DAA:
Your most excellent question has a lot to do with why I want to look closely at my MP3s.
The LAME codec (the software plug-in that GoldWave uses to perform file compression in encoding MP3s) has a variety of compression levels. The minimum acceptable near-CD quality that everyone seems to look for when swapping music files is 128 Kbps--but there is some noticeable discard of stuff at the high and low ends, and this is what audiophiles complain about concerning MP3s.
Just saying or writing "mathematical compression algorythms" makes me a little woozy, but from what I do understand, the codec has other ways to make file sizes smaller than simply stripping out frequencies, and that's what the higher quality recording settings attempt to do.
Apparently, MP3 codecs discard competing high frequency sounds immediately after a loud percussion; this is called "psycho-accoustic modelling". On the theory, perhaps, that if somebody shoots you, you'll not hear yourself scream. This doesn't bug me too much for our predator calling purposes because we don't have bass drum licks like rock-and-roll tracks.
But then there is loss-less compression, too. Imagine you had a city block full of empty lots and another one of cookie-cutter houses built on the same plans. You could go through and describe every lot, every building, and record the data for each and every one. Or you could save time and file space by saying say "this is a whole block of empty lots" or "this is a block of all the same kind of buildings" described by the cross streets... an analogy for what compression schemes that are loss-less do.
First results of looking at my MP3s shows that ones I made on the PC from .wavs using 128Kbps have been stripped of frequencies above about 17 KHz. However, later ones I've made from .wav files using the 320 Kbps MP3 setting (there's a slider on the GoldWave "Save As" menu, that you have to slide north, above the default setting to find) seem to be capable of reproducing frequencies up to 22 KHz. This is the highest frequency of the .wav standard, i.e. the highest sounds that CDs capture and reproduce.
My PoGo! MP3 device has a line-in record function and a built-in codec that allows 256Kbps samping at 48 KHz (in theory that's 24 KHz per channel, or 24 KHz for a mono recording like I use). Higher dynamic range, theoretically, than the other components of my caller. (My homebuilt is presently amp, speaker, and wireless mic limited to 20 KHz, but that's still a heck of a lot more than most commercial predator calls are capable of already.) What I'd like to determine is whether the PoGo! codec strips out freqs that the LAME codec doesn't.
I'm looking at some of my most used, PC-recorded tracks of squeaky things like live-caught baby mice and Crit R Call sequences--there are definitely harmonics and elements of these sounds that extend beyond 20 KHz. But the majority of it seems within the gray area of 15-20KHz.
(FWIW, 12-15 KHz is where the horn speakers fall off.)
Long post, but what I've noted over the years is that owls and bobcats and gray fox respond well to high-pitched sounds. The whole point of this exercise is that I want to squeeze as much out of the recordings as I can in order to photograph more of them.
LionHo
