B
Bulls Hit
Well-known member
In his Mastering Audio book Bob Katz has a great way of describing the difficulties in explaining the audible differences between similarly spec'd audio gear from different manufacturers. He likens a piece of gear to a building with lots of windows but no doors. Each window we peer through will give us a slightly different view of what goes on inside, but because we can't walk in, we're never going to get the full picture. To me it seems a very analagous situation to quantum indeterminacy, the Heisenbery Uncertainty principle and the fact that we can never measure all the properties of a subatomic particle.
Anyway one of the things he seems particularly hot about is the area of signal degradation as a result of digital processing. Everything I do in my DAW while mixing, from a simple volume envelope to inserting a reverb plugin on a track are achieved by algorithims and calculations that result in rounding errors. Because of the many thousands of calculations that are occurring, these errors accumulate.
Different manufacturers have different ways of dealing with rounding errors, longer internal word lengths, temporary registers etc. resulting in an output signal that's affected to a greater or lesser degree depending on how 'carefully' these errors are treated. This means that all audio processing software would leave a kind of 'digital footprint' characteristic of their particular error handling regime that could have an audible effect on the output signal.
Has there been any attempt to compare the likes of Sonar, Pro Tools, Nuendo, Cubase etc. in terms of how much their native processing corrupts, deforms or otherwise affects the signal? Or is this even possible?
Anyway one of the things he seems particularly hot about is the area of signal degradation as a result of digital processing. Everything I do in my DAW while mixing, from a simple volume envelope to inserting a reverb plugin on a track are achieved by algorithims and calculations that result in rounding errors. Because of the many thousands of calculations that are occurring, these errors accumulate.
Different manufacturers have different ways of dealing with rounding errors, longer internal word lengths, temporary registers etc. resulting in an output signal that's affected to a greater or lesser degree depending on how 'carefully' these errors are treated. This means that all audio processing software would leave a kind of 'digital footprint' characteristic of their particular error handling regime that could have an audible effect on the output signal.
Has there been any attempt to compare the likes of Sonar, Pro Tools, Nuendo, Cubase etc. in terms of how much their native processing corrupts, deforms or otherwise affects the signal? Or is this even possible?
