MP3 vs. WMA: Understanding Bit Rate
June 6th, 2006 Jason Dunn
Much has been said over the years regarding MP3 and WMA music formats, and as a result there’s a lot of confusion about the strengths and weaknesses of each format. If you’re going to rip your own CDs, what format should you use? And what bit rates? Better yet, what does bit rate even mean? You’ve probably seen “128kbps WMA”, but what the heck does that mean? Here’s a quick primer on these two formats, starting with a discussion of bit rate. And for the sake of keeping things simple, I’m only going to cover MP3 and WMA – support for other formats, such as Ogg or AAC, aren’t widely supported across Windows Mobile devices, though there are some third party tools that enable support for those formats. The discussion on bit rate applies to these other formats just the same however. Lossless is another topic entirely - Damion Chaplin at Digital Media Thoughts has written an article on the topic.
When you listen to a CD, you’re listening to digitally encoded music. The quality of a digital sound file is measured in bits – 0’s and 1’s – the more bits, the more information there is, and the more information, the more sound there is. More sound means the digital audio is closer to the original recording, and when you’re listening to a song, you want it to be as close to the real thing as possible, right? So that CD you bought has a really high bit rate – 1410 kilobits per second (kbps) to be exact. That means that a CD, which is the consumer benchmark for audio quality today, dishes out audio at 1410 kbps (remember that number for later). Audiophiles will tell you that DVD Audio or SA-CD is the real high-water benchmark for quality, but considering those formats have been colossal failures with mainstream consumers in terms of adoption, I don’t consider them to be pertinent to this discussion.
So if a CD is 1410 kbps, why would we drop the bit rate when creating WMA or MP3 files, and thus the quality? Storage space is the answer why. Doing some quick math tells us that a 1410 kbps song requires 176 kilobytes per second, or 0.176 MB. Figure on an average four-minute song, and we have an audio file that needs 42MB of space. 42MB per song, multiplied by 15 songs, and we have 630MB, which is nearly the capacity of a CD. It all makes sense now, doesn’t it? But do you want to be able to fit only twelve songs on your 512MB digital audio player? No, I didn’t think so – that might not even be a whole album. So in order to make those big 42MB songs smaller, we compress them by lowering the bit rate. We toss out the 0’s and 1’s – we actually remove parts of the song in order to make it smaller. So what does a song sound like with parts missing? That’s the magic of psychoacoustics – and what I’ll be explaining next!