Is one bit better? We’ve had a few questions similar to this recently. “Hey, from what I’ve come to understand, the most common specs that tell you about the quality of a DA converter are the bit depth and sampling rate. Why is it then, that as I look through catalogues, practically every CD player I come across is described as using 1 bit conversion?”
Back when digital tape recorders were first arriving on the scene sampling rates were usually 44.1k or 48k and bit depth was 16 bit. Brickwall filters had to be used on recorder inputs to prevent aliasing. Of course, these filters caused all sorts of other phase and distortion anomalies with the audio so it wasn’t long before the concept of oversampling took off and converters began to sound much better due in large part to the reduced impact of the brickwall filters. One-bit systems were then introduced and they basically did all the math a different way, which for a time produced even better sound quality than the oversampling converters of the day. Crossover distortion, in particular, is really low on one-bit converters. The one-bit converters looked at a whole digital word sort of as one bit of data that was then part of a much larger set of data. All the specifics of how the math was different are really beyond the scope of inSync (though I am sure someone out there will write in and tell us). At the time this conversion technique produced very good sound quality, but by today’s standards of 128x oversampling 24 bit converters it is considered inferior. It is still effective and sometimes used in CD players because the chips are inexpensive and CD’s are still all at 16 bits.