Rich Peet wrote:
> I have not met John Feith and would be interested if you know him
> Brian.
>
> In listening to the submitted sounds I note he often constructs short
> sound files out of multiple recordings often using different birds
> within a sound file. I resist that, and would rather see multiple
> recordings of a "real song" or "call" than a recordists use of
> editing construction in this application. I also noted that he likes
> having all his posted calls at 10db down. Makes for a pretty noise
> floor but I don't get that either. I don't mean to be overly
> critical but, I am guilty of that sometimes....
>
> What do you all think of instructional/illustrational sound files
> being a construction rather than a documentation?
Well, since it's more or less what I'm doing to construct Georgia's
frogcall CD tracks and my own website tracks I can hardly be critical.
I will use clips from more than one recording of a single species to
illustrate the range of calls. It is not practical to try and find one
instance when the range occurs at one time. And there are very distinct
time limits, particularly with files made for downloading via the web.
In order to have longer time you have to compress the files more, a
sound quality tradeoff.
And CD is no better. As a example, there are 32 species of frogs we will
have on the CD, that's a bit over two minutes per species. And we also
will be providing some chorus tracks. And there is narration. It gets
crowded quickly. I'm thinking in terms of 30 seconds or less of actual
ID call. And many of those will be filtered to remove other species, or
other sounds. Ideally I'd have found a nice quiet spot with just one
species calling for each. But the truth is that I may not manage that in
the rest of my life. The chorus tracks, of course, are the "real song"
tracks. Though frogs don't come in one at a time all that much.
As far as downloading, put up a whole bunch of files for each species on
the web and most will not download them. Put up one for each species and
many more will do that. I've tracked how folks vist my site. A typical
visitor will stay only a few minutes, and look at one species. Only
about 1 in 10 listen to even one call file.
I process my soundfiles at 10db down, I don't see any problem in this,
it's about the level I record them. That still leaves over 80 dB of
dynamic range for a CD, more than enough, and will likely include the
noise floor of the mic. We rarely have something we record that comes
close to that.
It sounds like he's just being practical, especially as the website is
for learning ID. It's not documentation as such. I've not listened to it
yet, took a bookmark for later when I have time.
Walt
________________________________________________________________________
________________________________________________________________________
|