The Grand Piano as Acoustic Synthesizer

A conversation with Spectrum Piano developer Doug Blackley
Publish date:

For years, musicians and inventors have experimented with fitting acoustic pianos with magnets to augment their capabilities. A carefully placed electromagnet can affect a piano string’s sustain and harmonic properties and give a player considerable control over those properties. One modern example of an augmented piano is the Electromagnetically-Prepared Piano originally developed at Stanford University’s Center for Computer Research in Music and Acoustics (CCRMA).

Another is the Spectrum Piano, developed by Canadian composer Doug Blackley and featured in his ethereal 2016 album, Vanishing Evocations.

The Spectrum Piano is a device that uses electromagnetic energy to vibrate a piano’s strings under Blackley’s control. The current version has 24 electromagnets, each placed directly over the two or three strings that produce a single pitch. Because he can vary the power of individual magnets and thus the intensity of vibration, he can control the pitch and envelope of the resulting sounds. And because his device targets specific overtones by selectively stimulating a string’s natural harmonics, he can control harmonic content and thus, the sound’s waveform.

You can hear audio examples from Vanishing Invocations, as well as read more details on the website

How do you describe the Spectrum Piano?

I call it an acoustic synthesizer because it can do things that are obviously within the boundary of synthesizers and produce sounds that normal acoustic instruments can’t. The difference between it and a conventional synthesizer is, of course, that a conventional synthesizer uses speakers to output the sound, whereas this uses strings and a soundboard; therefore, it’s acoustic.

With a synthesizer, you have control over properties such as harmonic spectra. Do you have that with your Spectrum Piano?

Absolutely. If I have a string tuned to 55Hz, one of the low A strings, it would produce the sound of 55 cycles. But when the string is hit with the hammer, it outputs 55, 110, 165, and all the other frequency multiples. I can get it to sound the harmonics without sounding the fundamental. So, I could take that string and tell it to make a pling on the third harmonic, and I could make it do a steady-state sound on the fifth harmonic at the same time. There would be two different sounds coming off of two different harmonics from the same string. So, I can determine the harmonic content. It’s an electromagnetic-drive system. I don’t use the hammers.


That way, I can bypass having to use the fundamental. In fact, I use low strings to get a lot of high-frequency sounds. If you excite the fundamental of a piano string, you get this wonderful, huge sound. It almost sounds like a pipe organ, because if you excite the fundamental, of course you also excite all the overtones above it. But if you excite the harmonics and don’t excite the fundamental, there’s a ton of stuff you could do by exciting the harmonics of multiple strings at the same time, which fuse together into a synthesized timbre made up of, say, five different strings, each outputting different frequencies on different harmonics. That all happens at the same time, and it produces this wonderful, rich, cohesive sound.

Once you set individual harmonics in motion, how much control do you have over them?

It depends. First, I can excite them with a variety of envelopes. I can give them the hardest envelope I can get, which is still not going to produce anything like a piano hammer, because if you can push it that hard, the sucker will melt. I’ve done that, and it melted. But I’ve done plucked sounds, and you can get a nice ping, ping, ping, ping.

Could you play a melodic pattern from harmonics on an individual string?

Yes, I could, but I wouldn’t bother because I’ve got lots of strings.

Then could you play a harmonic pattern on a lot of strings just by controlling the magnets?

Oh, yes, I can do that.

How do you send frequency into the magnets?

Electromagnets are conventionally used with direct current. You send in a direct current, and it sends out something that excites the core, and it sends out a magnetic signal that will pull things towards it. But if you send out alternating current into it, then it will alternately push and pull.

And then you have frequency.

If I send out a 440-cycle sine wave of alternating power, then it would push and pull 440 times a second. However, there’s an issue at the same time. You can’t push with magnets. You can only pull. So, of course, you wouldn’t hear 440 cycles from this when you heard this thing vibrating. You’d hear 220, because you’re only getting the pulls.

How do you trigger tones and control harmonics?

I use a MIDI keyboard, and it instructs my computer to launch into sending signals that eventually wind up at the electromagnets.

Does it respond to MIDI velocity to control dynamics?

