Mastering limitations

We've certainly struck a chord with this current topic, the sound of vinyl. Lots of good commentary and of particular interest to me, lots of great input from my readers. Thank you. One of the joys of writing this daily blog is how much I learn, how much I am reminded of what I've forgotten over the years. Great input, thank you, thank you. I want to start by being clear on my views on digital and its superiority over vinyl in technical terms. Digital has more dynamic range available than vinyl or tape. That is simply a matter of fact. Further, if we use a high sampling rate for either DSD or PCM, digital also has the possibility of extended bandwidth, beyond the capabilities of tape or vinyl–not to mention flatter frequency response and ruler flat linearity. Again, just stating the facts. But, here's the thing. All the wonderful technological advantages of digital does not mean it is better. I want to be clear about this because it's easy for folks to jump to conclusions. Else, they might say, why else would I bring up these facts if I were not convinced it was a better medium? I just want to be clear on what we're saying here. No one benefits from foregone conclusions–pro or con against the question at hand. Our goal is to learn, to share info back and forth, to further the cause of high end audio and our search for musical truth - in whatever form it may present itself. Fair enough? It has been suggested by reviewer Michael Fremer (in our comments section) that mastering engineers don't necessarily compress recordings when mastering vinyl. This is a fact, and each has their own way of getting what they consider best sound. As I had mentioned yesterday, if the transfer is from analog tape, no compression or limiting is needed. Tape and vinyl have about the same dynamic capabilities. But when the original recording has dynamics that exceed vinyl or tape's ability to capture them, as is often the case with digitally recorded material, the vinyl mastering engineer is faced with some tough decisions:

• Set the maximum level and let low level details get lost in the noise of the record (not often done)
• Turn the low level up (out of the noise) and peak limit the louder parts to fit (this is often done by mastering engineers concerned with good sound)
• Dynamically turn up and down the low levels with an AGC (automatic gain control) and peak limit the louder notes (commonly called compression)
• Turn up the low levels of music and let the peaks clip, using forgiving electronics (like tube gear)

And then there's combinations of all the above. But, one way or the other, if the original source material significantly exceeds the dynamics of the storage medium, some combination of "compression" (as detailed above) has to be done–or ignored–with obvious sonic consequences either way. Why is this important? Because any of the above methods changes the sound, and gives us another piece of evidence in our search for understanding. Tomorrow I'll point out a number of frailties in the digital audio chain - and you might be surprised at what I have to write.