I just got back from the 1998 AES show that was held in San Francisco. As I meandered from display to display ogling over all of the new digital recorders, digital consoles, and digital workstations, I often heard someone whining about not having enough features on their whiz bang digital system. The audio industry has given them wings with which to soar, and they want to soar higher. Well I remember the days when I had to walk 40 miles to school in snow up to my neck, and it was uphill both ways.

For just a minute I am going to flash back (it’s not just Viet Nam vets that have horrible flash backs) to the days long ago, before digital recording, when the reigning tape machines were the 24 track and 16 track recorder that used two inch wide analog tape recorded at 15 or 30 inches per second. Performing overdubs on these machines went well, unless you wanted to record multiple passes of, let’s say, a vocal, and then combine them to make a master track. When you copy a vocal from one track to another on a piece of analog tape, you loose the transients and add additional tape noise and harmonic distortion to the signal. This is not acceptable if you are striving for the best possible sound quality.

If you will not accept the generation loss, then you only had two choices. The first choice was to "punch in" on the original track if you wanted to change anything in the vocal. There was a very high "pucker factor" if the performance was very good, but the vocalist insisted on an attempt at a better performance. You would roll the tape, punch in on the vocal line to be replaced, and… whoops, it wasn’t as good. Too late! The only thing left to do was try the punch over and over until the vocalist got a take that was up to par with the one you erased. There was no "undo" in analog recording.

To make matters even worse, there was no "auto-punch" on analog machines. You had to perform the punch-in and punch-out over and over each time manually, without making a mistake. Analog punches were not clean, perfect punches, either. Analog machines would erase a little spot where you punched in and leave a little hole where you punched out, so you had to be good at it to replace short vocal phrases, or even words, on the master vocal track. If you screwed up, it was like trimming your sideburns. Each time you missed the punch you would have to punch in a little earlier or punch out a little later to fix the botched attempt. The "punch on the lead vocal track" was the method we used for Steely Dan vocals, and more than once my sideburns met at the top of my head.

The second method was to record multiple passes on the tape, and then leave them where they were, without bouncing, and select between the tracks during the mix. This is the way we did guitar and sax solos. Let’s say that we just recorded three great guitar solos on three separate tracks. After microscopic appraisal, we would decide on the pieces of each track that would comprise the final solo. We would then erase all of the pieces that were not going to be used, so that during routine playbacks we would hear a complete solo without having to make switches on the console during each playback. There was only one problem with this method--- you very quickly ran out of tracks. I must remind you that Murphy’s law applied to analog recording as well, and you always needed one more track than you actually had available.

On one occasion, we had a guitar part that spanned three or four tracks as outlined above, and we needed to overdub a saxophone over the same section of the song. We had no more open tracks. We ended up recording in the holes on the guitar tracks. The first few bars of the sax on one track, the next few bars on another track, and so on. Talk about pucker factor. If the sax player was playing a line that would go past the start of the guitar, you would have to punch out regardless, to save the master guitar track. I, uh, uh, I can’t go on.

Sharpie Automation

There was no such thing as console automation. Mixing was done by a team. We used the "three men to a knob" technique. One guy set it, the second guy checked it, and the third guy changed it when you weren’t looking. Automation by Sharpie. Thousands of little pieces of tape with marks on them. Marks for intro levels, marks for verse levels, marks for chorus levels, marks to remind you which marks to use… If the mix was simple enough, you could make it all the way through by changing the knob positions when you got to each section of the song.

Each guy would have a few knobs to deal with. They wouldn’t always be the ones right next to each other, either. If there was one guy in the middle of the console working on vocals throughout the mix, another guy could move the drums up for the chorus, then run around to the other side of the console to catch a move on a guitar, and then run back around to change the drums at the next verse. The producer usually became the choreographer during the mix. If there were any mistakes, you would try the mix again. Sometimes you would record a bunch of mix attempts on the two-track machine, and then later listen to the playback and decide that you liked some things about each attempt. Out came the razor blade and editing block, and the pieces of each mix were edited together to form the final master.

Sometimes the mix was too complex to do with the above method. Then what? Well, we would set all of the levels and effects for the intro, and mix the intro onto the two-track machine. We would then set the levels for the first verse, and mix the verse to the two-track machine. Before we went any further, we would edit the intro onto the verse and make sure everything was ok. If not, we would make a correction, put it on the two-track and edit it on again. After the intro/ first verse was satisfactory, we would mix the first chorus and edit it onto our master mix. Each time we would check the edit and not move on until the transitions were seamless. Sometimes we would have to edit one bar pieces to get a solo or a transition just right. By the time we got finished with the mix, it looked like the master two track tape had more edits than the Monica Lewinsky phone tapes.

Early Automation

The earliest console automation was the Allison Engineering 64k automation system. Forget about moving faders, it was VCA based. A complex signal was recorded on the analog tape that contained the position data of each fader on the console during each automation frame. To update the data you had to playback the information on one track and record the modified data on another track. This left you with a 22-track machine, and you were asking for trouble if you recorded a vocal next to the automation track because of the crosstalk inherent in analog tape machines. If you recorded the two automation tracks next to each other you could end up with unusable garbage for automation.

The first moving fader automation system was Neve’s Necam automation system. The mixes were stored on 8" floppy disks and you could do some offline editing and merging of mixes. The faders weren’t fast. As a matter of fact, studios that still have this system call them "crawling faders." It did sound better than VCA based automation, and if you wanted to change something, you just grabbed the knob and the old fader data was instantly updated.



Finally

Ok, so enough of the whining, already. No matter which hard disk recorder, or eight track disk/ tape recorder, or digital audio workstation you buy, remember that it will be thousands of times better than anything we used to make all of the Steely Dan records. If we had this new stuff back then, we would have really produced some good records. And when you are frustrated about performing all of those edits with a mouse while staring at a teeny tiny video monitor, think about what you might have done after hundreds of edits to a piece of analog tape with a nice sharp razor blade so closely poised near your WRISTS! Someday I’ll show you the scars.