19 April, 2007

Clinical compression

It's my son Ben's birthday today and my present to him is about 2,500 tracks on the iPod his Mum is giving him, all hand-picked to introduce him to music as I know it after my 40+ years of avid listening: Mozart, Captain Beefheart, Abba, Zero 7, Miles Davis, Underworld - everything I think he might enjoy encountering on a wonderful shuffle journey.

I've been ripping a lot of CDs in the process and I spent some time thinking about codecs. I settled on Apple Lossless this time, and am replacing all the tracks I originally ripped at 192 kbps AAC or MP3. While I was doing this I was musing about two huge negative impacts codecs have had, or are about to have, on society.

Codecs were developed because of limited bandwidth for broadcasting/streaming rich media, but they've been applied willy-nilly in non-broadcast situations where there is a limit on storage space so that people can squeeze more music onto portable MP3 players. This is another example of our culture sacrificing quality at the alter of quantity. Nobody has to compress their music to put it on an iPod: you could have one tenth as many tunes all in perfect CD quality (itself, of course, only a digital approximation) - and yet millions of people have not understood the implications of compression and have ripped their music at 96 or even 64 kbps. When you get rid of over 90% of the information in a music file, there are consequences! The gaps between the dots become larger, and your brain has to work harder to join them up. The clever psychoacoustic theory that allows swathes of sound to be removed is based on you reinserting it while you listen by dint of imagination, which takes constant effort. Richness, subtlety and depth are still lost, and there is a strong temptation to turn the volume up because the sound is just thinner.

Of course the very nature of mobile music is different from the intense, focused listening practice I grew up with (headphones in a darkened room, avid reading of sleeve notes and so on): music is now a background activity to life. Or is it the other way around for MP3ers? Either way, the reduced signal to noise ratio heps to disguise the loss of quality they are experiencing due to their brutal compression practices.

I predict that research will show that there are serious social consequences of all this listening to highly compressed music. I believe it is creating stress, tiredness, irritability and antisocial feelings because so much subconscious brainwork is being required. Also, the tendency to turn the volume up and up must be creating more and more incidences of noise-induced hearing loss (NIHL), which is already a major problem due to ear bud headphones. Any time you can hear someone's personal stereo leaking out, they are almost certainly damaging their hearing. The rule of thumb when listening is that you should still be able to hear people talking to you. Most people ignore this even if they know it. We may well be raising a deaf, antisocial generation.

That's one of the two bad social effects. The other is coming soon: a vast waste of time for millions of people as they have to go back and redigitize all their music without compression. Storage space gets cheaper every year. At home I have a terabyte hard drive with my iTunes library on it - unthinkable only three years ago. In just a few years we will see the terabyte iPod. Why would you want to squash the quality out of your music when you can get your whole collection uncompressed into your pocket? All those 64kbps tracks will look (and sound) pathetic, unless someone has the obsessive need to carry around several million tracks - in which case I suggest they go to MP3 Anonymous and get help.

From today, my son has access to over 2,000 tracks at perfect CD quality. I have educated him about safe listening levels. I very much hope that his experience is the model for portable music in the future. I urge anyone reading this to put quality first: invest in your ears and your health by buying enough storage to rip uncompressed - or if you must compress, use only lossless codecs. That way you will enjoy the full richness of the music you love, listening will be a relaxing experience instead of hard work, and your digital music will be truly future-proof.


  1. I equate listening to a lot of today's overly compressed pop music with eating too many (or any) Big Macs. They both may taste good at the time, but quickly wear you down and leave you feeling sluggish.

  2. Interesting stuff, and what a great idea for a present!

    But may I be a little controversial?

    You state, 'I believe it is creating stress, tiredness, irritability and antisocial feelings because so much subconscious brainwork is being required.'

    This sounds right, but in practice I think it conflicts with our what neuroscience tells us about sensory perception. Our brain effectively models what is around us - we just get to see the edited highlights.

    Take vision as an example - if what you were saying is true, then simply opening our eyes would make us feel sluggish because we are doing an enormous amount of unconscious work extracting turning the input into the primary visual cortex into the 'image' we see.

    There's also a huge amount of subconscious brainwork involved as we walk down the street, for example, in coordinating our motor actions.

  3. War-n, I think your analogy is spot on. Jamie, I see what you mean but I think the difference here lies in the direction of the process: our brains (from what I understand of current thinking about perception) are good at extracting what we need from background noise - though the level of messaging we get these days has caused a lot of marketing gurus to claim that we are at saturation point. However, we are talking here about putting information back in - to use a visual metaphor, joining the dots is a lot more effort than seeing a shape against a background. So I stand by what I wrote about compressed sound... IMHO it hurts the brain!


I welcome your feedback!