It’s the time of year for saving money!
It would be uplifting to be able to write that pressure from audiophiles killed off the “loudness war” of increasingly more compressed pop music, but they didn’t. In point of fact it seems that audiophiles were not even part of the equation, really…
No, it was the streaming services that killed off dynamic-less music, here’s why…
Basically, the introduction of “normalized levels” made it so when streamed “loud” tracks with very little dynamic differentiation between the loudest and softest tracks are played they sound much worse than tracks with expanded and less compressed dynamic range.
Since ALL the major streaming services have adopted normalized levels, including YouTube, Apple Music, iTunes with Soundcheck, Tidal, and Qobuz; highly compressed pop tracks that used to sound louder when maxed-out levels were the only consideration now sound thin and gutless when streamed through any of the more popular streaming services with normalized levels.
For a detailed video presentation about the why’s and how’s of this new streaming world I heartily recommend this YouTube video of a presentation by mastering engineer Alan Silverman – he goes into much more technical detail than I will here.
Basically, there was a period of time when it was considered “good technique” by mastering engineers to make a track sound as “loud” as possible. And how was loudness defined? By overall maximum signal level. The track with the highest level, which is an objective measurement of one particular variable, such as SPL, watts, volts, or RMS, was “louder.”
But loudness is not merely about objective levels. Instead perceptual loudness is about many factors including pitch, balance, tonality, transient response and perhaps most importantly, dynamic range. This is why two tracks that have the identical maximum levels can have entirely different dynamic ranges…
As an example of what is being done now in the pop realm Let’s look at the Gammy-winning track “Bad Guy” by Billie Eillish. The song begins with an aggressive bass line coupled with an almost whispered lead vocals. Almost immediately it’s the contrast between the vocals and the bass line that sets up a strong dynamic counterpoint. I heard this track played through the three-inch ceiling speakers in my local Planet Fitness locker room…what a difference with no bass – only the vocals and the slightly goofy synthesizer line…which brings me to the fact that when heard over a good pair of headphones the dynamics of “Bad Guy” will move you.
Another clear example of how dynamics can drive a track is “My Boy” from Eilish’s previous album. The initial bass line is not only huge, but also tunefully dynamic with energy and power. This is the sort of thing that you can’t achieve from a compressed track; much of the track’s power comes directly from its dominant dynamics.
So, we audiophiles won…and we didn’t even have to raise our voices…so the next time someone utters the words “loudness wars” as an excuse for why modern pop doesn’t cut it, you might want to direct them toward listening to something new that reflects the current dynamic renaissance, such as the one of these Eilish tracks…or try “You Should See Me in A Crown” to see if your system can really take it…
Just when i was wondering about modern recording…..spot on article.
To be more specific, perceived “loudness” has much more to do with the total energy than the peak level of sound at a particular moment. A simple analogy is to think of it as the amount of sound “under the curve” of the waveform envelope at any particular instant. Looking at it this way, a waveform which has lots of peaks and dips in volume would have a lot less energy than a waveform of the same measured “peak level” and little variation in level. As an example, take a rim shot on a snare of peak digital level and a pipe organ at the same digital peak level. The pipe organ will sound a lot “louder” than the rim shot overall because it has more energy over time; the rim shot has the same “peak” energy, but it only does this for an instant in time where the organ is constantly producing sound energy. More to the point, the organ will trigger normalization based on not just peak level, but total energy. Volume normalization methods which look for energy level rather than simple peak levels of waveforms will see that compressed music has a lot more energy and thus reduce the level of these signals. A signal with lots of dynamic range has less energy over time, thus its peak levels are not limited to the same degree.
A long time ago, I made a “TV commercial zapper” which looked at energy levels rather than peak levels because commercials are compressed heavily and regular program material is less so. When high total energy levels were sensed, the audio was muted. Worked fairly well using the then-available analog techniques.
The “equivalent loudness scales” take not just peak levels but also total energy into consideration.
Good to HEAR. The only CD I immediately returned to the shop because of its disgusting sound, Amy Winehouse: Back To Black. I really didn’t care that the shop only gave me half what I paid, it was that bad. I bought very few modern recordings after that, once bitten, twice shy.
That one is indeed horrible. As are hundreds, likely thousands of others. I have declipping software; it helps a lot with some, but there may have been multiple compressors and limiters on any given track. No way to undo those…
There still is a major problem with dynamic compression in recent studio masters. Just check out the loudness war CD database. Even CD remasters in 2020, have much less dynamic range than the same CD mastered in the early 80’s. These new CD remasters sound terrible. I can’t believe this is still going on today in 2020 and the audio engineers and musicians are still ruining the music.
What’s crazy is that I think if the dynamic range had been compressed to a DR8, if that was what all music was compressed to (with the exception, of course, of jazz and classical and certain genres like this) then most of us, myself likely included, would never have even had a problem.
Even a DR7, with high-quality recording and judicious, skillful mastering, can sound quite good; but a DR6?!?!
They just HAD to push it to the point where it’s impossible to NOT notice. Sure, I prefer something in the range of 10-13 personally, but 8-9 is still quite enjoyable and lively.
If the industry simply moves to DR8s as the standard, I’ll be very happy.
I checked the DR database and Happier Than Ever the latest album from Billie Eillish is being rated DR5. One track has DR3. Which if true isn’t very good. It would love to think that record companies have started releasing titles with more DR. But it is 100% not the case. I think that a lot of people have simply got used to listing to music this compressed and may even like it. For me I’ll keep searching out the non remaster version of albums……………