iPod Battery Life, with Rockbox
2008-10-05 19:56 - Tech
Rockbox is open source firmware for various PMP devices. I've known about it for a while, and very recently they've fully released (after quite a long time) version 3.0. I won't go into detail about Rockbox, information is easily available.
One thing of interest, however, is how Rockbox affects battery life. Besides all the potential wonderful things Rockbox can do, it has (had?) a reputation for draining the battery much faster than the stock iPod firmware. So, I decided to put it to the test.
I set up the onboard sound (which I never previously had a use for) on the linux server at home. I figured out this particular command:
arecord -q -d 5 -t raw -D default
Which records the raw bytes data the sound card, 8 bits per sample, to standard out. A quick C program summed up and averaged these bytes. Taking the absolute value gave me an approximation of the overall volume. I was able to leave the iPod playing unattended, and log exactly when the sound started and finished.
Turns out my quick and simple approach was very prone to false-negative, from the quiet parts of songs I expect. Nevertheless, from a simple shell script which repeatedly took measurements, I got a first "start" and last "end" time. I let the iPod fully charge, hooked up to the wall adapter. Then, with shuffle and repeat on, to most closely mimic my usual listening habits with no chance of stopping, I let it play.
That's right, Rockbox lasted just over two hours longer than the stock firmware. Now, the stock firmware has "sound check" turned on and all sorts of options tweaked, where Rockbox was running at the defaults. I'd probably configure some power draining features on if I really used it, but it certainly isn't worse! Remember however that I've just got a single sample point for each version here.
Unfortunately, the iPod remote built into my backpack doesn't work with Rockbox, so that's pretty much a dealbreaker for me. But batterly life certainly isn't a reason to pass it up.
