Hacker News new | past | comments | ask | show | jobs | submit login
The Loudness Wars: Why Music Sounds Worse (npr.org)
41 points by bugs on Jan 4, 2010 | hide | past | favorite | 23 comments



First heard of the Loudness War in IEEE: http://spectrum.ieee.org/computing/software/the-future-of-mu...

Didn't know it's turned into an entire meme by now: http://en.wikipedia.org/wiki/Loudness_war. Lots more links here for those who want to find out more.

Illustrative list of waveform and peak level examples: http://www.cdmasteringservices.com/dynamicdeath.htm. This page really goes to town on graphs of various popular songs to show how and where the clipping occurs.

Death Magnetic is definitely one of the best recent examples especially due to the contrast with the GH version.


While they explained audio compression well enough, they really botched the explanation of uncompressed PCM and MP3, and failed to mention 24-bit/192khz at all. The result is that one means of information compression used in MP3 (masking) is posited as an improvement over uncompressed PCM. Here's a quick summary of why your music sounds like crap tho:

reality -> microphone -> pre-amp -> digital conversion -> mixing -> mastering -> compression -- MP3 -- decompression -> analog conversion -> EQ and/or "enhancement" -> amp -> headphones -> ear

Every single arrow is a loss of quality. The studio makes sure that everything from reality to mixing uses the best equipment. Then they purposely butcher the quality to make it "pop". Next is where you muck it up based on the format you buy. MP3 has terrible detail reproduction (most obvious in cymbals), CD is okay and by far the most common, and DVD-audio is best but rare (vinyl is the most rare, and comparison with digital audio is complicated). Then, shoddy MP3 decompression aside, most consumers elect to reproduce their music using underpowered amplifiers and cheap headphones or cheaper speakers.

So, half of it is the studio's fault, and half is yours. I bet most people don't even notice that an iPod has better analog sound reproduction than any computer. But the studio's errors are more damaging, because there is nothing we can do about them. No matter how much you learn to love your music, or how much equipment you buy, you can never undo the one step in which they destroy the quality of the product.


They've confused digital compression and dynamic compression. Death Magnetic sounds worse on CD because of dynamic compression, not digital compression.


No, The 2nd para of TFA: We're talking about compression here, the dynamic compression that's used a lot in popular music. There's actually another kind of compression going on today one that allows us to carry hundreds of songs in our iPods...


Right, but then they quote someone saying that the video game version is better because of "digital domain compression", followed by a statement that "Digital compression is the process that allows a song to go from being a very big sound file in its natural state to a very small file in your iPod"


They are still talking about dynamic compression there. The music went through a different mastering process for the CD than for the game, and the game version was left with more dynamic range. His use of the phrase "digital domain compression" is misleading here, as it refers to dynamic compression performed in the digital domain.


Which is why I said it's confused.


To expand on this (and repeat what I just posted at NPR), Death Magnetic sounds worse on CD because of dynamic compression, not digital compression. If digital compression was the problem, then the Rock Band version would sound WORSE than the CD, because the CD has no digital compression (and Rock Band does). The CD is what you make MP3s out of, after all.

The CD master of Death Magnetic has had its dynamic range heavily compressed (which squashes the waveform), where the Rock Band version is uncompressed because the game mixes the song on the fly.


Also, "digital domain compression" isn't a thing.


Meaning there is a digital domain and compression can be performed there. It is still ambiguous as to what kind of compression is being described. I agree this was intended to describe dynamic compression.


The problem here is how musicians ( I myself am one ) tend to listen to tracks after mastering.

The 'loudness wars' tend to lead to listener fatigue because of the lack of dynamic range over the course of many songs. Most musicians don't listen to masters this way. They listen to the same track repeatedly to see if it is what they want before moving on to the next mastered track.

The number of musicians I know who listen to an entire mastered album all the way through multiple times before it goes to press is 0. 1, maybe 2 times at most for the entire album.

For individual songs, that big loudness boost sounds REALLY good. Its the over an entire album that is sounds bad. And especially with the return of a singles culture via iTunes et al, can you blame the many musicians who maximize for the single song experience rather than the album experience?


Frankly, albums never really advanced very far as an art form. Most albums are just a loose collection of singles tied together by fairly coincidental thematic elements.


I suspect that at this point, a somewhat dynamically crushed sound just comes across as "right" for listeners. I'm surprised that no one has launched a study to see how listeners actually react to the crushed sound, once the apparent loudness is neutralized.


For some listeners, yes.


Some of the last metallica album sounded really bad because of this.

http://en.wikipedia.org/wiki/Death_Magnetic#Criticism_regard...


As a long standing heavy metal fan I can tell you that all of the album sounded pretty bad and the compression was the least of their worries :)

It just goes to show that even VERY rich and VERY powerful bands are making a basic mistake of not listening to their masters before turning on the presses.


I've been buying more and more vinyl. It's a little expensive, but sound quality is simply superb. I spent some time trying to recognize sounds in Pink Floyd music on MP3/CD/Vinyl and realized that current technology is not yet suitable for music as sophisticated as Pink Floyd. If they could find a material that's a little harder, gramophone records might just be the medium of the future.


Actually LPs have much less dynamic range potential than CDs. It's just that most producers do not utilize that range, they instead push all the levels to the top.

I used to have a link to a study that claimed that many "long term" popular hits (music that has remained popular for decades) has a common thread of being produced with a wide dynamic range. I can't find it now...


This article also contains a very good and easy to understand explanation of the basic principle behind lossy audio compression.


George Graham reviews AAA for NPR, and always includes a note about the dynamic compression at the end of the review. Highly recommended if you like this genre.

http://georgegraham.com/reviews.html


Unfortunately, MP3's are here to stay. While some open source folks are pushing OGG Vorbis and FLAC, these technologies do not have the convenience and universal support of MP3's. The only thing that can sort of compete is AAC, but only while iPods are in majority.


More than a few of my friends have been switching to good-quality, noise blocking headphones (like etymotics HF-2s). We can turn the music way down and enjoy it more.


That's not really what they mean by loudness, but that's a good way to prevent ear damage.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: