Recently I’ve talked a little bit on my Twitter feeds about how I’m prepping a new hardware based live pa, and I’ve had a few people ask me questions about it. Namely, why hardware and not Ableton Live anymore, and how do I go about creating a strictly hardware based live set. So, I’m going to talk a little about that for this week’s blog entry.
To start with, no I’m not ditching Live and the APC40 completely for my live sets. I’ve been happily using that combination for a couple years now, it’s just time to revisit my past a little bit. If you were one of the bored people who made it all the way through my “History” blog post from a few weeks ago, then you’ll know that my very first introduction to writing electronic music was to put on live pa’s in the late 1990’s.
I’ve revisited the idea a few times over the years since then in a series I call “Morphing Mechanism”, but for the last couple of years I’ve really been itching to put together a brand new live set that doesn’t involve a laptop at all. It’s both a challenge to me, and I think a way to sort of set myself apart a little bit from the plethora of laptop based performers in Seattle these days. I’m sure one day I’ll revisit the laptop based live set (in fact all this hardware work has given me some new ideas on how to do so), but for now I’m focusing strictly on hardware grooveboxes and drum machines to perform with.
I started work on this project early last year with the intention of it being based around an Elektron Machinedrum-UW, and an Access Virus TI2 Polar. In this instance, the Machinedrum (MD) was going to be doing all of the drum sounds, as well as being the sequencer driving 4 tracks of synths in the TI. Unfortunately, after 8+ months of work (and literally on the day I finally considered the set done and ready), I ran into a nasty bug in the Virus OS. An hour before I was to record a demo of the set to pass out to promoters, I lost all of the sounds TI and there was no way I could get them back. Yes, I had been making daily sysex backups, but the bug was such that the backups the TI sent were corrupt and I had no way of knowing this. So after loading one of these corrupted sysex backups back into the TI, all of my sounds were over-written with garbage noise. To say I was upset would be a huge understatement.
A few days later Access confirmed the bug (and released an OS update correcting it shortly after), but by then I was pissed off and fed up, so I sold the TI. Of course, this left me in sort of a quandary. With the TI and all of my synth sounds gone, what was I going to replace it with? In the end I decided to finally take the plunge on an Elektron Monomachine (MnM). The Machinedrum is my favorite bit of gear ever, and I figured it was time to see if the MnM equally as good when it comes to synth sounds. Based on other user reviews, I was a bit fearful that it might not be a sound that I liked, or that it would be too simple for me, though luckily in the end these fears proved to be completely unfounded. The MnM is a very deep synth, and while not as oriented to performance as the MD is, I knew it would work nicely for my new live set.
By this time, I was beginning to think it would be best to just scrap everything from the last live set attempt and start over with a clean slate, so that’s what I did. All the MD sounds and patterns got deleted, and I started with an empty palette on both the MD and MnM. Because the MnM has it’s own built in sequencer, there was no longer a need for me to use the MD to sequence my synth parts either. So for this go around, I’d still be doing all of the drums on the MD, but both sequencers would be running slaved, with the MnM being the master clock. No real reason why the MnM is the master rather the MD, other than the fact that I have the MD on the left and MnM on the right, and it just feels more natural to hit start and stop with my right hand. I again decided to use only 4 synth sounds on the MnM, which leaves two of it’s six track free to assign as effects.
One of the things I find most helpful in preparing and performing live sets, is sticking with a set layout on all my gear when it comes to instrumentation. For instance, I know that no matter what song I’m playing, Track 1 on the MD is always my kick, Track 2 is snare, Tracks 9 and 10 are the HH’s, etc. Likewise on the MnM, Track 1 is the bassline, Track 2 is my lead, Track 3 is the effects for the lead, Track 4 is a random synth, Track 5 is my pad or fills sounds, and Track 6 is the effects for Track 5. Setting things up this way right when you beginning writing and prepping the live set really makes it simple to know exactly what you are controlling at any time in the set. Not to mention trying to troubleshoot things in the heat of the moment when something doesn’t sound right.
The other thing I do when working with hardware live sets, is to treat each pattern like it’s own song. In most hardware grooveboxes and drum machines, your sounds and sequences are organized into short segments called a pattern, typically from 4-32 bars long. So when I’m crafting the set, I basically treat each of these patterns as a distinct song in the live set, and I write between 10-16 patterns to last me an hour or so. Of course, this means that all of the drops, build ups, and variations in each song need to be done on the fly, they can’t be programmed in advance. Normally this is accomplished via muting individual sounds, and tweaking the parameters of different sounds as I play. This is actually my favorite part about performing, as it means that each time I do a set it’s completely unique, and I get to orchestrate it on the fly depending on my mood.
