A Steely Dan horror story

Denny Dias recounts the "absolute nightmare" that still haunts Katy Lied.

The freak erasure of “The Second Arrangement” was a nightmare come to life for Walter Becker, Donald Fagen, and their loyal studio corps. But arguably the most bone-chilling tale of recording misfortune in Steely Dan’s history took place a few years earlier, during the making of Katy Lied. A technical calamity befell the band that still haunts the sound of the group’s 1975 album.

Guitarist Denny Dias, writing for an early Steely Dan website, recounted the details of the dread incident, which may or may not have involved a mysterious mist straight out of John Carpenter’s The Fog. “Something happened during the remix of ‘Doctor Wu’ that scared the hell out of us,” Dias recollects in the essay, reproduced here with the author’s permission. It’s the kind of spooky story that will make any audiophile wonder whether ghosts hang around recording studios too.

Happy Halloween!

---

Katy and the Gremlin

By Denny Dias

She is lovely yes she’s sly, but we were ordinary guys. I still have a 30ips safety (copy made from the master tape) of Katy Lied that Roger Nichols signed in blood! Yes, Katy was a bug. She got to all of us. I remember trying to master the record by myself because no one else wanted to work on it any more. But I’m getting way ahead of myself. 

In 1975 we had great expectations and lots of enthusiasm. We had excellent musicians ready to perform in a state-of-the-art recording studio. We had our radical super hi-fi monitor system that consisted of electromagnetic flat panel Magneplanar speakers with three amplifiers and two subwoofers and active crossover tuned to the room with a real-time analyzer. They sounded great. The songs were great. The musicians were grateful. What could go wrong? Well, things happened. Some could be attributed to human error. Others could be blamed on mechanical failure. The rest will never be explained. 

Anyone who watches old science fiction movies knows that strange things start to happen when you encounter a mysterious mist. I am thinking of the day that the steam generator went berserk. It was supposed to keep the air in the studio at a perfect 50 percent humidity, but on this day it felt more like Biscayne Bay. The air was so thick you could cut it with a knife. The glass was foggy. Everything was damp, including the sound. The drums actually sounded like they were soaking wet even though they had been recorded on a normal day. It was almost funny except that we couldn’t work that day. We all went home, and the steam generator was soon fitted with a new control unit. The studio was given a clean bill of health, but I can’t help thinking that there was some unseen oxidation that caused the studio and even the tapes themselves to fester. 

The actual recording went fairly well for the most part. There was, of course, the tambourine fiasco when Roger accidentally erased part of Victor Feldman’s track on “Rose Darling” (Roger had never done anything silly like that before or since), but the worse mistake didn’t become apparent until much later when it was time to transfer the record to vinyl. 

Mixing was an absolute nightmare. Every song was mixed at least twice, and not because we were being fussy. In fact, we had mixed the entire record before we realized that there was a problem. We were using the new dbx noise reduction system, which was supposed to give us a better signal-to-noise ratio than Dolby, and for some reason the dbx units could no longer decode the mixes on tape. They sounded dull and lifeless, and no one could explain why. After all, all of the equipment had been properly aligned for each session. This was especially puzzling since each mix was played back immediately upon completion. How could the sound deteriorate so quickly? Even if there had been some awful mistake, it couldn’t have happened the same way twice and certainly not more than twice. 

Several of us formed a contingent to storm dbx headquarters. We packed up the tapes and the dbx units, and Gary and Roger (and one or two others) boarded a plane to the East Coast. They confronted dbx and discovered that no one could fix it or explain it. The people at dbx built us a special pair of units with adjusting knobs that could alter the settings that are normally sealed inside at the factory. This too was a miserable failure. Could the tapes have been exposed to gamma rays? Why didn’t any one else using that studio have a problem? And why only the two-track mixes? The two-inch 24-track masters were still sounding good, so we decided to remix the entire record using Dolby. 

I just dusted off my copy of the first mixes and gave them another listen. They still sound quite dull as expected, but I wanted to see if there was anything lost in the remix. Let me assure you that the new mixes are better in every case. I’ve heard some people describe the mixing process as a “thankless task,” but I think it’s more like a performance. It’s done with feeling, and depending on the mood of the day the result can vary quite a bit. Here are just a few differences that can be described with words: 

• We were so impressed with the performance of Phil Woods on “Doctor Wu” that when it came time to fade out at the end of the song we couldn’t fade Phil. In the first mix, everything fades out except the saxophone! 

• On “Daddy Don’t Live in That New York City No More,” there is no special effect on the lead vocal. For the remix Roger implemented a manual phasing technique that required him to stand near a tape machine and slowly turn a dial by hand. 

• On “Bad Sneakers” there is a drum bash that happens at least a half dozen times during the guitar solo. In the first mix it sounds kind of ordinary, but in the final mix Roger found a compressor that he could set to make it sound really special. 

Something happened during the remix of “Doctor Wu” that scared the hell out of us. I mentioned that we were impressed with the performance of Phil Woods, so you can imagine how we felt when his saxophone suddenly sounded dull and lifeless! This required immediate investigation. The 24-track master was encoded with dbx so there was big tension while Roger did some troubleshooting. When he cleaned the heads on the tape machine the sound cleared up for a while. Then it got dull again. It seems that the tape head developed an irregularity on track 17 and was scraping bits of oxide off the tape! We decided to keep working on the mix, but we would avoid playing any part of the tape containing that saxophone. When everything else was ready, Roger cleaned the head once more, and we recorded the mix on two-track tape. 

Then it came time to make master disks. Records that are re-released with the indication “direct from the master tape” are generally disappointing because here is the last chance for the people who made the record to correct any mistakes made during mixing. A little adjusting here and there can really make a record sparkle. “Katy Lied” needed more than a little adjustment. I remember the first time we brought it to Allen Zentz’s mastering facility, it became obvious that there were things on the tape that couldn’t be transferred to vinyl. Roger said, “I’m getting out of here,” and left before we knew what the problem was. Apparently, his use of condenser microphones in close proximity to the cymbals required too much acceleration for the needle to track. 

The mastering process now became a desperate attempt to produce a vinyl disk that could be played on an average phonograph. We moved to Kendun Recorders, where they had more processing equipment. The speakers there were foreign to us, and all we could do was look at each other and shrug. The engineer suggested some compression, we shrugged and when we brought the disk to a more familiar setting it was awful. As I said, the job was left to me because no one else wanted to work on it any more. I brought my own speakers to Kendun and, after a week of disappointing attempts, Walter stepped in to produce the final release version. It was still disappointing, but it sounded better on more sound systems, so it was the better choice. 

At this point it might be a good idea to read the back cover of the original record. There are some comments about bandwidth and transient response that should have new meaning now. However, the music is still on the tape, and the tape is well preserved. The sound of the digital CD version on Citizen is better than any vinyl by far. It’s interesting that after all these years there is finally a released version that sounds good. 

The 24-track master tapes may not be doing as well. There is a song called “Mister Sam” that didn’t make it onto Katy Lied. The track was good, and Donald recorded a very nice lead vocal performance, so, when we began recording the next record, we thought that perhaps we could include “Mister Sam.” When we placed the two-inch master tape on the 24-track machine, it sounded dull and lifeless and could not be used at all. No one wanted to check the other master tapes, but I am certain that (to paraphrase the lyric) “Mister Sam” don’t sound so good no more.

---