Convoluted Thinking

Over the past few months, there's been a certain amount of nervousness in the music software industry, with software piracy pushing small companies to the limit of bankruptcy, or beyond. Every week new stories emerge — one software company reported something like 90 cracked copies in use for every one purchased. Interestingly, the big record companies in the US have decided to make a stand against record piracy via MP3 downloads by claiming punitive damages against a number of end users as a deterrent. Technically speaking, there's nothing to stop music software companies doing the same thing, so think twice before downloading a chunk of cracked code — the software police might just be watching you!

Regular readers of this column will know that this isn't the first time I've dealt with the thorny subject of software piracy, and I've no doubt it will not be the last, but what particularly interests me this month is a whole other sort of 'piracy', one that seems to be perfectly legal. If somebody copies a section of your record without permission, they're clearly in the wrong, just as they are if they copy your software, your lyrics or your melodies. But where does the law stand in relation to the new breed of convolution-based signal processors that can measure and accurately replicate not only the acoustic signatures of real rooms but also the exact sonic characteristics of existing hardware processing units?

Whether the user is doing anything wrong in taking an impulse response of their favourite vocal plate setting from a rich friend's Lexicon 480L is open to question, but do you think it is ethically correct for a manufacturer to sell a convolution-based device (whether hardware or software) that comes with a library of presets taken not only from vintage equipment, but also from their competitors' current hardware models? I'm not coming down on any particular side of the fence you understand, I'm just posing the question. It's not only reverbs and echo units that can be digitally 'cloned' this way — this technology and its increasingly ingenious adaptations can seemingly be applied to mic preamps, equalisers and even dynamics processors. In fact, I don't know why guitar amps haven't yet been recreated even more accurately using this technology — though I'm sure we'll see it happening before too long.

Should we be asking the same question of physical modelling now that companies are building devices or plug-ins to model big-name guitar amplifiers, stomp boxes or even guitars, or software synths that emulate vintage analogue instruments? I'd say not, as physical modelling can only ever be an emulation. The designers have tried to get their sound as close as possible to the real thing, but they've done it by designing their own algorithms and then fine-tuning them in comparison with the original device to get the best possible results.

Convolution, on the other hand, allows the exact characteristics of an existing device (albeit for only one set of front-panel settings at a time) to be transferred into a computer or a bunch of DSP chips and reproduced. And if it really takes off (and how can it not?), what incentive is anyone going to have to design any more great-sounding signal processors in a future, when all their competitors have to do is run a sine-wave sweep test signal through their unit and, in effect, clone it for free in just a few minutes? It's my guess that we'll see some interesting legal test cases along these lines in the future, but, as with most new technologies, it can't be uninvented, so we simply have to learn how best to live with it.

Paul White Editor In Chief

Home | Search | News | Current Issue | Digital Editions | Articles | Forum | Subscribe | Shop | Readers Ads

Advertise | Information | Links | Privacy Policy | Support