Interesting video. I also thought it was bold to post 3D models on the web. Although the article says the models are not useful without access to a high-end 3D printer, I disagree. Why do I need to print those models in the physical world? Couldn't we create a "virtual cartridge and needle" that tracks the grooves and recreates analog signal encoded in them?
Of course this is a crazy and inefficient way to copy much less distribute music (guessing the 3D groove models are huge) but it's naive to think the models are worthless without access to a high-end 3D printer.
I have a demo CD they made for people who are interested. I have to say, I'm rather impressed. If I had >$1m in the bank, I would definitely invest in one.
That's the physical world. Besides, somehow this doesn't sound like it would suit the purist who believes everything digital taints the sound. Although I didn't do any research, if I had to guess, the laser samples the grooves and those samples then go through a D2A of sorts. That would be enough for analog purists to avoid it at all costs.
I was suggesting the 3D models (still in bits and bytes) could be interpreted by a "virtual turn table cartridge" which could decode the variations of the modeled groove. There's no lasers. This is just software. Of course the laser turntable sampling techniques and software could be fed virtual groove samples from the 3D model rather than the actual laser-sampled ones -- if that's what you meant.
I guess it depends on your definition of 'useful' since, rather than the 3D model, you might as well just post the original data. No doubt the music can be recovered through software and the posting of the models would constitute distribution in violation of copyright law. The same holds true for an MP3... it is encoded with an algorithm. The MP3 data is not copyrighted, the decoded music is copyrighted but people are still sued for distributing MP3 files.
Strangely, the algorithm determines what you get from the input file just as much as the input file itself. A set of data that decodes to one song with a particular algorithm could decode to a completely different song with another algorithm so who is to say which song the file represents? A real life application of this would be to split an MP3 (or AVI) file into 2 parts... alternating odd bytes and even bytes. IMO, neither of the 2 files produced would be copyrighted material by themselves (singly they would decode to gibberish) nor would someone be breaking the law by distributing ONE of the files. Only by distributing both files or someone else re-assembling the two files into the original material would copyright law be broken.