Mastering Abuse Sucks Dynamics From Digital Material 20May20

Referenced in some email threads is my dive into why I prefer "the sound" of some albums over others, when all other things should be generally equal. My research led to understanding the "Loudness War" that's been raging largely since the CD was introduced. The Wikipedia article linked is very good. Basically though, during mastering, a recording engineer can compress and equalize tracks with digital signal processing - making louder parts quieter and quieter parts louder, and deliver more instances of peak amplitude (of which a CD is only allowed "a few"), making a track "louder" overall. The effect often for me however is that dynamics and timbre have been mastered/compressed out of songs in the studio and I get a deader overall feel. This effect is measurable (more later) and a clear illustration is ZZ Tops' recording of Sharp Dressed Man which was released 3 times using the original, with some remastering done for each release.

Hmmm... nothing can be done about this unless an artist or producer or engineer chooses NOT to make these tradeoffs, and some do choose that path. Dire Straits stuff under Knopfler often had great dynamic range. Brother's In Arms was released when loudness manipulation was well underway and it is awesome, as almost all his releases are.

Unfortunately in many cases louder is the "average" listener's preference and bands are motivated to sound loud on the radio, since they are in comparisons. This might also be at the core of vinyl's resurgence; some say records sound better to them - and it very well might be better because an engineer has many fewer chances to equalize master material in a more purely analog domain. There is a keen database of albums and tracks I found that was produced using standardized measurements of audio dynamic range... it's cleverly called Dynamic Range Database and you can search for artists or albums and folks have uploaded their measurement of material. Supporting the vinyl DR notion I looked up Vampire Weekend and show measurement of their different albums and release media below. Red is low DR, yellow is mid DR and green is good DR in this database - it's been dumbed down for easier and quicker interpretation. It's great.

The database was largely compiled from measurements done using MAAT's software. I used their DROffline Mk II to measure some tracks and albums myself and it's fast and simple.

There were not standards on how to measure a track's or an album's dynamic range for many years, but now there are and MAAT's pro audio software measures to those standards as well as does an interpretation of goodness on a 0 to 14+ scale (as shown in the DR database). I used another app from Sweden's LTS called MasVis which also analyzes tracks regarding dynamic range and delivers output which is analytic and comparative - it's only available for Windows however.

Now I just need to figure out how I am going to behave, listen, buy and alter my own equalization practices. Bit depth & sample rate are insufficient to determine the "quality" of a digital source recording, we need the dynamic range published too - at least you can look up what others have measured about the material you intend to buy. This all ignores streaming which is much lower quality on published & unpublished metrics... my system @ the Nest renders stock YouTube videos decently - but could do much more if the source material was better.