(Warning: this post contains potentially hazardous doses of geeky tech-speak, which may cause drowsiness, dizziness, or irreperable damage to one’s dating life.)
Thanks to everyone who’s posted and emailed with ideas, suggestions, and offers about the 5.1 Echoplex gigs in NYC.
A few days after my previous post, I was laying awake in bed, with visions of MIDI data routes, audio signals, and synchronization relationships dancing through my head. Sorting through different combinations of hardware, sync networks, and control options, trying to figure out how to configure everything to let me do what I wanted, without having to scroll through five dozen footpedal banks or suffer a nervous breakdown.
Finally, I stumbled onto an arrangement that seemed like it would do everything I wanted to be able to do, with a mimimum of hassle. Having put it to the test now, I can say (knock on wood) that I’ve figured out my approach for five Echoplexes at the same time.
(“And there was much rejoicing…”)
Here’s the setup:
1) I’ll use two MIDI controllers – a Digitech PMC-10 footpedal, and a Roland 626 drum machine. The Digitech pedal I’ve already detailed in a previous post. The Roland 626 is the very same drum machine I got back in the summer of 1988 from my dear mom and dad (over half my lifetime ago!), a good half year before I started playing guitar. (It’s also the unit I programmed all of the drum beats for Disruption Theory on.)
But I’m not using the drum machine to make beats, trigger drum samples, or anything like that. Instead, I’m using it for the 16 buttons on the front panel. Ostensibly, each of these buttons is used to trigger or program a different drum sound, when programming beats on the drum machine. But each button can also be set to send its own specific MIDI note. Which means that I can configure each of the 16 buttons to trigger a different Echoplex function and/or loop number. Voila! Instant hand controller! (This is what I used as my first Echoplex MIDI controller, before I got my PMC-10, so it’s kind of fun to be bringing it back into the equation.)
The PMC-10 and the 626 go into two seperate inputs of a MIDIman/M-Audio 3X8 MIDI through box – which is a box that will take the signals sent to the three different inputs, and then route them to the eight outputs. The real beauty of the M-Audio unit, for me, is that any of the three MIDI inputs can be routed to any of the eight MIDI outputs, or – crucially – any COMBINATION of the eight outputs, using physical three-way switched located on each of the eight output channels, which I can switch and alter by hand, on the fly.
So this means I can set all five Echoplexes to be controlled by the footpedal alone, by routing all five MIDI outputs to be controlled by the footpedal input. Or I can set them all to be controlled by the drum machine buttons, by assigning all MIDI outputs to the 626 input. Or (and here’s where it starts getting fun):
Set two Echoplexes to be controlled by the footpedal, so I can drop Replace fragments into just those units, while the other three Echoplexes are controlled by the drum machine, where I can perform overdubs, multiply, or a myriad of other functions completely independently of the commands being sent to the first two…
Or control three Echoplexes with the footpedal, one with the drum machine, and set the fifth one to the third (and unused) MIDI input, so that I know for sure that it won’t be tweaked by mistake…
Or have five independent loops, completely unsynchronized, of different lengths, switch all five MIDI outs to send data from the footpedal, and then do an Unrounded Multiply to impose exactly the same new, strict length onto all of them…
Or send continuous controller data from the footpedal to make a feedback fade on four Echoplexes, while the fifth one sputters away untouched…
Basically, I can have two completely different paths of control data going to any of the five Echoplexes at any point in time. I can even start a function on an Echoplex with one controller, switch the input for that channel, and finish the function with the other controller, if need be.
The beauty of this approach for me is that I don’t really have to change anything that I’m already doing. Each Echoplex can be set to the same MIDI channel, to recieve the same commands, and I can control any of them using exactly the same footpedal banks that I’ve been using for the last three years. Rather than come up with a whole new series of footpedal banks, controller options, etc., I’m expanding my “regular” approach in a modular fashion, so that I’m building on what I’ve already done, instead of coming up with a whole different way of doing things.
2) Synchronization is done by “BeatSync,” which is the most basic and limited form of synchronization available. But that’s the beauty of it, too: it’s kind of idiot-proof – a master unit sends an audio “click” at the downbeat of every cycle, and any slave Echoplexes read that “click” as their reference point. What’s particularly nice about BeatSync for my purposes is that it’s completely analog – it doesn’t involve MIDI at all, so I can keep the MIDI pathways dedicated exclusively to transmitting controller data, without having to worry about how changing control information might gum up synchronization pathways.
Disadvantages to BeatSync? Well, it’s not sample-accurate, which means that the loops will always be a few milliseconds apart from one another. But that doesn’t bother me, since I’m not looping any stereo images. I do hear a bit of flanging if I overdub the same signal into two units at the same time, and then monitor through a mono source, but I don’t see that being a problem with 5.1 – on the contrary, I actually think having a random few milliseconds difference when the loops hit will make for a subtle but very musical variation in the “space” of the sound. And even though the loops are a few milliseconds apart each time the loops repeat, they don’t drift further apart over time, so the sync is very tight over long periods. The slave Echoplexes are sort of “retriggered” every time they get a new “click” from the master, so each “retrigger” happens at a slightly different time… but always within a very, very narrow (and thoroughly acceptable) margin of variation.
Also – and I could be completely imagining this – but there seems to be a very slight extra “depth” of sound when I have two Echoplexes synchronized together running BeatSync, with different overdubs sitting on different units… could it be that the very subtle, but real, variation between sync betwen the units is causing a microscopic variance in the sound field?
(I just re-read the immediately preceeding paragraph, and was reminded of Chicken John’s advice to me at the Odeon Bar last year: Time to get a girlfriend, man!)
So, uh, where was I?
Aside from that, the only other problem with BeatSync is that one Echoplex has to be designated as the master. But that’s not really a problem either, because if I find myself, for instance, fading out all loops except for Echoplex number 4 (which happens to be following Echoplex number 1’s master clock), and then want to change the loop length of number 4 without adding any sound into any of the others, I can simply assign both 4 and 1 to the same controller, so that any expanding or contracting of the basic loop length will be the same on both units. Meaning that Echoplex 1’s length will change exactly the same as number 4’s will, so it’ll end up providing a master clock that’s identical to the length of that of number 4.
God, this is my geekiest post ever. I promise to do my best to follow up with a thoroughly non-technical post, filled with obscenities, crass sexual metaphors, and meditations on guitar colors, before too long…