I also noticed a huge drop in battery time. From 1-1.25hrs to less than 30min.

> Use (i.e. google to learn about) powertop.
>
> This should help identify the bottlenecks.

I've been checking to see wackups with powertop, and cpu usage of apps with
htop.

> While I hate to admit it, pulseaudio could be waking up your system a
> bit more often than just plain alsa. This will change however with the
> next release of pulseaudio which will use dynamic buffer sizing and some
> advanced cleverness to actually cause far fewer wakeups... less than
> pure alsa surprisingly (if I've understood it correctly!)
>
> So you can always try some metrics by disabling PA and see if that helps
> battery life. Also you can disable bluetooth and other wireless if you
> don't use them as they are real hogs.
>
> Col
>

I agree. On an idle computer with no sounds playing, pulseaudio regulary takes
2 of the top 3 spots for cpu usage (htop). artsd is also has 2 entries in the
top 6. So... between pulseaudio and artsd, 4 of the top 6 cpu users are sound
related.

By powertop, (again idle computer), pulseaudio usually only rates around 0.4%
(1.2) of total wakeups where as artsd usually has the 3rd spot with 5.8% (17.8).

All this suggests soundservers & systems are a bit of an issue for battery life.

I will do more time based testing to follow this.