Teaching an 808 some new tricks

TR-808 and Korg KMS

I’ve been really happy with my Roland TR-808 ever since I got it a year or so ago. I bought it because I thought it would magically solve my inability to make good drum patterns. Percussion has been a weak point for me, and I thought that maybe I needed a better tool.

I first tried using a TR-707. It’s much cheaper than an 808 and yet it’s got a really useful-looking matrix that shows the drum pattern for all the different drum sounds at once. I thought that a visual aide was just what I needed. But it turns out that the 707 doesn’t let you change the pattern while it’s in the middle of playing. This was a major blow, since I wanted to be able to experiment with the pattern while it was playing.

So I eventually shelled out and just got an 808. It’s very simple and intuitive to program, which is exactly what a rhythm dullard like me needs. However, I confess that despite my satisfaction with the 808, I have found myself staring longfully at some of the other drum machines from the 80s. Sequential Circuits’ Drumtraks, the TOM, the Oberheim DSX or DX, E-mu’s Drumulator. The list goes on (though that is mostly it for me). However, these are all just drum machines with digital samples. I didn’t like the idea of amassing a bunch of drum machines (no matter how nice they look…and they do look very nice) just to access the samples. Yeah, I know that some of them have analog filters. Well, I also have samplers with analog filters. Problem solved. Almost.

I still liked the 808’s method of programming patterns. The best world would be to use the 808 but have some additional sounds. The 808 has three trigger outputs, and these could potentially be used to trigger other devices. But I usually want more than 3 sounds in a pattern. Instead of using the trigger outputs, I decided to run the individual audio outputs from the 808 through a box that would convert them to MIDI notes.

808 outputs

I first tried using an Alesis D4 to convert the audio into a MIDI signal. I always thought of the D4 as just a way to get some cheap drum sounds, but it’s also got a ton of trigger inputs. Alesis intended for folks to hook up drum pads to these trigger inputs, but it supposedly also accepted regular audio as triggers.

The D4 didn’t work so well at converting audio into MIDI. It worked ok on certain sounds, but less reliably on other sounds. I could never get it to convert the hi-hat sound, for example. I didn’t have much patience for the D4, and an Akai ME35T popped up on eBay for cheap. I actually meant to bid on the auction and then got busy and accidentally forgot about it. But luckily no one else bid on it and the seller relisted it. Hurray! Plus, I just think Akai gear looks nicer than Alesis stuff.

Alesis D4

The Akai worked like a charm, and it was also much easier to configure all the various triggers than the D4. Eventually my master plan is to sample various drum machines and build up a library for either my ASR-10 sampler or the Emulator III. In the meantime, I hooked the MIDI output on the Akai ME35T into the D4 and just used the stock sounds on the D4. It’s got a variety of classy-sounding 80s drum kits that are fun to switch through. Anyway, I made a little demo of the process:


[pro-player TYPE=”MP3″  image=”http://thecompleatsynthesizer.com/media/images/808.jpg”]http://thecompleatsynthesizer.com/media/audio/tr808triggerdemo.mp3[/pro-player]

Tascam Tape Time

When I was in high school, I bought my very first multitrack recorder from my church. It was a four-track Tascam tape recorder, a Tascam 234, and it had sat tucked under the sound booth unused for many years. My older brother often ran the sound board for church services, and so I used to sit back there with him as a little kid. Many times I stared at that curious tape machine and wondered why it was so different than other tape recorders. Why so many knobs and meters just to record a tape? I definitely didn’t understand the concept of recording four independent tracks to a single tape.

Once I started to record my own music (with a single lone synthesizer), I finally realized the importance of being able to go back and record multiple tracks on top of one another. Our music pastor knew that I was starting to record music and he sold me the tape machine for $100. I used it a little, but I quickly realized that I needed a way to synchronize my sequencer and the tape machine. I didn’t know much about tape sync, but the idea of sacrificing one of my tracks as a sync track was not attractive. Only three extra tracks left? Not for me! I retired the Tascam and soon found a digital 8-track recorder that could sync to MIDI clock. I’m actually not sure what ever happened to the Tascam tape machine, but it’s been gone for a long time.

