CD masters were initially prepped/mastered by Bob Ludwig at Masterdisk in NYC. Many days were spent matching master tapes to different brand analog ¼” machines to get the best transfer possible….It was also realized at that point, that the analog tapes were in VERY poor condition due to continuous use for cutting master lacquers to keep pressing records.
At the end of the mastering, it was also decided to do a 1 pass FLAT transfer of the analog masters (after each album was mastered for CD), would be made to the 3M 4Track digital recorders. This was in the VERY early days of Digital recording. The masters that were created for CD were done with the Sony PCM 1600 system….
Move forward several years…[and] a request for a new set of CD masters was sent to the MCA studios. Not realizing that there were NO copies of the Ludwig created masters in their files, the studios pulled the 15IPS Dolby EQ tape copies, and transferred THOSE tapes to the Sony digital tapes, and sent those to the US plant(s) for replication. Nobody ever compared the US made CD’s to the original CD’s, since it was assumed they were all the same.
Fast forward to mid 1990’s. Roger Nichools is now living in Nashville, and I (Glenn Meadows) have been working with him mastering projects for him. MCA decided to add 4 cuts to the then 8 cut Steely Dan Gold CD. Roger asked for the master tapes be sent to Nashville so that he and I could master the 4 additional cuts and create new CD masters. When the tapes arrived, they were a stack of ¼” analog Dolby copies (see above), as well as the 1 1630 master that was the 8 cut Solid Gold album. Roger realized that they had sent the wrong original album masters had been sent.He made some phone calls and told the MCA vault that they needed to find Sony Digital masters made at Masterdisk, and that they were the correct CD master source tapes. What he subsequently found out, was that the wrong tapes had been used YEARS earlier when production switched to the US. Panic ensued, and Steely Dan management threatened to sue, and MCA stopped all CD manufacture.
Roger was able to locate the FLAT 3m 4 track masters that were recorded when the fist mastering was done for CD release. Those were sent to me in Nashville, where there is/was still a working set of 3M digital recorders. It was at that point that the re-mastering that later became the box set occurred. All 7 of the CD’s were remastered at that time to update ALL CD releases. Transfers were made from the 3M system for mastering…. Processing included all digital EQ/Compression signal processing. Also, several functions of CEDAR was used. On the older albums, a light processing of De-Noise was applied. This was used in conjunction with the EQ processing so that when the noise was removed, there was also no loss of hi-end (a typical false accusation about CEDAR, the reality is an overly aggressive application of the process)….. That was how all 7 albums were re-mastered, and re-released. The ONLY notation was a sticker that said “Recently Re-Mastered by Artist”….
The [Citizen] box set was simply a re-edit adjusting sequence to match that provided by MCA to fit on 4 CD’s. Since we were about a year and a half later, there were some improvements in the CEDAR system, and there were several tracks that I was unable to completely remove the tape his without having resulting “space monkey” artifacts. With the later version of CEDAR, I was able to successfully remove the last vestiges of the tape hiss.
So, according to Meadows, the original ‘80s Steely Dan CDs were mastered by Ludwig, presumably in conjunction with Nichols. (I’ve made an edit to my TBVO to reflect this.) These CDs were based on new EQ’d masters recorded to Sony PCM 1600 digital tapes and sent to CD pressing plants. Subsequent CD remasters – including the Meadows-mastered Citizen box and Nichols’s 1999-2000 remasters – were based on a flat-transfer of the analog master to 3M digital tape. (I’ve clarified this distinction in my TBVO, too.)
The key claim that I missed in Meadows’s account, which the SHF folks noted, was the story about early-‘90s CD remasters by Meadows, which were only identifiable by a “Recently Re-Mastered by Artist” sticker. Further, the subsequent Citizen box was identical to these CDs, according to Meadows, save for an additional round of noise reduction.
Why does this matter?
It means that what I identified as the ‘80s Nichols mastering may, in fact, be the pre-Citizen Meadows remaster. Like the Citizen box and Nichols’s ’99 remaster, it clearly comes from the 3M tape. As noted in my column, all three have the same glitches, which aren’t present on any of the other CDs. So they all originated from the same digital source. I thought that was because the original ‘80s CDs also came from the 3M tapes. But, according to Meadows, that’s not the case. Further, what I identified as the ‘80s Nichols mastering is incredibly close to the Citizen mastering, differing only slightly (usually in the high end) on some songs. Could this be due to the additional CEDAR processing noted by Meadows? Very possibly. Finally, it’s possible that – despite being identified as 1984 on the artwork – the CD I refer to as the Nichols ‘80s mastering was actually released later if, in fact, the label recycled the ‘80s artwork for Meadows early-‘90s remaster and added only a “Recently Re-Mastered by Artist” on the (long gone) cellophane.
If the above is true, where does that leave us?
If what I called Nichols’s ’80s mastering is actually Meadow’s early-‘90s pre-Citizen mastering, it means that the common ‘80s mastering I attributed to Hoffman (in line with his own claim) is either the Hoffman mastering or the ‘80s Nichols (with Ludwig, possibly) mastering. Both could not have been released.
Which is the case? I don’t know. The plot thickens.
1. There are dozens of CD releases of Aja, including some very obscure Japanese and Russian pressings. To my knowledge, all of the existing releases as of this writing feature one of the seven masterings mentioned above. The only mastering excluded from this list is a new MQA CD release. It’s excluded not only because of questions about the fidelity of the MQA format, but also because it’s exceedingly rare. Despite my skepticism about MQA, if I manage to get my hands on a copy of the MQA CD, I’ll update this piece.
2. Audacity was used for waveforms.
3. For the subjective analyses, all editions were ripped with XLD and played Audirvana Plus. The DAC/Amp combo was a Schiit Yggdrasil/Ragnarok stack. Most critical listening was done on a diverse set of headphones: NAD HP50, Sennheiser/Massdrop HD6XX, and Focal Clear.
4. I wondered if the Hoffman’s seeming edge in fidelity had to do with differences in EQ. To test this, I used Har-Bal’s mastering-cloning feature to apply the Hoffman mastering’s EQ to the Takiguchi file. While there are limits to what Har-Bal’s filters can do (the Hoffman master still had more low bass), it was relatively difficult to tell the resulting files apart when comparing them side-by-side in Audacity, and I doubt anyone could detect differences in clarity between the two a blind test.
5. There’s been great debate as to whether “Mastering 3” and “Mastering 6” as defined in these posts is the same as the confirmed Hoffman mastering (“Mastering 7” in the posts). Some of the CDs listed have slightly different peak levels, and the track times on many differ slightly. (On some versions, the end of one track even contains the first moments of the next track.) However, if the audio files are trimmed to contain the exact same parts of the songs, the three different graphs available in Har-Bal all show that masterings three, six, and seven are all Hoffman’s mastering, at least according to the copies of each that I have.
About the Author
Josh Mound has been an audiophile since age 14, when his father played Spirit's "Natures Way" through his Boston Acoustics floorstanders and told Josh to listen closely. Since then, Josh has listened to lots of music, owned lots of gear, and done lots of book learnin'. He's written about music for publications like Filter and Under the Radar and about politics for publications like New Republic, Jacobin, and Dissent. Josh is currently a postdoctoral fellow at the University of Virginia, where he teaches classes on modern U.S. politics and the history of popular music. He lives in Charlottesville, Virginia, with his wife and two cats.