Absolutely, yeah, and it does pitch bends, too.

Do you control envelopes by changing the power of the magnets?

Correct. Ramping them up and ramping them down and by adjusting the frequency content that I send into the magnets. And I can send many different frequency [spectra] and run them up and down into the same magnet at the same time. So, you can do anything you want.

Can you use the electromagnets to stop a string from vibrating?

No. It’s a good idea if you could. In theory, if I could somehow measure the exact frequency vibration of the string and be correctly in phase, I could use it to stop a string from vibrating by sending the same sound as that but 180 degrees out of phase, so when a string goes away from the magnet, I pull it, and when it comes in, I wouldn’t touch it. That would dampen it. But I have no idea how. I mean, that would be like a neural network, rocket science beyond my world of dreams.

Then do you suppress vibration mechanically, like with a damper?

Oh, sure. There are a number of issues with dampers. First, if my piano had a sostenuto pedal, I could have a bunch of strings with a sostenuto pedal down, and they could ring endlessly. And then I could use the conventional damper on the other ones. But when I chose to make [the Spectrum Piano], I chose to leave behind the whole idea of hammers and all that stuff, because this is not an augmented piano. It’s a synthesizer, which is a different thing. I can play sounds with the damper up, and they’re much quieter because they don’t get this big, long resonance that they have when the pedal is down. Normally, I block the sustain pedal down so that it’s on all the time.

But if you hold it not quite down, you can get sounds that decay away much faster. Of course, if it’s all the way up, it will just stop, which is normal, but it’s quite a bit quieter. The issue with that is, at this point, mechanically speaking, I don’t have a way of holding the sustain pedal anywhere other than just on or off for start.

The second thing is the music that you heard, for example, the first piece, which is called “Flow,” it will do that all in real time, all at the same time. It’s not multitracked. In that, you heard lots of plucking short sounds and long sounds and different sounds. If I were to put some kind of dampening on the string to make the plucking more articulated, then it wouldn’t be able to make the long sounds anymore.

What kind of technical challenges have you encountered and overcome?

Oh, I have so many.

What comes to mind first?

I can think of the day when I went downstairs to get something, and the system was on, and I heard this weird sound. I thought it was like a foghorn from a ship outside. Then my wife said, “Hey Doug!” I went running upstairs, and it was just booming loud, and there was smoke pouring over the top of the piano. You don’t want to light your piano on fire. And, of course, it didn’t, but that was me melting down a bunch of magnets.

I did all the metal work. I soldered the thing together. I came up with the physics of the idea of how to do the thing. I did the software programming to allow me to drive the magnets in the way that I wanted and all that. But I’m a composer; I’m not a technician. And one of the reasons why I think the thing works so well is because I had a specific musical need. I understood the physics and thought I ought to be able to do this in the way that it did. But then it came to, well, gosh, I’ve got to get a saw and I have to cut metal up. So, the whole thing took a lot of time, in terms of technical stuff.

How long have you been working on it?

I started in 2010. I went into Master of Fine Arts at the university here. I chose to study music for the acoustic piano, because my whole life I’ve been using synthesizers. I first plugged into modular in 1978—a long time. And then I did music for film and TV, and then I taught. But I got tired of that, and I wanted to go back and work on improving my melody, my harmony. So, I chose to study acoustic piano because that wouldn’t allow me to be Mr. Synthesizer and dress it all up in cool sounds.

So, I did that, and the time comes up when you have to present something for your thesis. And because I studied acoustic piano, I’d have to write a bunch of music for piano. I play composer’s piano, but I’m not a concert pianist, so I’d have to get someone to play my complicated stuff. People said, you could do music for piano and synthesizers, for example, but I really wanted to stick to my guns. I went there to study piano, so I wanted to do my project with piano.

I thought back 30 years ago or so to a theater show that I did a live score for. At the time, I stuck an electromagnet on a raw piano frame with the strings, and the whole score for this theater production was me playing that. I wished at the time that it could sound continuously, but it couldn’t. So, I tried a magnet, and I made a little functioning system. It was weak and feeble, and I didn’t know what I was doing, but I knew what could happen. I know there have been electromagnets on piano in the past to make the string sustain longer to augment the sound.