But no longer! I’d been thinking nostalgically about that tape machine for a long time. From a design standpoint, it just looked really nice. I also like the concept of recording stuff to a tape machine. I’ve been hunting around for one on eBay, but a lot of them are in disrepair. It seems that these things have a lot of internal problems. When they do sell, they go for a lot more than the $100 I paid our music pastor. But last week I was driving across the country and made a little stop in Kansas City. A guy was parting out the remainder of his studio and had listed four of these on Craigslist. Naturally, I offered to give them all a good home.

Tascam 234 stacked


He said that they were mostly working. I suspect that I might have two working units and two parts machines. I like the idea of having one (or two) for spare parts. If eBay is any indication, these tend to break down in a variety of ways. That was one of the things that made me nervous about getting one again…I knew I’d be buying something fragile. But it’s a little nicer to know that I’ve got a large cache of spare parts.

Tascam 234


I’ve got a couple other projects to finish up first, so these are going in the corner for right now. Hopefully I’ll have at least one working soon and I’ll try recording a song on one.

Patchbay Afternoon

I’ve been putting this day off for far too long: the day that wire up my patchbay. Without any patchbay hooked up, the studio has been a messy tangle of temporary cords strung across equipment and laying like little booby traps around my chair. I used to use a Furman PB-40 patchbay, which had convenient 1/4″ phono jacks in the front and back. I retired the Furman when I discovered that it was the cause behind some mysterious noise problems in my signal chain.

Vowing to only buy quality components, I scored a Switchcraft patchbay off eBay for cheap last year. I think it was cheap because the rear hookup was not very convenient. Instead of 1/4″ jacks or something, it had a snake on the back to ended in a totally separate 2U rack plate with a ton of little pins. My original plan was to cut the snake and just splice my own 1/4″ jacks onto the wires.

Switchcraft patchbayThe picture is a little confusing. The old Furman patchbay is sandwiched in between the two Switchcraft pieces. The red/black/white thing on the bottom has all the pins.

Anyway, my plan for chopping up the snake changed when the patchbay arrived in the mail and my dad saw it. My dad works in the switch room at the local phone company, and he immediately recognized the connections. He told me that I just needed to get a punch tool and I’d be able to easily connect bare wires to the red/black/white pins.

Bare wires

Hooking up all those wires seemed daunting. I started to wire up an old Hosa snake that I had laying around. I stopped myself after half an hour and thought, “What am I doing? I’m connecting crappy cables to this nice patchbay and I’ll just drag the sound quality down to the Hosa level.” So I delayed and picked up some better bulk cable. A wise idea…but with that step the patchbay became yet more work, since I’d have to wire up phono plugs onto the new wire. And so for months I just totally bypassed the patchbay and hooked equipment directly into each other. I would probably be doing that still today, but a new purchase forced me into action.

Apogee 800


I bought an Apogee Rosetta 800 as my audio converter. Actually, as you might expect, it’s a little more complicated than that. I first bought a two-channel Apogee 200 from a Guitar Center in Virginia. It came with the critical firewire option card that I needed. Two channels of input was fine, but I needed at least four channels of output (two for monitoring and two for outboard effects). My thrifty side thought that I could assemble some sort of Rube Goldberg concoction that involved a cheaper audio converter (maybe an old Mark of the Unicorn box) and then tether it to the Rosetta digitally.

After I got everything hooked up, I realized that this complex arrangement wasn’t possible with the Rosetta 200. Only two channels of audio could come in and out of the firewire connection, period. And so I decided to sell the Rosetta 200 (keeping the firewire card) and pick up an eight-channel Rosetta 800. The 800 showed up last week, and that’s when  I realized that somebody didn’t do his homework.

The Rosetta 200 uses XLR jacks for all its audio connections. This worked fine and I had plenty of these cables. The Rosetta 800 uses DB25 plugs for all its audio input and output. “DB25” is the technical name of the parallel port connection that old computers used. Apparently it lives on in the world of high-end audio converters. It seems that the more expensive the audio converter, the more inconvenient it becomes to hook it up to any other equipment.

