From: "Secomb's"
> Hi all
> I can see the trick of any good reflector is good shape to it which is
> going to be challenging. Apart from the site that Tom used
> http://graffiti.virgin.net/ljmayes.mal/var/parabola.htm
> is there any were else to source a simple way of getting parabola shape.
> An aluminum dish would be good but I seem to remember a while back it
> was mentioned that metal reflects give a metal tone to the recordings.
Metal dishes tend to be like metal bells, they ring. Plastic rings too,
but damps out pretty quickly.
By the time you work around and make a mold, then make the dish, you may
find it cheaper to just get a Telinga spare dish, a proven design. (or
Edmond has aluminum solar cooker dishes) Unless, of course, you are just
doing it for the challenge. Tom's site has probably the easiest way as
you don't make a mold.
> Walter, I have read Telinga sight several times but as yet I have not
> read the reading you recommended
Ok, as the Naturerecordists file section seems stuffed with soundfiles,
I put a pdf of it up temporary on my website (WARNING 1.4 meg):
http://naturerecordist.home.mindspring.com/Parabolic.pdf
It will not be able to stay up there too long...
> http://home.att.net/~rsl/APPR1.htm by Randolph Scott Little,
Note that this older article may make the error of thinking of sound too
much as a transverse wave or at least mixing theory from the two types.
Transverse waves do describe light or radio waves and there is a large
body of formulary and theory for that. Sound is not a transverse wave,
so such theory and formulas and their conclusions are inappropriate for
using a parabola for sound, though they work well in designing radio
antennas. Mixing the two types in our theories is the primary source of
confusion about parabolics. Here is a webpage showing the difference
between transverse waves and longitudinal waves like sound:
http://www.gmi.edu/~drussell/Demos/waves/wavemotion.html
In light of that how much of that article is correct would have to be
evaluated carefully. In any case, the paper I reference is a later
paper, hopefully a little more correct. Even it probably needs to be
reevaluated though it's general ideas seem to work. A parabolic like the
telinga was not around when it was written, nor were our modern mics.
> From what I can gather
> 1;- a reflector becomes more directional, when increasing the dish size
> and have the mic. more into the dish. Would this also increase the
> distortion created by the edging effect when a sound is just off the
> focal area of the dish especially having the mic more into the dish?
>
> 2;- Most gain is in the higher frequency. I was surprised that the
> response curve was not smooth for the mid frequencies. This is nice to
> know but the response curve to a lot of mics isn't flat either. The
> combination of the 2 factors is to much of an unknown (for me any way)
> to theories what may happen. Could be a headache for one doing specific
> project concentrating on frequency variations within species or
> subspecies for instance.
When you consider how the sound gets to the parabolic, the effects of a
parabolic, if anything counteract some of what's happened to it since it
left the source. Sound attenuates unevenly in the frequency scale with
distance in approximately the reverse of the increasing gain of the
parabola. One of the telinga papers points this out.
In addition in any real life site there is a jumble of reflecting and
absorbing surfaces between you and the caller. Without accounting for
all of that you are fooling yourself if you think the mic is giving you
a accurate measure of the sound intensity of the caller at any frequency.
In general each species calls in a pretty limited frequency band. This
greatly simplifies the problem as we don't have to account for
variations in the mic across the entire frequency spectrum, but just in
that narrow band. Where the mic variations will be lost in the
variations imposed by the site. There are ways to analyze the sound to
extract at least the theoretical gain with frequency back out of the
recording.
Anyway, sonograms of calls recorded by parabolic come out as well as
those from any other mic. What your really look at is the relative
intensity of parts of the calls, not absolutes.
> 3;- imperfections in a DIY reflector may soften the edging effect plus
> giving lesser gain in the higher frequency, giving a slightly flatter
> response curve (instead of being skewed to the high end of frequencies).
Think about it carefully, how much of the above three are ideas due to
thinking of sound as a transverse wave?
I believe by edge effect you are thinking of a transverse wave effect?
Not really sure on this.
Nothing will flatten the response curve of a parabolic. At least not in
any predictable way. You can work with it. You can attempt to calculate
it back out in the computer.
> 4;- a reflector act as a barrier for sound coming from behind. This very
> dependent on the material used. The plastic umbrella I was using gave
> very little protection from background sound. Fiberglass gives some.
> Interesting to know how much protection there is from polycarbonate
> plastic and metal. Thick cardboard is good but weight is a big issue there.
True, the material of the dish does make a difference in this. I expect
even your umbrella does absorb sound a little. Both the polycarbonate
and the metal seem to be good barriers. If you wanted better the back of
the dish could be coated with acoustic foam.
> By the way when mention this
> "You have some low frequency falloff in a *notch* due to interference if
> focus is at rim depth, and even less if you stay inside"
> What did you mean by notch?
See the paper I put up. Figure 5 shows it most clearly. The explanation
makes sense to me. That, and some have reported this notch in their own
parabolas.
I really hate to get into all of this, religious wars are more peaceful
than parabolic theory discussion. The best way to know what a individual
parabolic does is to take it out there in nature and record your
subjects of interest. Whatever theory is, what you record is reality.
And that's how I judge parabolics. I do know that using the Telinga I
can record cleanly with gain even things like bullfrogs, who's calls go
down to 50hz or so.
I more consider that theory may give some hints of things we might need
to avoid in our design. Emphasis on the "may". The derived formulas in
the paper I put up may even need revising. What is known about the
effects of sound on mic diaphragms (and our own eardrums) needs to be
compared to how parabolics are viewed on the theory level, for instance.
The truest part of the paper you found is the statement that we know
parabolics seem to work. Without them our ability to record distant
subjects would be greatly reduced.
Walt
|