From Egyptology and Commodore Amiga to reflecting on AI and selecting the best physical modeling tools, Ed and Andy (Plaid) open up about their work and practice. We’re turning this year into the next with talks with our favorite artists.
Plaid performing live with Benet Walsh (guitar) and Emma Catnip (visuals). Photo: Matthew Bergman.
Exiting 2024: David Abravanel returns with another in a series of conversations that have been in development over a span of time. David has woven together interviews with Plaid 2023-2024 for this piece. -Ed.
It’s only fitting that Plaid created their own utopian planet.
Let’s rewind: visuals have been a consistent element in Ed Handley and Andy Turner’s music. In the 90s, as part of The Black Dog*, rich visuals with a demoscene/early-CGI look and inspiration from ancient Egypt wrapped the music of the first three Black Dog albums in a distinct mystique. In 2006, Plaid released Greedy Baby, an album-length collaboration with visual artist Bob Jaroc that featured eclectic videos for every piece. Then there’s the soundtrack work for director Michael Arias’ films Tekkonkinkreet (2006) and Heaven’s Door (2009). To say nothing of some truly unforgettable music videos.
Back to that planet – it’s called Feorm Falorx, just like Plaid’s last album, and you can visit it thanks to some cutting work with artist Emma Catnip to bring it to life as an AI-augmented, animated world. There’s an album-length exploration (also called Feorm Falorx), the live visuals, and even a nice comic book detailing the adventure.
While forging ahead, Plaid also took time in the last couple of years to celebrate their storied past: reissues of The Black Dog’s long out-of-print Warp albums Bytes and Spanners showed just how long Handley and Turner have been delivering unique melody and mystique. Plaid were generous enough to sit down with me twice, once in 2023 as an American tour had been postponed and again in 2024 before a performance in New York. As we hear through the grapevine that there may be new Plaid music to come in 2025, let’s take a look at the last couple years – and the last four decades.
Photography in this article courtesy of Matthew Bergman.
Plaid & Emma Catnip’s full-length audiovisual presentation for Feorm Falorx.
Project names and members – cheat sheet. Before anyone gets confused — Ed Handley and Andy Turner have worked as a duo since the late 80s, mainly under the name Plaid, alongside other aliases such as Atypic, Balil, and Close up Over. Together with Ken Downie, they formed the original lineup of The Black Dog, whose debut album, released under the moniker Black Dog Productions, featured compositions written solo and by various combinations of the three. Following the third Black Dog album, Spanner, Handley and Turner split to become Plaid full time, while Downie retained The Black Dog name and was later joined by Martin and Richard Dust – this trio still records as The Black Dog today.
CDM: One thing I wanted to start with, perhaps a bit silly, is the naming of your albums and tracks. The latest album is Feorm Falorx, and you’ve got tracks like “Eshish” or “Crumax Rins.” Where do these titles come from? They almost sound like they could vaguely be another language.
Andy Turner: It’s a variety of things. Often, it’s a permutation of words, mixing letters together, which is the case for Feorm Falorx. There’s usually a real word meaning in there, but we tend to mix it up a little bit. So yeah, it may come across as a bit confusing, but I guess we see the titles as just labels for the music, not to be too directive about the feel of the music.
Ed Handley: Yeah, it’s to do with how the words end up feeling, and the fact that it’s got some internal meaning for us – that’s almost irrelevant because we feel that the final words, the final title, conveys something about the track.
Andy: To give away a secret, I suppose – you gave the example of “Crumax Rins”. That was called that because it was the last track we made on a synth called the Crumar Spirit. We made the bassline for the track, and then the machine died, so it was kind of rinsing the last notes out of the Crumar – maximum rinse of the Crumar.
Plaid – “Crumax Rins” (video by Bob Jaroc).
Another thing that has always stuck out to me uniquely with your music is a command of melodies and chord progressions. It can get extremely lush, compared perhaps to some of your contemporaries who are more timbre-focused with more minimal melodies. I’m thinking of tracks like “Non Hoi”, the morphing of guitars into synthesizers as the chords change. Did you have any kind of classical or jazz training?
