Ogg take more battery?

Will ogg take more battery life from the Fuze? I like ogg, but I don’t mind mp3. I just want the longest battery time possible.

I’m wondering because I know that ogg is more complex than mp3 and on most mp3 players, it does shorten battery life. I’m wondering if this is the case!

Message Edited by gabe565 on 10-04-2008 01:13 PM

Ogg is more computationally comlplex than MP3, so it’s expected that battery run time would be impacted.  Posts on the Clip forum indicate it might be around 10% difference, but I haven’t seen anything resembling a scientific analysis, so you’ll probably have to try it yourself to find out.

BTW, FLAC is computationally simple to decode (compared to lossy formats), so it might even have a positive effect on run time.  Although most people probably won’t want to sacrifice the required storage space for a little longer run time.

Actually,  FLAC drives more data (I/O)  thru the system and uses more power than OGG.

@sansafix wrote:
Actually,  FLAC drives more data (I/O)  thru the system and uses more power than OGG.

Obviously, with spinning hard disks that’s the case.  I didn’t realize that would be true of flash memory.  Isn’t the flash memory powered on all the time whether it’s being accessed or not?  Or maybe it’s related to power management of the memory controller itself?  Anyway, interesting to know.

It would be really interesting to see some tests comparing battery runtime between the various supported codecs.  Anyone out there with too much time on their hands? :smiley:

BTW - It’s pretty cool that the forum spell checker always suggests “assassin” for “sansafix” :stuck_out_tongue:

The flash memory is power managed like everything else on the player.  Writing uses the most power.

@skinjob wrote:

[snip]

BTW, FLAC is computationally simple to decode (compared to lossy formats), so it might even have a positive effect on run time.  Although most people probably won’t want to sacrifice the required storage space for a little longer run time.

Most modern architectures run memory at a different (and usually much lower) clock speed than the processor, even when the memory controller is on-die and technically able to run synchronously. This is the reason why, for example, lookup tables are generally less efficient these days than just recalculating values: performance is mostly dominated by how fast you can fetch from RAM, not how fast you ALUs crunch math and logic operations.

FLAC is definitely computationally cheap to decode, but as sansafix confirms, I/O is the biggest power draw. Decoding one second of audio requires far more compressed bits to be copied from flash storage to work RAM if you’re using FLAC than Vorbis or MP3. Even a format with computationally heavy decode requirements would mostly draw power because of the number of RAM reads and writes than how much actual number crunching is performed.

(as an aside, the storage and battery tradeoffs still don’t offset the huge convenience of being able to play archival-quality FLAC files without having to transcode during the copy-to-the-player phase. Bravo, SanDisk, and thanks)

All very interesting points.  Never really thought about the I/O aspect of it.

That makes me wonder more about Ogg Vorbis.  Everyone always says it reduces battery run time because of the increased computational complexity relative to mp3.  However, at a given quality level Ogg should also be smaller than mp3.  So I wonder if the CPU penalty is offset somewhat by reduced I/O with regard to battery run time.

Hmmm, maybe someone needs to actually test it to see… I always thought that lossless took less power, but I guess I was wrong!