Parabolic microphones.

[Greg Miller]

Interesting hobby article, except for the misleading use of the word "Parabolic" in the title. "Curved" would have been more accurate. e.g. I doubt that the skillet or trash can lids were really parabolic.

Based on the photo of the trash can lid, which is relatively flat, I would have guessed that the actual focal point (if there is one at all) would be much farther out. But the photo could be misleading.

Actually I disagree somewhat with the comment that the collander would not be useful because of its punched holes. (But a strainer, made from screen, would be entirely different from a collander with some small punched holes.) As long as the hole dimension is a small fraction of the shortest wavelength desired, and the overall perf area is a small fraction of the overall area, perforations would have little appreciable effect on the reflection. That's why perforated satellite dishes, and WiFi dishes constructed of wire segments, still work effectively despite the apparent open spaces. Of course holes would allow a small amount of leakage -- mostly HF -- from behind the dish.

[Don Palomaki]

Originally Posted by Greg Miller

Don,
... I would love to see a 3-foot or 4-foot diameter dead cat! Sort of like a giant furry garter belt...

Maybe visit a fabric store and check out the fake fur. I've not tried it, but it might work. specially if used to line the edges

[Bob Hart]

QUOTE:


"I would love to see a 3-foot or 4-foot diameter dead cat! Sort of like a giant furry garter belt."

That can be arranged. I have a roll of deep fur-fabric I originally got hold of for custom deadcat muffs for the wider enclosures I made for the Sony mikes.

My inclination would be to deadcat the rim and the ECM-55B, its holder and maybe its support spar which is made of square furniture tube. To reduce the front-to-back mechanical acoustic couple through the dish itself, I might try some old carpet cut into pizza slice segments and glued onto the rear surface

[Bob Hart]

Well here's a quick unscientific test.The MixPre and Zoom H4n recorder were each gained up a fair way, about 60% of the gain cointrol on the MixPre and 52/100 on the Zoom. Normally the Zoom's gain would be kept down to about 0.5 - 1 when taking feed from the mixer.

The dish was aimed blind through the bushland to see what was there. It happened to point towards a main road junction about 3km distant. Rejection from off the axis was good. There was a gas cannon operating on an orchard about 2km distant to the right and it was not coming through in the audio.

EDIT. Youtube spat the dummy on the first upload which would not process. Excep for cuts and crossfades, no other process was done on the sound.

PARABOLIC TEST - EARLY MORNING BIRDS. - YouTube

[Greg Miller]

Hey, Bob, that sounds pretty good! Glad to see you're having success there.

Just for the sake of reference, what are the rough dimensions of the dish? (Forgive me if you mentioned this above, it's midnight and I don't have the fortitude to read through the entire thread again.)

[Bob Hart]

Greg.


The dish is ovoid in shape. The dimensions are in metric as I could not find my imperial tape measure. The dimensions are 660mm narrowest and 720mm widest of the active reflecting surface. The 720mm width is oriented in the upright direction. Add about another 32mm diameter for the rolled edge. Actual depth of dish as measured from a line running directly rim to rim is only about 65mm.

The profile is an angled flat faced crop of a larger parabola. Focal point to dish is about 490mm for best acoustic performance.

The satellite transducer was about 25mm closer but the active element is probably furthur inside the plastic shell, which would make it the same or near to.

The thickness of the material can only be guessed as it has a near double-rolled edge. It looks to be about 1.5mm pressed steel rough-finish powdercoated.

My guess is a proper full parabola dish would work better but this one is adequate as is.

I gave it a spray coat of gloss black to make it go faster, ( just kidding ), to smoothen the surface for better reflectance.

The stand is from a 2 x 500watt halogen worklight pair. ( Your Home Depot brand equivalent ).

The dish has been inverted. The nuts which lock off the vertical arc of adjustment have been replaced with wingnuts for faster adjustment.


The birds were not in sight. They were very distant to hear by ear, possibly in the ballpark of about 50 metres in the trees or garden on the opposite side of the road. The crow, I did not hear by ear at all. However this is not a fair judgement. Leaving the muff off one ear is probably too tricky for the brain to cope with. The really loud bird in the last segment had flown closer to about 20 metres for water I had sprayed up into my own trees to draw the birds in.

[Greg Miller]

Bob,

Thanks for the dimensions. Don't worry, I can convert to imperial, even though the US dropped the ball on official changing to metric back in the 1970s. I studied a fair amount of science and engineering, so "25.4" and "39.37" are firmly ingrained in my aging gray matter. (And somewhere in my shop there's a wooden folding rule that I picked up in Germany back in 1979.)

I guess any original "wrinkled" or "bubbled" finish might de-focus the dish very slightly, but probably not enough to worry about. Let's assume the original RF frequency was 10 GHz; that's a wavelength of 0.03 meters = 1.18 inches = 0.0983 ft. Assuming speed of sound to be 1100 ft./sec., that corresponds to one wavelength at ~~ 11,180 Hz. My point is that if the dish was flat enough for 10 GHz RF, it is flat enough for > 11 kHz audio.

Another way to look at it: a dimensional error of 0.04 wavelength will cause negligible gain decrease; of course a smaller error will cause even less decrease. So unless the original "roughness" exceeded about 0.047 inches (sorry, ~ 1.2 mm) it shouldn't cause much gain reduction until you get up to 11 kHz.

Anyway, kudos on your experiments. I will have to start looking for "road kill" dishes around here, and maybe try to replicate what you've found, once the weather warms up.

[Bob Hart]

Greg.


"Roadkill" That's a new one for me. I shall have to remember it.

