Making Copies: Digital Perfection? Part I

So we know that storing analog sound can be done using a variety of methods including lacquer platters, analog tape and vinyl LPs. The entire music industry was built on these formats and the fidelity of our listening experiences has advanced since the original cylinders that Thomas Edison hand cranked well over a hundred years ago. But the production process of making commercially available recordings requires copying. Audio engineers have to make transfers during the preparation of the production master and then of course, there’s the mass replication that’s required to get the products into consumers hands.

What if there was a method that could be replicated without any loss? Imagine that we could capture the original music making, do all of the overdubbing, mixing and mastering without reducing the fidelity along the way. Of course, I’m referring to digital recording and playback. This is not to say that digital is without its problems but it does mean that a source digital file can be copied exactly. This is a huge advancement for the science of audio recording and has forever changed the landscape of both audio production and the consumer experience.

What is contained in a digital file and why is it possible to make perfect copies? The essential concept that I try to explain in my audio engineering courses is that the information stored in a digital file is not the actual sound or a model of the sound. If you look through a magnifying glass at the grooves pressed into a vinyl LP you can actually see an “analog” of the waveform undulating back and forth. If I printed out the ones and zeros in a digital sound file (it doesn’t matter if it’s encode in PCM or DSD format), you wouldn’t be able to see any “musical” patterns or waveforms. There is nothing in a digital sound file that relates directly to the sound that the musicians created. It has to be turned back into the analog signal using a DAC.

Instead the information stored in that file represents the “blueprint” necessary to reproduce a completely new iteration of the original acoustic sound.

Lately, I’ve been spending too much time on a side project that is a 3-dimensional puzzle. I’m trying to figure out how to make lots of copies of a very cool project that I originally conceived and produced over 30 years ago in a wood class (I’ll invite you all to see the item in an upcoming Kick starter campaign!). Anyway, the original 10 copies that I made were done by hand using a milling machine. They are beautiful but and if you looked at them you think that they are “identical”. But they’re not. They are all slightly different…not perfect copies. They were produced by hand.

My new effort is to make lots of copies that are exactly the same. So I’ve taught myself to use AutoCAD and have created a “digital” representation of the puzzle in the memory of my computer. It is a digital set of X, Y and Z values (ones and zero) that when interpreted in the right way; describe the length, width and height of the puzzle. The next step is to send the file to a production shop that can physically create the puzzle using fancy new stereo lithography machines…3D printers. After the prototype, they will be injection molded in mass quantities (I hope). Each and every copy will be absolutely identical. They won’t be copies of a master…they’re actually all first generation masters!

So if I store the instructions or blueprint of a sound file instead of an approximation of the waveforms cut into a piece of lacquer, it is possible to recreate the sound as a first generation “master” each and every time. This is a tremendous step forward and it is now the norm for manufacturing, architecture, film and music. It’s allowed audio engineers and the music industry great potential fidelity than every before.

But it does require that the resolution of the “master” digital information be high enough to ensure that we got all of the fidelity of the music in the first place.

I’ll talk more on that tomorrow.