kaernbach wrote:
> Dear list,
>
> I am new to this list, and probably won't be long on this list, once
> I have bought my equipment. I hope you won't mind my selfish
> information sucking attitude.
>
> I am not a "naturerecordist". I am an experimental psychologist,
> focussing on hearing. I do most of my research with lab sounds
> (sinusoids, white noise, and the like... see http://periodic-noise.de
> for a typical project of mine). However, sometimes I need ecological
> sounds.
>
> My present project is on ugly sounds. Sounds that make you goose
> pimples. Finger nails on the blackboard, cutting styrofoam, etc. We
> want to have good samples of such sounds before we start looking into
> the stimulus properties that make them this ugly.
>
> My present question is on a cheap solution to record uncompressed 24
> bit 96 kHz sounds. Why these specifications?
> * 96 kHz because it is supposed that high frequencies play an
> important part in generating these goose pimples. High levels of near
> ultrasound might contribute to the effect.
> * 24 bits because that would make things easier. One would not
> need to adjust the gain control very varefully in order to have a
> 16-bit version that makes use of the full dynamic range. Also, with
> equal SPL high frequencies have much less amplitude. In order to
> capture them with some precision, 24 bits would be good.
I doubt that the sounds you want have all that great a dynamic range.
You might want to do some measurement to find out. You have 90+ dB to
play with in 16bit, a little more with the newer ATRAC which uses bit
shifting.
To accurately record the sounds well, you will have to adjust the gain
control well. It's more than just dynamic range involved. Mic pre's have
characteristics that change with the gain setting too, for instance.
> * uncompressed because we just don't know whether a compression
> algorithm (e.g. ATRAC) would change anything of importance to those
> sounds. While one might not be able to tell the difference between
> compressed and uncompressed bird songs, music, speech, ... it might
> just make the difference between a GPS (goose pimple sound) and other
> sounds.
Only recording the actual sounds and trying them will tell this. I'm
guessing that ATRAC can get this effect, but would have to record to
find out for sure. ATRAC does not change as much as people think. It
should, in fact, be thought more of as a sound synthesis system. None of
the original sound is recorded, only what amounts to instructions to the
decoder, which synthesizes a replica of the original. A extremely good
replica.
> I have studied the offer of Raimund Specht,
> http://www.ultrasoundgate.com/usg124.htm
> which comes to 2552 EURO (#31242 + #10212, incl. 16% VAT). O.k., it
> would be nice to have a hand-held device like that (and even nicer if
> it had some own recording technology, say solid state). But let's
> suppose I had much less funding available, say 500 EUR or so. Is
> there any solution that would allow to do uncompressed 24/96
> recordings at this price? It needs not to be a handheld solution,
> though laptop-based would be nice.
There are some other USB inputs that do the higher sampling rates. I've
not researched them.
> [Well, I am aware that the next thing to consider would be the
> microphone. I have a nice one here, I'll have to look up its
> specifications, but I suppose that an appropriate microphone would
> not cost thousands of USD/EUR.]
If you are of the belief that ATRAC will not get what you want, then you
are into a high quality mic. ATRAC gets everything cheap mics produce.
You may also have to be very careful about which mic you use as each mic
has a different sound characteristic it imparts to what it records.
To be sure of getting the high frequencies running into ultrasound, if
they are really necessary, I'd recommend the Sennheiser MKH-800. It has
specs designed to match what 24bit 96k can do, one of very few that do.
Most mics top out at or below 20k. You are going 96k to get the higher
sounds, make sure you have a mic to match. And, if the sound components
of importance are of a very low level, the MKH-800 is a very low noise
mic, so your low level sounds are less likely to be masked by the mic
self noise.
If you do find that these sounds are ugly because of ultrasound, you
will need to have something that can play them back accurately to use on
test subjects. Ordinary headphones may not do it.
Another thought is that it could be, in some cases, that a full
soundfield will be necessary, not just mono. In thinking about it most
of these sounds give me the perception of movement. Makes me wonder if
it's tied into our fight or flight reactions.
What you probably should do is record with what you have, and see if the
ugly effects record. If not, then go for something new. I'm not at all
convinced you will need all that exotic equipment to record these sounds
well enough. I do think it's probably going to come down to the mic more
than the recorder.
Walt
________________________________________________________________________
________________________________________________________________________
|