I thought, what if I could go back to that? Surely, I could make it do different sounds other than just vibrate the strings, because the physics of the situation told me that I ought to be able to do it. I ought to be able to make it make synthesizer sounds. And if it could do that, I would be able to do my thesis performance using only acoustic piano, and I still get to have all my wonderful synthesizer sounds. So, the thing to do would be to change the piano, rather than sort of change the focus. I thought it would work, and it did.

How many versions of the Spectrum Piano have you designed?

I did the first performance for the thesis in January 2013. And over the next year-and-a-half, I built another version, which is the really spectacular-looking one. That’s the second version. I did a couple of small concert performances with it. I got a concert pianist [Andrew Czink] who’s really into avant-garde stuff to be my second player for it, and he just loves the Spectrum Piano.

After that he said, “Doug, we’ve got to do something with this. We’ve got to have a duo.” So, we made the duo where we do all this sort of modern, strange, wonderful music. He found these various things that we could do. We did a performance at a place in Vancouver called the Western Front. We also were invited to the Newfoundland Sound Symposium in 2014.

Although the mechanism just comes out of the piano—you can put it into a different piano—it was too heavy to ship. In order to get the thing onto the plane, I had to build another one out of aluminum, which would be lightweight and small enough to fit.

[After that] I finished my album, which I just put online and sort of announced the instrument out to the general world with my website. That was last fall. I also applied for a grant, and just three weeks ago they said that I got the grant to present funding for building a new one.

Has anyone else recorded or performed with the Spectrum Piano?

Unfortunately, no. Part of the grant, however, was that I make one mechanism that I can hand off to somebody else to do so, and I like that idea. I will either give it to a couple composers, or I might pass it off to the University of Victoria to see what they could do with it. But I’m perfectly open to the idea of other people working with it.

Could you envision it as a commercially available product?

People ask this. I don’t especially want to be a manufacturer. Absolutely, I could, but can you imagine launching into that?

You could license it to an existing manufacturer, couldn’t you?

I could. I would be open to that. This is obviously what I should do. But then the question is how much of it is patentable and how much not? What’s really original about it is the use of the summation of all the strings together as a collective series of harmonics [that] allows me to turn it into a synthesizer and produce those sounds that I have not heard off of any other electromagnetic instrument.

Are those heat sinks I see in the photos? Does it use a lot of power?

There are two big heat sinks along the top, and then below that you see 24 electromagnets. I am running 24 amplifiers at the same time, and the amplifiers are not normal amplifiers. If you put a speaker on these, your speaker will expire, because they’re suicide amplifiers. Those do take a fair amount of power in the sense that the heat sinks get pretty hot. When I did these performances, it might be 45 minutes. I would have to adjust certain parts to make sure the whole thing didn’t overheat, because if it does, the individual magnets will shut off. The other thing is, if I push too much power into the electromagnets, the sucker will overheat and burn up.

What makes the technology behind the Spectrum Piano different from sustain technology developed for guitar?

The way an EBow works is it senses what frequency a string is vibrating, and then it sends out a sine wave at that frequency to sustain the sound. It takes a while to pick up the sound and figure it out. But with the Spectrum Piano, I know exactly what the strings are tuned to already, so I don’t need that. I can hit it with something right away and know that it’s the third harmonic on the 27th string without having to do any analysis. The strings in a piano just sit there, and I can excite them all in an array at the same time, rather than just exciting any individual string. I can excite them [in a way] that uses 18 strings to make one sound and make one note, as opposed to exciting one string.

So, you can create much more complex sounds?

I would think so, because I can do anything to excite an individual string—hit all the harmonics you want and do all that sort of stuff. But at the same point, with this one I have 24 strings, and the next one will have 48 strings, and I can use any combination of them all at the same time in order to fuse together like an orchestra, rather than one player. So, you can have one violin player, where here I have 24 violin players, and you can make a lot more sounds with 24 players doing different stuff than one magnificent, unbelievable player, which is more how I saw those guitar ones as, not that there’s anything wrong with that. They’re fantastic. It’s just two different animals.