The Week of Deep Apple Electronic Music Patent Mysteries continues! Behold as Apple submits a patent for — as near as observers can tell — detuning oscillators with common beats. Let’s switch to synthesis 101 for a second. Detune two oscillators, and destructive interference between them will create beats. Apple’s patent claim: “The present invention relates to a music synthesizer and a method of generating a synthesizer output with a constant beat.”

1. A method comprising: generating a constant beat parameter; adding the constant beat parameter to a pitch signal to derive an input for a first oscillator; and combining an output of the first oscillator and an output of a second oscillator to generate a music synthesizer output with a substantially constant beat.

2. A method as in claim 1, further comprising: deriving an input for the first oscillator from a linear pitch control signal.

3. A method as in claim 2, further comprising: combining a detune parameter with the linear pitch control signal.

Patent: Music synthesizer and a method of generating a synthesizer output with a constant beat

The patent, in other words, covers a basic synthesis technique, and that’s raised some controversy as this patent has spread online among synth geeks: why would anyone want to patent something so basic?

If you detune by an amount in cents (or any other fixed interval), the rate of the beating will change proportional to the amount of the original frequency; that is, your beats will change in frequency as you change pitch. This is especially amusing to me, because before I read this patent, I was explaining this very property to my Computer Music class at Brooklyn College. Happily, Chad the Physics major helped remind me how arithmetic works, and we got our detuned oscillator equation working perfectly. Normally, you want to maintain the interval of the detuned oscillators and not the rate of beating, because that’s the traditional detuned synth sound to which we’re all accustomed.

The situation described in the patent, while not as common, isn’t new: it’s natural to want to use the beating as a rhythmic device and keep its rate constant as you change pitch. In fact, this was the feature we “fixed” in our Max class example. It’s much easier to translate up and down by Hertz in digital synthesis; it’s just less typical.

“Less typical”, however, does not mean you should run out and patent it. I checked with a colleague as I first heard about this, and he immediately mentioned Kurzweil synths.

In fact, this has even confused people on the Core Audio API Apple developer list. They’ve singled out specific Kurzweil models and, prior to that, even modular synths used for the same technique (plus the obvious ability to do this in modular synthesis software like Max, Reaktor, Csound, and so on).

How simple and banal and obvious is this? In the judgment of the developer list participants, it’s “give it to your class as a work problem” simple … “patenting breathing” simple.

I’m not sure Apple’s patent lawyers have some evil plan here. This is so banal, no one is likely to go to court over it. In fact, if anything, it seems like a waste for Apple to even bother with a patent application in the first place. It certainly wouldn’t hold up under a legal challenge. Or else we’re all missing something fundamental. (But the patent is very explicit about what the result is, so even if the method were novel, it’d still be useless. It’d be like inventing some wild, new technique to sync oscillators.)

I’m curious: does anyone have any idea how this stuff happens in the first place? Is the legalese so thick that when you describe flying cars, it comes out as oscillator detuning? (Well, except that this isn’t legalese, and many of us understand every word it’s saying, and it’s wrong, because this isn’t new.) Is there some sort of bizarre miscommunication going on between developers and lawyers? Should I go patent my next Max pitch transposition patch?

As regular readers know, I’m someone who’s happy to learn I’m wrong. So if someone can explain what seems like utter nonsense, I’m all for it.

Thanks to Matrixsynth: Apple Patents New Synthesizer Method

  • anon

    This is hardly unusual…

    Not to mention the ridiculous cost of filing a patent precluding the general public from filing them at all.

    The whole system's a total sham.

  • They might be ensuring that some future device of theirs will not result in a patent suit or "we did it first" scenario. I doubt that they are looking to take on the synth industry.

  • flintwall

    The patent filed acknowledges 'prior art' and the extract is thus a pointer to some future feature/product under development which is of course not mentioned – all quite normal in this weird world.

  • Francie White

    Just like with the iPod, you'll find out, tee hee!

  • bliss

    I'm thinking that they are just doing it as a precaution to avoid litigation rather than to pursue litigation.

  • Well, just because it's submitted doesn't mean it will get approved. In order for a patent to be approved it has to be new, useful and not obvious. So, hopefully the USPTO has at least one synth geek on hand to reject this one.

  • Adrian Anders

    And let's face it. Apple DOESN'T need to have the patent stand up in court… it just needs to have it so that it can extort licensing fees from small companies and private individuals who don't have the money to fight a lawsuit in court. The same as 100s of other bogus patents that have cropped up these past few years.

    Another practical example that relates to digital music: The patent for zero-latency convolution reverb is held by a company (I forget which one) that effectively has eliminated any possibility of seeing a free/cheapware zero latency convolution reverb while that patent is in existence. I think that patent is a pain for music software development, and not worth protecting even though that software method is pretty damn non-obvious. Now take these complications, and add it to something a bit more common like synthesis techniques. Pretty soon we might see Synthedit "developers" breaking so-called "protected ideas". But one has to consider if a technique can be reproduced in something that's as simple as Synthedit, a program that existed BEFORE the patent was filed, by someone with only a moderate amount of mathematical knowledge, then the idea must be pretty damn obvious.

    Try convincing the USPTO of that though. If they can't figure out obvious patents in regards to something as common as "text-messaging" or "internet applications" etc. then they have ZERO chance of understanding something as esoteric as synthesis methods. They'll just rubber-stamp that bad boy and leave it to us to handle the consequences of their broken system.


  • Brendan

    Check out brainwaves and binaural beats, yo. This shouldn't be patentable, IMO.