... thoughts on Taylor Swift's REPUTATION CD...

dbwalek said:

" Complaining about DR and compression in modern pop is like complaining that rock and roll has too much drumming"

Totally disagree. There is nothing inherent in pop that requires, needs or benefits from DR compression. And there is no reason that pop can't be enjoyed for SQ either.

What I would like to understand is why especially voices often seems to fare so bad under modern compression/digital processing. When Taylor Swifts voice is made louder all her natural voice (which is fine) disappear and we are left with something that hurts our ears (at least my right one). It might of course be the voice itself, but I don't really think so. Very apparent on the Red cd, less so on Reputation. Is it because of some "radio-editing"? 

It is my understanding that compression is often used for voices in order to reduce sibilance (harsh 'esse' sounds). If the engineers get the DR compression so wrong, maybe they're mucking that up as well.

But to my ears harshness and stridence are the effect that most bothers me in recordings with really narrow DR. 

It is also my understanding that all music played on the radio goes through various digital algorithms to keep loudness levels flat throughout the broadcast. That's why the whole loudness war deal is so stupid. It does not accomplish anything.

On albums I like with high loudness which causes shrillness in female voices (Neko Case) comes to mind, I'll use software EQ to turn the dB down and other minor tweaks to reduce harshness and ear fatigue. All I'm probably doing is a poor version of Apple's Sound Check.

The harshness in her voice (and other artists today) is mostly due to the use of effects specifically Autotune. These pop and rap recordings are not using a sine wave to organically correct pitch, a square wave is created and the gain is pushed high into the vocals. Next add the loudness part with the use of limiting and compression and you have buzzing, electronic sounding vocals.
It's the style of pop music today.

And then each radio station adds a fair amount of compression. And for streaming (iTunes) the original files have the data compressed into a low-res version of the song.
 

For sibilant voices engineers use filters and a mild amount of compression. This is standard practice. 
There may be sibilance thru your speakers due to tipped-up highs from the mix and/or mastering.