Andy: No – Ed has some engineering training, and I can read music and play sort of brass instruments, but neither of us studied to any sort of high level, and neither of us can really play keyboards particularly well. It’s kind of an exploration with us, just slowly trying to find the notes, the chord the comes after the previous chord. You can kind of hear it, but it’s just a matter of feeling it out. These days as well, there’s some great technologies that help out a lot, like Scaler 2, but historically, we’ve just learned by ear as we’ve gone along.
Ed: It really is about finding and feeling your way based on what you’re hearing. It’s not based on someone else’s idea of music theory or how things should be. I’m sure now we’ve developed these kind of muscle memories for how things should go, and what we know about scales and key changes and all of these things that very early on we didn’t really recognize. I think there’s some value in that – there’s value in both.
There’s value in being highly educated because then you can move beyond the standard forms and do something original. But there’s also great value in a sort of naivete, and we’re in the naïve camp. In some ways it’s fun, because there’s all of this discovery.
Ed Handley
There’s value in being highly educated because then you can move beyond the standard forms and do something original. But there’s also great value in a sort of naivete, and we’re in the naïve camp.
For the artwork and tour for Feorm Falorx, you’ve been working with generative AI?
Andy: We’ve been working with Emma Catnip for all of the AI artwork, very closely. She was fortunately on the artist program for Dall-E, so she had early access to outpainting; some of the images in the graphic novel use outpainting. There are various other diffusion models that she’s used – it’s just developing so quickly that over eight months, basically from completing the music to release, the technology was changing every couple of weeks. The artwork started looking one way, and then completely changed by the final package, especially with the graphic novel because we didn’t have to finalize that until about two weeks before the release.
You’re touring with Emma Catnip for live visuals as well. This seems like the most significant collaboration you’ve have with a visual artist since Bob Jaroc and the Greedy Baby album. How is it different, in terms of working on ideas and taking them on the road?
Andy: I think, after we split with Bob Ed basically took on the responsibility for the visuals for a couple of tours. I think he wanted to maybe focus more on the music instead of the all-consuming visual stuff. Emma was just starting to work with the AI tools. We did one show [together] in Berlin as a sort of experiment, prior to the album being finished, and that went really well. It seemed to fit with the whole “fantasy” of the album, you know? We could do very surreal psychedelic things, and that was a bit of the vibe of the album, being given a spaceship and going to a fictional planet.
Part of Plaid’s live setup, including Live running Link and Max for Live. Photo: Matthew Bergman.
How are you both communicating on stage, mainly with your laptops?
Ed: [Ableton] Link really is a great thing, and we have certain tracks that we lead. So one person’s leading on the track, and then the other person’s adding stuff or messing around and that’s it, really. It’s a combination of live mixing, triggering, a fair few live synths as well.
Were there particular synthesizers or instruments that you used frequently on Feorm Falorx? Are you mainly working with hardware or software?
Andy: Well, other than Benet [Walsh, guitarist also known as Mason Bee], everything was in the box. UVI Falcon is sort of a go to, and we’re both quite into Newfangled Audio’s Generate. It’s got a really interesting quality to it because of the oscillators.
Ed: We’re fairly obsessive about software and trying new things out. We’ve got access to a lot of them, and play around with some, and they kind of fall away eventually and you stop using them. But no one’s really doing those kinds of chaos oscillators like in Generate. It’s quite unusual to get that, and kind of nerdy to have that technology in a plugin. I’ve come across similar [oscillators] in Max years ago, but it’s never really made it to a plugin until recently, as far as I know. For sampling and physical modeling, UVI Falcon is pretty great, and for additive, Loom is still actually really lovely as well.
Andy: And Audio Modeling, the company – they make all the cool physical modeling instruments that we tend to use.
Plaid – “Wondergan”, featuring the Audio Modeling Solo Brass instrument.
AI was part of the visual process for Feorm Falorx and the tour – are you using any of the music or audio applications of AI/ML? What are your thoughts on the tools that are emerging?
