If we take a hi-res recording, there are two main potential sources of error – two ways in which data can get misplaced. If we look at original recordings made with high-sampling rates, then like any other digital recording they are prone to errors in the recording process or technology itself. In fact, in digital terms, this is the same whether you are talking about an original recording or a digital transfer of an analogue recording – although an analogue master will embody both the errors of its recording and replay technology, be that tape hiss or limited dynamic range, age-related deterioration or discrepancies between recording and replay technology.
…but I have no idea what it means!
But if you look at early digital recordings – often 16bit/44.1kHz – then up-sampling them to ‘higher resolution’ formats involves creating interpolated data points between the existing samples. Now throw in possible changes in the native clock frequency, differences in jitter levels and the differing transfer functions of the original encoding chip-set and the chip-set that’s re-encoding the signal and the potential for error in the value and placement of the interpolated data points becomes significant. Throw in the obsession with ever-higher sampling rates pushing the bandwidth capabilities of both record and replay chains and you risk creating a perfect storm of digital degradation. I’m not talking about total breakdown here. I’m talking about bending the performance out of shape, altering its inner relationships, straight lines between the dots where there should be curves (or vice versa), a subtle erosion of the realism in recordings. Rather than real, the sound becomes flattened and plastic, smooth and synthetic, lacking texture, bite and body, dimensionality, immediacy and vitality. If you find that proposition hard to credit I suggest you take a listen to Glass CD. You’ll not find it wanting for detail, information, musical integrity or realism – and yes, it’s 16/44!
Not what it says on the tin…
I’ve been doing a lot of listening to streamed music recently. Time and again, supposedly higher resolution files disappoint when compared to the lower-res versions – or simply disappoint on musical grounds. Amongst ‘Hi-Res’ files that do deliver, DGG’s downloads are consistently excellent – yet they are ‘only’ 24bit-96kHz. Yes, the files in question are original recordings, made at (or at a multiple of) this sample rate. Yes, DGG has a remarkable roster of musical talent available – along with the usual quota of marketing department ‘faces’. Yet in the wholesale rush to embrace the ‘musical future’ these are distinctions (and lessons) that get ignored or overlooked, by manufacturers and customers alike. The reality is that sticking a ‘Hi-Res’ label and sample-rate on a download is a bit like sticking a GT logo on a car: it indicates at least some claim for performance but THAT could mean almost anything: from the frighteningly good to the genuinely frightening – and everything in between.
