Interesting replies to the "voice quality" claim. I concede that
voice quality is simply part of a larger gestalt (GIZ?) that goes into
the recognitions process and that, consequently, it's hard to argue that
it alone (even for voices of friends on the phone) is the key. More than
likely some hearers, when in doubt, begin to focus more on one thing than
another. Another concession: many birds have more than one quality
register - scolding wrens are in a different 'voice' than singing wrens,
for instance - and thus the 'test' suggested by Chris of having people
listen to one of the lesser known vocalizations of a species could surely
be designed to trip "voice quality" hearers up. Of course hearers who
focus more on pattern or who quantify are going to have problems with
rare vocalizations as well.
I haven't tried sorting out "Hoo" is "Hoo" at night much, but I
can't imagine it would be any different from sorting out all the unseen
birds one hears during the day.
To turn the thread into a segue: size/distance, yes! To me the most
difficult feature of identification. To tell a story on that, when I was
in the area of Papua New Guinea where I do research, I found myself
beginning to believe that there were two nearly identical types of
triller present: a tiny one and a medium-sized one. A distressing
conclusion, but I was new to birding then. I let it pass until one day it
began to dawn on me - after I had become more attentive to song - that
the "tiny" triller had an identical song to the medium-sized one, but it
always emitted this song *very faintly.* Hmmmmm.

Harriet Whitehead
Anthropology, WSU