Andy: Yeah, I’m following it very closely. Until very recently, there wasn’t anything that I considered good enough sounding to be usable. MusicLM is the one that I think has got the most potential. I’ve done a crash Python course, to try and get up to speed, because everything’s sort of happening so quickly. But not in the way that it’s intended to work. I think with a lot of music technology – like the 303, for example, was to replace an electric bass, but then it turned out it could make other stuff. I think it’s going to be interesting on the edges of their technology. They’re saying that with MusicLM, it keeps coherence for about two minutes and then starts to degrade. And then they give you examples of the “cool” two minutes. And I’m like, “I want to hear what happens when it degrades!”
I’m not at all threatened, because there are already tens of thousands of kids writing electronic music and churning it out every day, and they’re humans. They’re much better than a model, and they’re making really competent stuff. So I mean, these tools will be able to churn out more generic music, but we already have tons of that stuff. The stuff that will stand out will either be using that tool originally or just writing something that’s new that hasn’t been done. I find it incredible fascinating.
They’re saying that with MusicLM, it keeps coherence for about two minutes and then starts to degrade. And then they give you examples of the “cool” two minutes. And I’m like, “I want to hear what happens when it degrades!”
Andy Turner
Live trip to Feorm Falorx from Brooklyn, NY. Photo: Matthew Bergman.
Ed: I’ve used a couple of small specialist AIs – there’s one called Drumgan, which is incorporated into Steinberg’s Backbone software. It’s a drum AI that Sony came up with it. It’s not amazing, because obviously it’s trained really on very quite conventional sounds – an 808 and a 909, say – so it never really comes up with anything that radical because of its data set. But it’s nice that it’s this very specialist tool just for making percussive drum sounds, and you could expand it. I’m sure if you started incorporating something like FM into its data set, you’d get much more interesting sounds.
I think I can see a future in using these very small, very specialist AIs, rather than these AIs that just make a whole track. That’s kind of fascinating and mind-blowing, that it can do that with resynthesis, spectrally from any kind of audio data. It’s like a kind of Frankenstein dream and I think in the computer music world it’s been one of the “holy grails”, to just be able to spectrally produce a piece of music from nothing. It’s exciting – but I like the idea of, say, a specialist bass AI synth. It can be parameterized as well, much more easily than a “make a track” AI.
Andy: All the models communicate differently, so you have to work with them for a while to know how to “speak” to them to get what you want. I imagine the prompts will become much more sophisticated and the algorithms will become better at understanding and giving us what we’re asking for. It’s a really crazy time for music, with this technology.
Generating sequences or tracks, it can be an uphill battle, right? Some of it, you’re thinking – “oh, this is just terrible. No human would make this decision.”
Andy: We’ve worked with generative algorithms, certainly for MIDI, for years and years. It’s a fairly natural process for us to have an hour’s worth of generated stuff, and then have to listen through and curate the bits we like and work out which bits go together. And I imagine it will be similar with these new systems. I mean Magenta have something like an outpainting but for music, where you can have two bits of a track and it will work out how to get from one bit to another bit.
Ed Handley immersed in Feorm Falorx. Photo: Matthew Bergman.
Do you ever improvise with these tools in a live setting, or is that just a little too dangerous or too processor-heavy?
Andy: For MIDI stuff, we might have something sort of randomized. We’ve had smaller elements, but not something that would bring down the whole circus tent, if it went wrong.
Ed: The Magenta MIDI stuff that you can use in Max is fairly quick, because you just sort of press a button and it’ll generate a four-bar sequence. So you could use that in real time. I don’t think we have.
Music doesn’t necessarily need an AI that much. Certainly with MIDI, the algorithms for generative components and music theory are good enough that it doesn’t need some artificial intelligence. It’s much more rule-based and pattern-based, which also makes it much more predictable than AI.
And as this advances, you get AI that intelligently breaks these rules – something like a virtual Ornette Coleman of MIDI, these kind of avant-garde departures.