So I started shopping for DB25 snakes. Once again, I found some Hosa snakes but decided that my Hosa days are behind me. I thought about just buying the raw materials and soldering up my own snakes, but that sounded like it might be a lot of work. Indeed, folks told harrowing tales on internet forums of trying to soldering all 24 wires inside that connector. I decided to stop being thrifty and just buy it pre-made. I found a guy out in Brooklyn that sold nice-looking snakes that terminated in bare wires. The showed up yesterday, and I knew that if I wanted to start using the Rosetta 800, I’d need to break down and finally start wiring up the Switchcraft patchbay.

So that’s what I’ve been doing today. I’ve got all the outputs working great, and I wired up a couple custom cables that are running to my monitors. I took a half hour break to write this thoroughly tedious post, which has surprisingly boosted my enthusiasm for going back and wiring up the rest of the patchbay, as I now know that there are actually more boring things in life than wiring up patchbays. Namely, writing posts about wiring up patchbays.

A swing and a miss

Working on music takes a good deal of persistence. Stuff generally just doesn’t work. That applies to both musical ideas (I can’t count how many times I’ve written up a supposedly great counterpoint line only to find that it clashed badly with the melody) and with the technical side (the extremely fickle Emulator III is back in the non-working camp). Part of the difficulty in sitting down to work on music is coaxing myself past the knowledge that what I produce this evening is probably going to be crap. No, really. I’m not being coyly modest. My ears have heard things that sound very, very bad, coming straight out of my very own synthesizers. But the only way forward to better songs is through the slough of cheesy, disinteresting, annoying and unoriginal song ideas. In honor of that, I present the latest mis-attempt: wiring up a monitor output for my Roland sampler.

The Roland samplers from the mid 80s (the S-550, S-750 and S-770) are a pretty cool bargain right now. They all sell for less than $150 and they’ve got some cool features. One of those features is a monitor output, a rarity for music equipment in the 80s. They actually have two monitor output jacks on them. One is an RCA jack and it’s really easy to hook it up to most TVs. Most tube TVs since the 90s have a composite RCA video input (color-coded yellow). This is very fool-proof, but sadly the display is monochrome. You just get black and white or blue and white or whatever the TV decides to do.

The other output is a “digital RGB” output, and it offers that elusive possibility of color output. I’ve wadded through many forums and read contracted arguments over what this term exactly means, and what standard it ascribes to. Roland apparently made a monitor that was perfectly compatible, but it has been wiped off the face of the earth. My friend Jacob has been a quest for this holy grail of monitors and can scour up only a few ad promo photos. In the absence of this monitor, people have devised several ways to convert the output to VGA, CGA, SCART and who knows what else. I surveyed the mess and decided that the easiest thing would be to hook it up to my old IBM CGA monitor.

Roland monitor attempt 1


The circuit is supposed to be pretty easy: just a handful of resistors and two transistors. Roland includes a schematic in the back of the sampler’s manual for the “RGB-25I” cable, and I think this is supposed to be compatible with CGA monitors. The transistors are labelled DTC 11 AFE. I have no idea what that means. So I tried standard 3904 NPN transistors. The trickiest part is actually the connections. The plug for the sampler is an eight pin DIN plug, which isn’t very common anymore. I found some plugs on Mouser but I wanted to test out the circuit immediately, so I just stuck wires into the appropriate holes in the DIN jack. On the monitor side, I needed a female DB25 plug, which I picked up at Radio Shack. An hour or so later and I had this unholy mess:

Shameful breadboard


Yeah, it’s pretty shameful. Yeah, it actually took about an hour. I’m a little slow.

Eagerly, I switched on the sampler and the monitor and found…



…nothing. Just a quiet, very blue monitor. It’s an attractive blue but sadly, a non-working blue. I checked my circuit a little more to no avail.

But I threw this together in such a slip-shod way that I actually would have been more surprised if it had worked. Come on! Wires in the DIN socket? Masking tape to keep the resistors insulated from one another? So I’m just going to do this the right way now: order up the actual plug online, and solder everything onto a real board. It’ll take a little more work, but hopefully I’ll soon have a post about a working color monitor output for my Roland S-770.