I wrote this blog entry several months ago, but never posted it because I thought I would have time to go back and evaluate the sounds more. Since I don’t see that happen any time before I finish my dissertation, I just go along and post it now:
I have to say I am a bit puzzled about how boring some of the sounds sound. They resemble the XP sounds, which is probably a feature, but are very conservative and far from implementing ideas from recent research on sonification and music cognition (e.g. the Sounding Object project). Many of the sounds are based on tonal features (rising and falling intervals), and some are earcons (trash can). On a positive side, I think the new set is more coherent than the XP set as the various sounds seem to work better together. They are also somewhat less obtrusive than previous sound sets (which people tended to turn off).
Back in my undergrad days I wrote a term paper on systems sounds, where I also analyzed various computer soundscapes. At some point I will try to translate, update and publish it somewhere.