Ed: Yeah, some of the bizarre [AI] piano improvisations I’ve heard where it’s done a kind of resynthesizing of a piano. It’s utterly bizarre, because obviously the AI doesn’t really know about scales and pitches, it just knows about all these piano pieces that it’s been fed. So if there are a few in there with atonal chords, it’ll do the most bizarre sort of joining together, and then you get this strange chord in the middle of a pop track. Nothing a human would likely do, which can be interesting.
Andy Turner piloting things towards Feorm Falorx. Photo: Matthew Bergman.
In the graphic novel that comes with the Feorm Falorx special edition, there’s the story about the two of you going to another planet and playing a beautiful festival.
Andy: Yeah. I think that sort of grew naturally from the obviously very real experience of us actually being taken to another planet and playing at an infinite festival [laughs]. And for the earthbound version of it, we just had a hell of a lot of fun recreating the environments that we saw and the things that we heard.
There is a kind of feeling of being there when listening to the album. I’m thinking about Drexciya – Gerald Donald said that he and James Stinson used to do a kind of meditation before working on tracks. Along the lines of closing the doors to the studio and saying, “we’re a number of ‘aquaknots’ below the surface, we’re in Drexciya.” Do you have anything like that?
Ed: I mean, I think all music-making is kind of that because you enter this world of the piece of music. If it’s themed, then obviously, you’re part of a larger, broader landscape that maybe contains tracks that are related or similar sonic landscapes. If we’re being very literal, they may just use similar palettes or technology. But on a more metaphysical level, they exist in kind of the same space.
I think I’ve always found that with music making, you lose yourself, and you’re immersed in a completely new space. I think you can do that with a whole cluster of tracks together where it’s kind of visceral, making music that’s actually based on a kind of experience you’ve had.
I’ve always found that with music making, you lose yourself, and you’re immersed in a completely new space.
I remember the press release for Scintilli suggested that you said that word 50 times together before working on music, something like that?
Andy: I’m not sure that’s true, actually [laughs].
Ed: I mean, it’s pretty esoteric, right? Music that gets you going to that hidden space. I think it’s probably different for everyone, but there is a ritual or ritualistic element that is useful. We haven’t employed a kind of group method so much, but we each have our own techniques.
Andy: I think it’s sort of a broad thing – finding someone you can communicate with on a level and those conversations where there’s no holding back. You’re just free thinking, and that’s so powerful creatively, and that’s where collaboration can lead to exceptional ideas.
The Black Dog – “Pharaoh”, Amiga demo c. 1990
The visual world-making and mythology goes back pretty far for you both. With the recent reissues of the Artificial Intelligence compilation and two of the Black Dog albums on Warp, I was thinking – there’s a significant influence of Egyptology, and perhaps the occult. Looking at some of the old Amiga programs, or the artwork for Bytes and Spanners.
Andy: I guess I hadn’t thought about it like that. I think with the Black Dog stuff, there was an interest in finding out about things. Ken [Downie, founding and current member of The Black Dog] was an Egyptologist – or, he studied as a hobbyist, so he had a lot of knowledge there. When we met him, we were gathering a lot of knowledge from him. I didn’t know much about that period when I was 18 or 19, when we first started working together.
The front and back cover for The Black Dog album, Spanners.
We were pushing the technology for [Spanners]. For example, Ed and I did the front and back cover for that, and at the time we were using 3D Studio. Each cover had to be rendered in nine panels to get it to the highest resolution we needed. And each of those nine panels took about 20 hours to render! I did the front, the Cerberus dog. You’d get your model together, spend your nine times 20 hours rendering everything out and then realize something was sticking out wrong, because you couldn’t lock everything together like you can with modeling these days.
Technology is probably more something that we’ve retained, than the idea of making worlds for people. But I would agree with Ed totally about the process of writing, which in itself is almost a meditative sort of practice.
Did you ever find that some of these frustrations or limitations with technology could also inspire you?
Andy: The first album we made, Mbuki Mvuki, that Ed and I did; we were sharing a flat at the time and we had a monophonic synth, an [Akai] S900 or S950, which was also monophonic and with very little space. We were sequencing with Dr. T’s. You make do with what you’ve got, very limited kit, but you can make music with anything. We could make music with this table!