3 ms·

The "Digital Compression" section feels misleading. Maybe that's my misunderstanding, though. As I understand it, when we compress audio file, so from a 100s M

The "Digital Compression" section feels misleading. Maybe that's my misunderstanding, though. As I understand it, when we compress audio file, so from a 100s MiB of raw wave data it'll become a several-MiB file, we perform a (discrete) Fourier transform, get rid of less significant frequencies (thus compressing the data), and record the coefficients. Please, correct me if I'm wrong. So I don't get the "rough edges you end up with in the digital recording" part. It is applicable to pre-compressed data, but that's another topic.

jerf 16y ago
While that is called "compression", it is not what is being referred to. This sort of compression takes audio that is sometime soft and sometimes lound and flattens out the volume differences in various ways. This, too, is not the crime per se, because this isn't always bad. Making something completely flat can suck the life out of a piece, but there are contexts where you aren't really "listening" to the music where it makes sense, such as background music in the stores or something, and you're unlikely to find what they are playing "full of life" anyhow. The real criminal here is clipping. Compression is only mentioned here because it lets them "get away with" yet more clipping after they've compressed it.
gabrielroth 16y ago
You're confusing dynamic range compression with digital file compression. Totally different things. The only thing they have in common is that they make something smaller.
- mambodog 16y ago
  Did you read the article? There is a section entitled "Digital Compression" which actually seems to be just describing the nature of digital representation of sound, and what happens when you reduce the resolution of that representation ie. bit depth/sample rate. Then it awkwardly jumps into talking about MP3 compression. None of this is relevant to the "Loudness Wars" topic, which as you've rightly pointed out deals with dynamic range compression, yet all these things are referred to in the article. GP is confused by the inclusion of the unrelated material. The article is sloppy and unfocused.
icegreentea 16y ago
I think it's conflating a lot of stuff. It seems like he's talking about quantization, but I think he's actually talking about the actual MP3 (or any lossy) compression scheme. It's really screwed up. Actual MP3 will result in the 'rough edges' and all the nice stuff he describes. Well, kinda. The hilarious part is that while we some error introduced that sounds kinda like noise, MP3 is actually horrible at representing 'noise'. See the sizzle on your cymbals. I remember reading this article when it came out, and I wasn't particularly impressed by it then either.

jerf 16y ago

While that is called "compression", it is not what is being referred to. This sort of compression takes audio that is sometime soft and sometimes lound and flattens out the volume differences in various ways. This, too, is not the crime per se, because this isn't always bad. Making something completely flat can suck the life out of a piece, but there are contexts where you aren't really "listening" to the music where it makes sense, such as background music in the stores or something, and you're unlikely to find what they are playing "full of life" anyhow. The real criminal here is clipping. Compression is only mentioned here because it lets them "get away with" yet more clipping after they've compressed it.

gabrielroth 16y ago

You're confusing dynamic range compression with digital file compression. Totally different things. The only thing they have in common is that they make something smaller.

mambodog 16y ago
Did you read the article? There is a section entitled "Digital Compression" which actually seems to be just describing the nature of digital representation of sound, and what happens when you reduce the resolution of that representation ie. bit depth/sample rate. Then it awkwardly jumps into talking about MP3 compression. None of this is relevant to the "Loudness Wars" topic, which as you've rightly pointed out deals with dynamic range compression, yet all these things are referred to in the article. GP is confused by the inclusion of the unrelated material. The article is sloppy and unfocused.

icegreentea 16y ago

I think it's conflating a lot of stuff. It seems like he's talking about quantization, but I think he's actually talking about the actual MP3 (or any lossy) compression scheme. It's really screwed up. Actual MP3 will result in the 'rough edges' and all the nice stuff he describes. Well, kinda. The hilarious part is that while we some error introduced that sounds kinda like noise, MP3 is actually horrible at representing 'noise'. See the sizzle on your cymbals. I remember reading this article when it came out, and I wasn't particularly impressed by it then either.