In 1995, my employer, NCR, laid off several thousand people, including three thousand in the Dayton area. Worse yet, they laid off about 90 technical writers, including me. In a market with as small a technology base as Dayton, Ohio, it was almost impossible to compete for the handful of tech writing jobs. At first, NCR still needed me (even though the managers I had been working for weren’t allowed to hire me back). So I spent about two years working from home as a contractor for various NCR divisions. For a time, I actually made more money than I had as an employee. Then I contracted for several more companies, often working at home. One effect this had was that I had to upgrade my computer and software repeatedly. So by 1998, I had the latest Pentium you could buy and above-average memory and hard drive space. I had also upgraded my Cakewalk software again and again, and – oh by the way – my computer was still surrounded by MIDI and audio gear.
This was still all in my dining room, so there was no question of leaving microphones set up like we did in our first house (the one with the relatively dry basement). But the ability to play sequences into Cakewalk whenever I had a few minutes “down-time” was quite handy. My process evolved into something like this:
- Establish the tempo, either by setting the metronome or by copying a drum track from a similar song, or by starting a simple drum track (bass on 1 and 3, snare on 2 and four, or some such).
- Play the whole song through on the weighted keyboard, using the SC-55 to monitor, not worrying too much about mistakes.
- Listen to and visually examine the track to figure out which verse and which chorus came out the cleanest.
- Cleaning up any errors in that verse and chorus, then using “cut and paste” features to use those as the building blocks of the song. Clean-up included:
- Looking for and fixing wrong notes.
- Moving notes that were more than a few milliseconds out of timing back toward where they were supposed to be. Cakewalk’s ability to quantize by percentage was very helpful, as it could snug things up a little without making every note “dead on” and boring.
- Adjusting volumes to make them more consistent. Again, Cakewalk’s Interpolate feature helped me narrow the dynamics of the track without killing them. (For example, you could specify that a track’s velocity go from 0-127 to 55-127. That would bring up the quieter notes while hardly affecting the louder notes, something like the way compressors work.)
- Looking for and fixing wrong notes.
- Once I had a clean chorus and verse part, I would use them as building blocks, cutting and pasting to build out the song. Then I would:
- Program in the bass and drums (both the MT32 and SC-55 had an “acoustic bass” sound that was more convincing than my real bass).
- Revisit the tempo by section, raising or lowering it as needed to support the mood of the song.
- Play in orchestral parts, etc.
- Sometimes I would program in a guitar or banjo part just to have a sense of context for when I got around to playing in the real part. Since I actually played those instruments, I could program in the exact part I played when I performed the song live. No it wasn’t all that convincing, but it helped me plan the “live” part of the arrangement.
- Leave it for a day or a week, and relisten later to hear if there was anything that is now grating or sounds out of place.
By the time I had piano, bass, and drums – at least – in the MIDI sequence, I was ready to add the live instruments on the Multichannel. I could even add an instrument or vocal when I “dumped” the output of the MIDI equipment onto the Dokorder. If I wanted to get crazy, I could bounce things around on the tape to get more than two additional accoustic tracks. I usually didn’t, though. As you can guess, there was more than one time I suffered from “where the heck am I?” syndrome.
To get better mixes, I bought a Sony DAT to use as my mixdown deck. For the most part it worked very well, although I learned that with the SC-55 and the Sony DAT using the same sample rate, certain sounds got a little “confused” as they went to tape, the same way a scanned newspaper photo can look checkerboarded on your computer if you look at it in the wrong resolution.
I don’t believe I have any tapes left over from that period – it was mostly a time of experimentation, squeezed in between billable hours. But one thing I determined; if I was going to keep doing this sort of thing, I needed a tape deck that I could sync to the MIDI sequencer (or vice versa), so I could dump the tape recordings and the MIDI units’ sounds at the same time to the stereo master, instead of wasting tape channels on the synth tracks.
I had been lusting after several commercially-available 8-channel reel to reel decks, but the first one that seemed robust enough and affordable enough for my use was the Fostex R-8, which had been introduced in 1989, but whose price was coming down. And it included the sync capabilities I needed.
I bought one in the early 1990s, after a big price drop. Then a month later, they discontinued the thing and dropped the price again, by about another $400. Ooops. I told someone I was glad I got one before the price went down, and he didn’t realize I was making a joke.
Like my very first Concord tape recorder, the Fostex R-8 got around synchronization issues by combining the record and play function on one tape head. But it also had built-in Dolby-C(r), which rivaled dbx for overall noise reduction and was, frankly, better suited to the R8’s narrow tracks. So the dbx units were no longer necessary. There were multiple ways to sync the thing to my MIDI setup, and I don’t remember which way I chose, but once I was using it that way, the synch was perfect. In fact the machine was perfect. It was ridiculously easy to use and sounded great.
I had Wesley over to record violin tracks on a bunch of songs that I’d never had room to put violin tracks on before. I used my harmonizer to add harmony tracks to songs. I added real flute, sax, and harmonica to songs where I had only sequenced those instruments before. I was in “Hog heaven.”
The following drawing should give you the general idea of how these things went together. I left out my harmonizer, digital delay, and multiple compressors and preamps for simplicity. I also showed using a separate mixer for monitoring and for recording the live instruments. If you have a mixer with lots of monitoring capabilities, you won’t need to do that. In fact, I didn’t do that, either, but that’s because I was feeding the microphones into the R-8 through outboard preamps and monitoring at the end of the signal chain (represented by the headphones plugged into the DAT). I also left out other tone generators I had picked up – half-rack piano and drum units from Roland. (I picked them up because drums and piano usually used the most channels on my SC-55, and these extra units kept me from running out of channels.) Even “simplified,” this is a complicated little chart. I hope this makes sense. The gray lines represent MIDI cables; the black lines represent audio cables.
Then Cakewalk, who had migrated their software to windows some years back announced that they were looking for beta testers for a new product that would fully integrate digital recordings into their system. (You could add wave files to their sequences before, but it was not, er, user-friendly).
Digital recording had been around for some time, of course. (CDs had been introduced in 1982 and ProTools’ predecessor “Sound Designer” was introduced in 1984.) But PCs were just becoming powerful enough to use this technology reliably.
I signed up and got several versions of what would be “Cakewalk Pro Audio” over the next several months. It ran very slowly on my computer at the time, but I had patience. Each version was more useful and robust than the next. Since I was at a point where setting up microphones and capturing fresh tracks was not convenient, I practiced on tracks that I “dumped” from the Fostex.
Before long I had dumped all of the tracks that Wesley and I had recorded to my computer. I discovered that it was ridiculously easy to cut out microphone pops, wrong notes, etc. on the computer. After I upgraded my PC again (in part to handle graphic and desktop publishing programs I was using for work), the program ran much more smoothly. I was soon experimenting with comping my vocal parts, singing the song three or four times and copying the best parts of each take to a separate track.
Frankly, it got to a point where EVERY track I had recorded to the R-8 was now on my hard drive, and I found myself using it less and less. Eventually I sold it. Yes, I miss it sometimes, but I had kids in college and wasn’t using it at all.
The following picture shows what it was like when I took the R-8 out of the “studio.” Like the picture above, I’ve left out ancillary gear and extra tone generators for simplicity. I was feeding the mics directly through preamps into the audio inputs on a fancy sound card I had bought. I was feeding the audio output from Cakewalk through the audio outputs on that card to the mixer. But at at that time, I was still using my tone generators for all of the instruments. So the MIDI output from Cakewalk was feeding the synths. Then the audio output of the synths was feeding the mixer.
For a time I had a sound card that had several audio outputs, so I could mix even the audio from the PC on my hard drive on the mixer. Then when I replaced my PC, the old sound card wouldn’t work with it, and the manufacturer never made a replacement. So I had to mix the audio tracks on the PC. However, I could have virtually unlimited number of audio tracks, as long as i didn’t overload my PC. I usually kept it to 12 or so audio tracks, though, and made a point of “muting” tracks I didn’t really need while I was editing other things.
Then, we reorganized many things around the house, and almost all of my gear went into storage in the garage. About the same time Cakewalk started providing better “virtual synths” in its package, and I got a computer fast enough to use them. So, for just editing tracks, I started using the virtual synths instead of my physical tone generators. Within several releases of Cakewalk/Sonar, and a number of additional patch purchase, I was able to get sounds on my PC that were nearly as good as the sounds from my outboard gear. Some of the sounds were, frankly better. And there were a lot of patches, (like hammered dulcimer) that my outboard gear never provided.
I loaned out my SC-55 to a friend who used it in his studio for a while, then realized he wasn’t using it any more either and gave it back to me. But by then, I had better piano patches on my PC than the SC-55 or my other tone generators anyway.
I haven’t set up my full studio in years. As of this writing, we’re moving, and I’m hoping to be able to set the studio up again in one of the bays of the new garage. But, ironically, I’m not missing these toys as much as you might think. As long as I can hook up a keyboard controller and a mic or two once in a while, I have most of what I need right on the hard drive of my PC.
I may add another chapter when I get a chance to get my studio set up again, but one thing I can guarantee is that it won’t look like any of the previous iterations.
Best of luck. Please use the contact page if you have any questions about anything on these pages. Or disagreements, or corrections, or faint praise, or whatever.
And above all, enjoy your music!
– Paul