In the case of the MD and MnM, they both have a maximum pattern length of only 4 bars though. This presents some interesting challenges when writing and preparing a live set. Namely, how do I keep things interesting enough and not too loopy sounding? With software, this is less of any issue, it’s easy to add in complex pre-recorded fills, or use longer patterns. So one of the things I’ve learned over the years, is to just not worry about that too much. I just embrace the fact that this is going to inherently be a bit loopy sounding, and focus on making the strongest grooves I can so people don’t mind listening to them for 3-4 minutes a piece. Again, this is one of the great things about playing live versus writing in the studio, in all likelihood your audience will only ever hear these songs this one time, so you can get away with a little more repetition.
That’s not to say I still don’t try and keep things evolving and interesting either. I try and keep each song pretty short, and add a lot of variations with real-time tweaking and mute variations. You only have two hands, so there’s only so much you can do, but I’ve been doing this for almost 20 years now so I have a good feel on how to pace things to keep it moving. It helps that the MnM has 3 really slow LFO’s for each sound, so it’s not too difficult to make things slowly morph over the 3-4 minutes I’m playing each song.
The other trick I’ve learned for keeping things interesting, is to not worry about the drum parts until later in the process. I try and really focus only on the synth parts initially, so that they are strong enough to stand on their own without relying on complex drum parts or familiar rhythms. When I do this, it seems that the songs ultimately are more interesting than when I start with drums like I normally do when writing music. It might seem odd at first, but when you have really strong instrumentation, it’s a lot easier to write drums to fit, versus the other way around. Especially when each groove is only going to be 4 bars long.
Of course the one thing most people ask me, is how do I transition smoothly from one song to another? Let me start by saying that you don’t always have to worry about this. I know a lot of really awesome live acts that only play one song, stop, load up the next song, and then perform it. It’s a perfectly valid way of performing, and arguably has it’s own advantages (like not having to stress about transitions). But, for whatever reason, playing electronic music live has always been about crafting a continuous piece of music for me. Because of this, I’ve always gravitated towards gear that has some sort of facility that makes this easier. Initially it was the Roland MC505 with it’s Megamix, then the E-mu Command Stations with their similar XMIX function.
The Machinedrum UW has a rather unique function in that you can sample both it’s internal output signal, and/or anything coming into it’s inputs at the same time. Samples are mono, can only be 2 bars long at most, and quite honestly sound rather digital since they are played back at a bit rate of 12bits. Still, despite being a limitation, it’s a lot of fun and offers me an easy way to move from one pattern to the next. I merely sample the MD internally at the same time as the MnM coming in externally, loop that, mute all other parts, do the pattern switch while the sample continues to play, and then slowly unmute the new parts from the next pattern. The whole time you can freely tweak and re-sequence the audio you previously sampled too. It’s a terribly difficult thing to describe succinctly, but trust me that it works great and is very simple to do once you get the hang of it.
Initially I was running the MnM directly into the MD’s inputs, but to be honest, anything coming through the MD directly like that ends up sounding rather flat and one-dimensional. All the depth and subtlety is gone. So now I use my RME Fireface400 as a small, but very high quality standalone mixer (it doesn’t need a computer connected to work like this). Both the MD and the MnM go into the FF400’s inputs, where they are summed and sent to a master stereo output. I also have a copy of the MnM’s audio signal going to a separate output which feeds the MD’s inputs strictly for sampling for these pattern switches. The best part about this set up is that both machines sound fantastic on their own, and I can still feed the MnM to the MD for sampling. If the Fireface is out of your budget and you’re interested in this idea for your own sets, the MOTU Ultralite can do the same thing at less than half the price.
So there you have it, a somewhat brief run down of how I’m prepping and preparing my new live set. Currently I’m about halfway through writing material for the new set, though it’s coming together a lot faster than I thought it would. If all goes well, I hope to have a demo recording ready to go in a couple months or sooner, with some live gigs to follow shortly after that. If you’re interested in hearing some examples of material performed like I’ve described, here’s links to two of my previous live sets using similar gear.
This set is done using only the Machinedrum and nothing else:
This live set was done with the Machinedrum doing the drums, and a Korg EMX-1 providing all of the synth parts:
Both sets were done 100% live and on the fly, with no additional editing or processing aside from normalizing done to them. Enjoy, and stayed tuned for the new live set in the near future.
Also, if anyone is curious to see what your’s truly looks like (shudder), I just recorded a new video introduction for my mastering business. You can view it here: