While I was dabbling in multichannel analog recording, another revolution had started, based around keyboard innovations. People had been programming electronic equipment to play or even invent tunes for decades, but that sort of thing was viewed as a novelty until rock musicians started “pushing the envelope.” Rock wizards like Todd Rundgren, Rick Wakeman, Steve Winwood, Alan Parsons, and even Pete Townsend were pioneering new uses of electronic keyboards and related technology throughout the 1970s and 1980s. And some of the things they invented for their private use influenced or became commercial products. (Roger Powell’s Texture software being a prominent example.)
Most of the early sequencers started as custom hardware devices that would play a series of preprogrammed notes, but they expanded to software programs that could run on early microprocessors being built by companies like the Altair 880, the Commodore 64, and the Atari 800.
Computer-based sequencers allowed bands to add other keyboard parts to their performances and record albums. But for a time, equipment from one manufacturer couldn’t communicate with equipment from other manufacturer. The Musical Instrument Digital Interface (MIDI) coalition changed that by specifying a binary communications protocol. MIDI signals were transmitted by wires with a kind of 5-pin plugs that were also used on some computers and old PA systems, so the cords were available from the start.
Most MIDI-enabled devices had three MIDI jacks on the back. The IN jack received MIDI information from another device, such as a controller or computer interface. The THRU jack simply passed any information coming into the IN jack to the next device, allowing “daisy chaining.” The OUT jack sent any information that this device originated.
At first, the only thing one computer or synthesizer could communicate to another a stream of ones and zeroes that represented which notes to play, how long to play them, how loud to play them, which patch (sound) # to use, and any controller information generated by non-keyboard input devices like sustain pedals or pitch wheels. Later, the “MIDI spec” was expanded to allow other information to be downloaded.
As a person who had already come close to maxing out the capabilities of his four-channel studio, and who had bought an 8-bit PC clone for his home office, I couldn’t help wondering what it would take to add sequencing to my own bag of tricks. At the time, very few sequencing programs were made for the IBM PC, though. I bought a Roland MPU-401 MIDI adapter card for my keyboard and started shopping for software.
I spent way too much money on a Roland software product (MESA) that was sold on copy-protected 5″ floppy disks. You could put the software on your hard drive, but it still wouldn’t operate unless you had the source disk in the floppy, unless you wanted to rewrite your config.sys file every time you ran the program.
In those days, PC clones didn’t have sounds built in, so I also spent way too much money on a sound card that sounded – to use the words guitarist Leo Kottke once used to describe his singing voice – like “geese farts on a muggy day.
Looking for something that sounded better than my sound card (but not being able to afford anything new), I found a used Yamaha FB01. That was a half-rack FM “tone generator,” which had a nice array of built-in sounds and was actually capable of producing more than one sound at a time if you didn’t demand too much of it.
FM synthesis was a way to generate tones by starting with a single wave, then adding additional waveforms to it to change up the sound. The “Tine Piano” sound of “Something’s Telling Me it Might Be You” in the Tootsie soundtrack is a good example, as is the DX7’s tine piano sound on most of Michael W. Smith’s early work. Unlike the earlier DX7, the FB01 actually had less processing power per note, but Yamaha’s engineers had had more time to design patches (sounds), so it actually had some nice tones.
In addition, I was looking for a keyboard I could afford that I could use to input piano parts – my old Casio first-generation piano did not have a MIDI output. I stumbled across a Casio MT-540 at a Meijer grocery store and was amazed at the quality of its sound for the price, which was still more than I could afford.
But when I realized that the smaller, much more affordable MT240 had the same tones built in and I saw one for half-price at an after-Christmas sale, I pounced. Neither keyboard had velocity-sensitive keys, but it was still far better than nothing. Both keyboards used what Casio called PCM sounds, which included samples of real instruments modified by adding other wave forms and other means. Through real speakers (not the built-in ones), the piano sound was much more realistic than the FM tones I’d been using so far. (Here’s an irony – while most of the equipment on this page has been sold off or given away, I still have this. When I upgraded to better keyboards and tone generators, I used to let the kids play with it, but it’s still in great shape.)
I could mix the output of the FBO1, the Casio MT240, and the least objectionable sounds of the sound card onto a cassette tape or two tracks of my 4-channel and get something that sounded like music. And I went to town, sequencing what I envisioned as background tracks for a dozen songs that I was performing live at most of my concerts.
The following drawing shows an example of a simple MIDI studio using the components I’ve described so far. to make the cable runs less confusing, the MIDI lines (which carry only digital data, no sounds or music) are shown in gray.
Of course the copy-protected Mesa disk failed after a few months, and Roland told me the product was unsupported so I was “out of luck,” and many hours’ worth of work was lost. Since then I have refused to put software with any kind of copy-protection scheme on my computers. (I make a sort of exception for Finale, since they’re good about helping if an HD goes south with your software on it.)
This was a pre-windows, CGA environment, mind you. So I shopped around again. Somewhere, I came across a disk full of demo software. I tried a bunch of those, plus I was subscribing to Keyboard and EQ magazines at the time, and eagerly devouring any reviews I could find.
By then, a number of software sequencing programs that had originally been developed for other platforms had been ported to the PC. So in addition to trying the demos on the disk, I downloaded all the demos of those from the product bulletin boards (at 56K speeds – this was pre-WWW, too). I liked “Texture,” Roger Powell’s creation, for creating walls and waves of tone, but my primary interest was in putting realistic bass, drum, piano, and string tracks on my computer. I played with the demos to narrow down the list of sequencers that were suited for that. Of the remaining choices, only one did NOT rely on any sort of dongle or copy protection – Cakewalk, an MS-DOS sequencer made by Greg Hendershott for the startup Twelve Tone Systems.
Even in its earliest MS-DOS incarnation, Cakewalk was a remarkably powerful little tool. One view listed all of the events, including note and note duration, pedal events, etc., and allowed you to skootch them around in time or adjust the volume if your original performance was sloppy. Another view allowed you to copy blocks of measures around, so if you had a complex part that would be repeated many times, you only had to get it right once. And you could manually program it if you never COULD get it right. With such features, it was far faster to use than the old Roland software had been anyway.
The first song I sequenced was a Contemporary Christian song (accidentally) inspired by a Who song. I was thinking about “Won’t Get Fooled Again” and came up with a sequence that I thought was reminiscent of the keyboard sound of that tune, without actually copying it. Later I realized that I HAD copied the Who, more or less – the song “Baba O’Reilly,” better known as “Teenage Wasteland.” Pete Townsend’s programming was far better than mine, by the way, and he had to use weird old gear to do it. I actually performed the whole song as a technopop version several times when technopop was “in.” Later, when I was going more acoustic piano and guitar-based, I left the programmed little riffs in place. If you don’t like CCM, that’s fine, if you don’t like my voice, that’s fine, if you don’t like my song that’s fine, but if you want to hear the sort of thing I STARTED working on before I even had reliable sequencer software click here.
At some point in there, my middle daughter started taking piano lessons. Her teacher had a top-of-the-line digital piano with “grand piano action” and we had a spinet that would play if you breathed heavy on the keys. So she would practice her little heart out, then get to her teacher’s house and struggle to play the parts she had learned. A Cedarville student I had met on the MIDI bulletin boards was planning to upgrade his rig and had a Peavey DPM C-8 for sale at an affordable price. It was a controller without any sounds, but it had much heavier action than most controllers. So we bought it and hooked up the FB01 to it for her to practice with.
As a nice feature, it also had MIDI out (multiple MIDI outs, in fact), so I could use it to input VERY realistic piano parts into Cakewalk. Unfortunately, the FB01 did NOT have the world’s most convincing acoustic piano sound, so when my daughter or I played the thing through speakers, the result was hardly impressive. So I started searching for an upgrade.
By then the MIDI committees had agreed on a new spec, the “General MIDI” patch map. Remember, MIDI didn’t send patch names, only numbers. And on one device, patch 7 could be a flute, but on another synth, patch 7 could be a gunshot. So whenever someone traded sequences or upgraded their synth, they had to go through and assign new patch numbers to every part of every sequence.
At the time Roland (who basically ran the MIDI standards committees) had this idea that, if patch names were standardized, they could make and sell MIDI sequences that would play on anybody’s equipment, just like, say, phonograph records or CDs. Not surprisingly, the final list of which patches correlated to which sounds resembled several setups Roland had been using on its own synths. Folks who already owned synths or who programmed their own sounds exclusively got up in arms about it and called it restriction of trade and all kinds of things. But the truth is, there was no reason to freak – the old synths still worked just fine, and nobody said any new patches you came up with would have to fit ANYWHERE in the General MIDI (“GM”) table. Nowadays, with powerful computers and programs that can control literally hundreds of sounds at the same time, from dozens of “real” and “virtual” synths, nobody seems exactly “constrained” by the GM spec, but at the time, keyboard users were pretty much divided between those who saw it as making their lives much easier and those who saw it as “the end of the world as we know it.”
Roland’s first “consumer-priced” tone generator that conformed to the General Midi spec was called the “Sound Canvas.” If that name sounds familiar, it’s because some pieces of software today include a library of sounds that is supposedly based on the Sound Canvas’ original sounds (they’re much weaker, IMHO).
Like the much simpler Casio MT240 keyboard above, the Sound Canvas used a combination of samples of real instruments and many other tweaks to achieve a very realistic sound. Most of the sounds were very believable if you took the time to program in parts that those instruments would actually play. Most of the complaints about the Sound Canvas sounds came from people who wouldn’t make that adjustment. They might, say, want the sound of a banjo, but play in a piano part, ignoring the fact that banjo has unique strumming and picking patterns that are very different from piano chords. Of course it didn’t sound like a banjo!
It would also play several sounds at a time, which was great. Of course I was excited when I saw the things come out, even though they listed for $699. After all, music stores in Cincy and Columbus were already discounting them to $500. So I went to a Dayton music store that specialized in Rock gear and traditionally offered good discounts on pro gear. They had the half-rack version, the SC-55 on a shelf, not plugged in to anything. I asked the X-gen sales guy if I could see it, and he handed it to me and asked me if I wanted to buy it. I said, “No, what I meant was I wanted to see it in action.” Then I proceeded to explain what it did, show him how to hook it up to a controller and to a stereo amp and speakers, NONE of which he seemed to know before I explained it to him.
After putting the thing through its paces for about ten minutes, I decided it would be a huge upgrade to what I had, and I asked him how much it cost. He said “$699.” I said, “Yes, I know the list price, but I want to know how much YOU’RE selling it for.” He said, “$699.” I said, “You don’t sell anything here for list price, why aren’t you discounting this thing?” He said, “Because these hi-tech devices are so complicated that we always spend a lot of time showing customers how to use them, and answering phone calls once they take them home.” And he was SERIOUS! I walked away.
That week I saw an MT32 come on the used market in Dayton for about $175. That was a “tabletop” unit that was in between the FB01 and the SC-55 in capabilities. It had a decent piano sound, and some very realistic string and brass sounds. In fact, it had my favorite chamber orchestra patch of any synth I’ve ever heard anywhere (think the Beatles’ “Yesterday”). It couldn’t play as many notes at the same time as the Sound Canvas, but I figured how to tweak my sequences to get maximum use out of it. (For example, triggering ANY percussion sound required only a single pulse, even if – like a cymbal or gong – the sound rang out long after the note, technically, shut off in the sequence. So if I had a passage where the cymbal was causing one of the piano notes not to be heard, I would back the cymbal up a millisecond, make certain the note duration was the minimum possible, and the full piano part could be heard again.)
By now, I was sequencing a lot of my songs, and using the sequences as backing tapes for my live performances in churches (where backing tapes had come to be expected). This was before home CD-burners were affordable by the way – when I say tapes, I mean Phillips cassettes. But I discovered that many venues that used cassette recorders to play backing tapes used cheap units that would actually change speed during songs, making the guitar or whatever we were playing live sound out of tune.
So when I had the cash and saw them come available on the used market at a price I could afford, I bought a Sound Canvas SC-55 and the matching MIDI playback unit, the “Sound Brush” SB-55, which played MIDI format 1 files on 3.5″ 720K floppies. The SB-55/SC-55 combination fit into a single rack space, shared a remote control, and performed flawlessly. Of course, 3.5″ 720 floppies became virtually unavailable in a year. But by then, I had bought a CD burner for my home office (at $500, it could burn CDs at 2X and was one of the first CD burners sold in the Dayton area). So I started burning my background tracks to wave files and saving them on CDs, which generally played back flawlessly in most venues.
The MT32 became used just for the piano. It now resides in Indiana, along with the Peavey DPM C-8, where my piano-playing daughter is a schoolteacher. It still sounds great. I DO miss that chamber orchestra patch, though.
For a time, when I wanted to do demos with voice and guitar, I could play the “mix” of my MIDI signals onto two tracks of the Dokorder 8140, then add the vocal and guitar to the other tracks. (Later multichannel consumer decks had the ability to sync to MIDI signals or vice versa, but the 8140 didn’t.) To get an idea of what that would look like, mentally replace the cassette deck in the photo near the top of the page with the dbx units and 8140. Then, imagine unplugging the synths, turning off the computer, and adding guitar(s) and vocal(s) directly to the remaining tracks on the 8140.
Additional “toys” I bought during that era included a rack-mount mic preamp that had compression and eq built in, so I didn’t, technically, need to run my voice through the main board. (Sorry, I don’t remember which model – I’ll look it up when I have a chance, but the dbx 286s is close.)
Since I had moved to mostly rack-mounted gear anyway, I replaced my Tapco 6200A “rack-hog” mixer with a sixteen-channel Roland rack mixer invented mostly for keyboard use. I THINK it was a 16E like the one shown to the right, but I don’t remember for certain, since so many of these looked almost identical. I’ll revisit this section when I have a chance to check out my storage area – I still own this one.
I also bought a rack-mount MIDIverb, which would allow to use my PC and Cakewalk to adjust the reverb depth, delay, and volume real-time during recording sessions. Although, I’ll be honest; I typically found a setting I liked for a song and left it there. This one’s still in my rack, too.
Perhaps the most fun toy I bought during this phase was the Digitech IPS 33B harmonizer, an ancestor of the Vocalist line. The IPS 33B was not as sensitive to chord changes as its later brethren, allowing it to generate harmonies that “educated” people wouldn’t sing but which would have been very appropriate in Appalachia a century ago. You could sing into it and get two more parts out, say one part above your note and one part below your note. It was a great deal of fun. I toyed with the idea of using it live, but I never did. I did, however, get a lot of use out of it in my home studio. I traded it later for a Vocalist I could control more accurately with MIDI, but sometimes I miss its quirks.
By this time, I was singing “out” with violinist Wesley Waldron a lot, and beginning to regret that I didn’t have more room for acoustic tracks on my 4-channel tape deck once I added the synths. I could have used a 4-channel with capability of synchronizing to MIDI tracks (or vise versa) to good effect. But by then, 8-channel reel decks with that ability were becoming more affordable.
But that’s the next volume. To go there, click here.