What exactly does "compressed" mean?

[Actorman]

I have been reading all the posts on this forum debating the various issues around digital downloads and iTunes. One thing that keeps getting mentioned is that the files on iTunes sound "compressed." Can someone please explain in plain English what this means?

I've only bought a dozen or so songs off of iTunes and they were songs I didn't have in any other format where I could do a side by side comparison. However, I have used the iTunes software to put almost 11,000 songs from my personal CD collection onto my iPod and I don't really notice any difference there.

Is this more of a "technical" kind of thing or is there an actual, audible difference in the sound that I'm just not hearing (or don't know how to listen for)? And is it only with the downloads you buy online or does this happen when you transfer a CD as well?

 

[Mike Blakesley]

I suppose it depends on how good of a system you have, but I think compression is most notable in the high end. It basically is a technique to reduce the amount of data required to store a song. After the song is digitized, the file is run through software which removes bits here and there that it deems unnecessary for whatever reason. The result is a much smaller file, but many people say it compromises the sound.

I've downloaded quite a few songs from iTunes and they sound just as good to me as my "regular" CDs, but then my system is not exactly the top of the line. I'm using 30-year-old Akai speakers at home (although everything else is 5 years old or less) and I do have a pretty decent Pioneer system in my vehicle, but then you have road noise and such to contend with.

In the case of these TJB tracks, it's hard to compare because WARM has never been issued on CD, THE BRASS ARE COMIN' has always had a bit of a controversy surrounding its sound due to the processing, and VOLUME 2 has the oft-mentioned tinniness that Herb doesn't like, which has variously been attributed to the studio, his trumpet, or the engineering. Add to this the fact that the masters are between 37 and 43 years old, and you have a real chore to figure out if any loss in quality is due to the compression, or due to the tapes themselves.

As for ripping from CDs: It depends on what format you rip to. If you rip to the "lossless wav" format, there's no loss in quality because you're making an exact digital copy of what is on the CD. If you rip to MP3 format, which is a much smaller (and yep, compressed) format, there IS a loss in quality but there are different quality levels of MP3 to rip to, so it depends on your choices and your needs. If you're only listening on a car stereo in a noisy dump truck, you could cram hundreds of songs on a CD and it would probably sound great to you, but if you're listening on a high-end system at home (and have good ears), you'd probably be driven crazy.

But, keep in mind that the CD format is itself a compressed format, which is one reason many LP listeners think that LP is the only acceptable format.

 

[Harry]

By its very nature, sound is a very complex thing, with tones and overtones that define it and make it unique. Capturing sound onto any "device" has always been about trying to achieve the most natural sound using unnatural equipment.

Compressed music is what happens when you take a recorded work and strip out the things that human ears supposedly cannot hear and will not miss. But many agree that even though the human ear might not hear a certain tone or overtone, its presence is part of what makes up that natural sound.

Analog reproduction (LPs, tape formats) captures a great deal of those intangibles but suffers from the noise inherent in its format by nature. LPs have the noise of the vinyl being touched by the stylus, along with the clicks and pops of the vinyl's imperfection. Tape, has hiss to deal with - the alignment of the magnetic particles in the tape itself aren't perfect and have a background hiss associated with it. For years, engineers spent time and money trying to make quieter vinyl, and find tape formulations that weren't so hissy.

Then along came digital recording. It's all just a sequence of 1's and 0's that tell the playback mechanism what to do. The rate of sampling the original sounds is what affects digital media. The more you sample, the closer you'll get to that natural sound, and by definition, the more 1's and 0's you'll need.

That's where compression comes along. Compression strips away the 1's and 0's that supposedly aren't needed by humans to properly hear the reproduction. The smaller that one can make this "file", the easier it is to store, transmit, and copy on things like computers, iPods, and the Internet.

There are many different formats of compressed files. An mp3 is a generic version that by itself can have many different sampling rates. Higher rates approch more natural sounds but use up more "space", while lower rates make smaller files and strip away more of the detail of the sound. It's all about compromise. What's the smallest file I can make of this song and still not deteriorate the sound quality too badly?

Someday, when everyone has huge device capacity and broadband access, I predict that compression will start going away in favor of better sounding and bigger files. But for now, it's all about squeezing 10000 songs onto a gigabyte iPod.

Harry

 

[Actorman]

Mike, it's actually a cement mixer not a dump truck! LOL! (Just kidding!)

Seriously though, thanks to both of you for those great explanations! Very detailed, yet easy to understand. I really appreciate it. I've been wondering about that for quite awhile now.

 

[Rudy]

There are two types of compression: lossy, and lossless. Lossy compression does what Harry and Mike describe above: bits are "thrown away" to make the data smaller. Sony did this with the MiniDisc (using ATRAC compression)--like others, it works on the principle of throwing out data that supposedly won't be heard or missed. MP3 is one of the more popular ones. Apple uses AAC for the iTunes files. Lossy compression alters the sound characteristics to make the file size smaller.

For lossless, there is Monkey's Audio (APE files) or FLAC, among others (including one by Apple, which works in the iPod). Lossless compression means that the original WAV file remains intact, but the data file itself is compressed. Ever use a ZIP file? Same principle. This type of compression just operates on removing redundant bits in the file structure, and uncompressing the data results in the exact WAV file you started with.

I sometimes use APE or FLAC to back up my CDs as data files onto a DVD. Typically, the files compress by 50%. If the CD ever gets trashed, I get a bit-perfect backup off of the DVD, burn a new CD, and I'm good to go.

 

[Vinylalbumcovers]

On the file compression end, there is also Windows Lossless and Apple Lossless. Both are mathematically lossless and both sound pretty darn good - at least to my ears.

The...er...technical part of compression is done at the recording/mixing/mastering end and is completely different from the kind of compression mentioned in the other posts. This kind of compression involves reducing the dynamic range of a sound to make it more "even"-sounding. If used sparingly, this can be a good thing (vocals, brass, etc.). If used relentlessly (as it is on most contemporary recordings), the results sound somewhat more immediate, to be sure, but also much more "squished". Music should have quiet and loud passages. Compression does away with the quiet passages so that they're all just loud. How that became standard operating procedure is entirely beyond me.

Ed

 

[Dave]

Bring Expansion Back!

Expansion Rules!

We Want Expansion in our Sound!


Expansion in DIGITAL Audio--Yes!!!​

Dave inkshield: