Well, that was far harder that it needed to be...
Anyway, for anyone who's interested, I've put two 1-minute clips
up on the Nature Recordists area on Soundcloud. I've taken them
more or less at random from the last 24-hour recording I did.
They are both straight out of the Songmeter, with no editing or
filtering whatsoever.
They were recorded at 44.1kHz 16-bit, as .wav files. No low cut
filter was activated on the recorder. The Songmeter was hung on a
tree in the middle of a flying-fox camp, with one mic attached
directly to the box, and one mic extended about 1m away.
The two files are at 0130, when it's very quiet in the camp - one
bat flies overhead and lands in a tree, followe dby a bit of an
argument.
The second file is just pre-dawn, when the whole camp has
returned from feeding and they are all talking together...
No pretence at artistic merit or anything similar, just to give
the people who asked some idea of the default sound out of the
box...
cheers...
TIm
--- In
<> wrote:
Later this evening, after I've had some dinner, I'll see if I can
post an example from the songmeter so you can see what they are
like. I'll extract a couple of minutes, totally unedited. As I
said previously, it's not the equivalent of a high end recorder &
mic combination, but it's not bad.
And they aren't silly money, when you consider that you are
getting a recorder, two (interchangeable, weatherproof) mics, a
weatherproof housing, and a computer to do scheduling...
Out here (Aus) they're used quite a bit for fauna surveys - both
microbat and bird. And my lab uses them extensively for our work
with vocalisations of Aus mammals...
Before we had SongMeters I mucked around trying to waterproof
stuff - with mixed results. These are much easier.
tim
--- In <> wrote:
> The specs say "-115dBV equivalent input noise", which I think
might put it in the "ok" bracket. I assume a lot of what you pay
for them goes towards robustness and programming flexibility.
Peter,
Apologies for technical mail.
"-115dBV equivalent input noise". Equivalent to what? You need
the "what"
defined. I don't take specs at face value, so I put my "toy"
Tascam DR-1
through its digital paces.
At 16 bits sampling, minus the parity bit and the biggest bit
being 50%, I
got dither noise on the smallest three bits. This is needed to
average out
the slight sampling noise if you record very low. That makes 11
bits clean
which is 66dBs. That is well swamped by the natural noise in my
recordings
even when recording low.
At 24 bits sampling I only found a 20dB benefit which was 4 1/2
bits more
than the 16 bit digitiser, not the 8 bits or 48dBs I expected.
I'll check
out another recorder. However, at the 15 1/2 equivalent (11 plus
4 1/2) bits
available, this toy machine is as good as my low noise MKH mics,
and SQN
mixer and the wind in the trees which the birds are singing from.
Summary. Don't believe all the sales hype. I worked out some
figures.
Inevitable thermal noise (from vibrating electrons) drowns out
the higher
(lower level) bits. A 24 bit - 22 active bits - digitiser goes
down to
-132dB which doesn't really exist, but you would feed that
expensive
digitiser from a pro mixer which would drown out the noise
anyway.
I've got a pro SQN mixer and I can't fault my DR-1 for noise, but
I prefer
mu DR-100, still in the "toy" category. It is easier to use,
which is more
important than theoretical specs.
David Brinicombe =
|