Dave, good find, that’s a very interesting video.
I’ll stick to the main point of why the recording industry shot itself in the foot with their releases over the past decade+.
To me, the issue is not how it was recorded (analog vs digital), as both can deliver pretty amazing master tapes.
Where everything can and does go wrong is in the ‘mastering’ process for each output format.
The overuse of compression on the CD mastering is astounding, with many discs having less than 15dB of dynamic range on them, and this from a medium that supports close to 100dB dynamic range.
This is strictly a choice by the record company to ‘compete’ for audibility in cars, boom boxes, radio, etc.
To help resolve this for myself, I have a friend who has a professional recording studio, and I got him to give me the ProTools HD 24/96 stereo mix-down of a 128-track master of a 3 minute song.
He then also gave me the same song from the ‘mastering engineer’ who generated the CD master for reproduction as well as a physical copy of the commercial CD.
So I got home and had the following files to compare:
- PCM 24/96 stereo-mixdown with no post-processing
- PCM 16/44 stereo ‘CD master’ file used to create the CD
- PCM 16/44 stereo WAV file ripped from the physical CD
- MP3 160Kbps VBR converted from the ripped WAV file
All files were played back on my Denon AVP’s internal file transport (which is clock-synched to the DAC’s, so this about as ‘perfect’ as digital can get).
The results:
PCM 24/96 stereo-mixdown with no post-processing
The clear winner here. Amazing clarity, every instrument and vocal clearly distinguishable. Dynamic range to kill for (even though this was pop-music). Sounded just like what I heard on the headphones plugged into the ProTools deck.
If all music were distributed like this, no audiophile would be complaining.
PCM 16/44 stereo ‘CD master’ file used to create the CD
OK, this is where it got interesting. Lost about 10dB in dynamic range, as this is their ‘standard’ mix setting. Also, there is a bit of EQ in the mastering, slight roll-off in the highs, a bit of lift in the 60 to 120hz range. I’m guessing that’s for radio and small-system compatability.
But clearly, this is not the same thing that came out of ProTools. Very easy to distinguish from the previous version.
PCM 16/44 stereo WAV file ripped from the physical CD
There must be further downstream processing, as this one is not bit identical to the ‘CD master’, can’t tell if it’s more compression or EQ, but it’s a little but different. Hard to tell if I wasn’t looking for it.
MP3 160Kbps VBR converted from the ripped WAV file
I can usually hear compression artifacts buried in the song if I have good reference to the uncompressed version, and relative to the 24/96 track, this is night and day. One can spot this version in an instant.
But relative to the WAV file, it’s a bit more subtle, yet still distinguishable.
So HOW a piece of music makes it to YOUR system has much more to do with how it SOUNDS than the original recording format.
For instance, that 24/96 PCM stereo mix-down, put on a ¼ Tape at 15ips might sound better than the CD. (Although tape saturation will limit the dynamic range a bit and/or raise the noise floor).
Which brings me to repeat: it’s a shame SACD and DVD-Audio are on their way out, as those formats were the only ones to actually deliver the equivalent quality found in the recording studio.
But even those formats, the recording engineer can still screw up the mix: E.g the Genesis SACD re-issues (with too much compression and EQ).
So again, it’s about HOW the mediums are used not so much WHICH medium.