What surprised me that the lower-frequency off-axis rejection was subjectively better than I expected. I distinctly by ear heard but not in the recording, the gas cannon going off ( tweetie blaster"). These are used as an alternative to lethal force against protected wildlife which will bankrupt stone-fruit growers otherwise.

Whether this rejection is attributable to directivity of the dish or less efficient capture of the lower frequencies or both aspects being interactive, I do not know.


FOOTNOTE:

In the early morning cool drift down the gully on a still day, the sweet scent of ripening peaches can be caught here. When the Fenthion ban is brought in, the peach trees will likely be pulled as the crop is the most sensitive to Meditteranean fruitfly.

[Greg Miller]

Bob,

I wonder whether it was low frequency off-axis rejection, or just low-frequency attenuation. The dish has a lot of gain at high frequencies, but is essentially non-functional at low frequencies. So if you use the high frequency output level as an audible reference, the low frequencies certainly will be attenuated.

In other words, at low frequencies (certainly below the frequency where the dish size < one wavelength) the dish gain is negligible, so you have an omni pickup pattern (just the mic element) with resulting unity gain of 0 dB. At higher frequencies you have a very directional pattern with a gain of some unknown amount ... probably +6 dB or perhaps a lot more than that. The gain tends to rise with rising frequency. So all the low frequencies, both off-axis and on-axis, are very much lower in level relative to the high frequencies.

That dish has a minor axis of 660 mm ~~ 2.16 feet. That's one wavelength at ~~ 500 Hz, so your system should be pretty much unity gain below that, and rising gain above that.

[Bob Hart]

ACOUSTIC FRESNEL REFLECTOR ??.


Those of us who are older may be familiar with the effect of a large vertical area of corrugated wall or fence upon a sharp explosive sound. There is an artifacted echo return which is time-stretched but also reproduced as a sharp high pitch. Corrugated iron is a sinewave shape.

My assumption is that the high pitch is created by the incremental arrival of each short intense echo in rapid succession over time from each of the angular corrugated surfaces which collectively direct their echo towards a common point. That point is not apparently critical along the wall's length so long as there remains sufficient wall area to provide the return.

More critical it seems is the distance from the wall of the listener in attenuating that return. Too close and each peak in the corrugation will mask the echo from its neighbour, thus eventually attenuating by reducing the effective reflective area to the listener.

Too far and the return is more akin to a normal echo, far less time-stretched as all reflecting elements of each corrugation are more equidistant from the listener. There seems to be a sweet spot distance from such walls.

So for the scientists among you, my question is this, could much lower audio frequencies be amplified to a listening point by superimposition of each individual increasingly delayed echo from the sine-shaped upright corrugations. I guess there would be all manner of frequency bands of amplifying and nulling.

If unwanted higher "artifact" peaks were graphed out of the audio recording, could an amplified lower frequency recording be constructed from that echo return additive to the direct sound from a distant object like an aircraft?

I definitely heard an enhanced return from a corrugated airport hangar wall, of sound of turbulent airflow over an aircraft's stalled wing. My subjective estimate is that there might have been as much as 12db worth.

The aircraft was in a favorable position for sound to be reflected off a gently sloping but almost planar paved apron angled towards the aircraft, reflected up at a shallow angle to the corrugated iron wall then concentrated back to my head-height where I was camera-operating.

I have been possibly chasing my tail playing with a parabolic reflector mike which amplifies higher audio frequencies.

A mike, or even a row of them along a "sweet spot" line parallel to the corrugated hangar wall and flight of the performing aircraft in a "sweet zone" might do the trick with the lower frequencies of air-buffeting and airflow detachment from a wing of an aircraft flying aerobatic figures.

Any advice will be appreciated.

[Bob Hart]

An interesting notion which came up was that of the WW2 and prior long-range acoustic detection devices which preceded early VHF ground-to-air radars.

These acoustic systems consisted either of huge squarely constructed conical feedhorns or large shallow parabolic reflectors. Neither of these structures were orientable or portable.

By the nature of its ability to focus off-axis sources to focal points in a limited area around the natural point of focus, the parabolic listening station was able to plot to a limited ability, off-axis direction and altitude of sources. The parabolic listening station reflector was uncannily similar to the cropped parabola of the modern Foxtel style dish, except on a much larger scale.

It would be technically feasable but highly impractical to construct a parabolic dish of concrete in a dug pit under the aerobatic display box at the airfield. Likewise, an aerobatic display might be scheduled to occur within the focal zone of a radio-telescope dish or a decommissioned older generation large communications dish in order to make a one-off specimen of a unique sound. Most of those dishes are now long gone from here.

Using a communications dish which is still in service would be hazardous to human health and recording equipment due to high-energy RF at the focal point where the microphone would be placed.

This is all of course swerving well off the course of practical application into speculative theory but an interesting excursion nevertheless.

[Greg Miller]

Bob,

I have read about, and have seen photos of, some of the dual-feedhorn acoustic locators. The ones I remember were all steerable, at least in the horizontal plane, and probably vertically as well. Big massive contraptions with ear tubes leading down to a centrally located listener.

I have also seen photos of some cast concrete dish reflectors (maybe parabolic, but maybe spherical sections) that obviously were not moveable.

One problem with a giant parabola is that the beam width is a function of the ratio of (wavelength / diameter). In other words, for a given frequency, as the dish gets bigger, the beam width becomes narrower, and aim becomes more critical. (Remember, gain rises at ~ 6dB/octave. And you get more gain by narrowing the beam width.)

OTOH as the diameter becomes smaller (allowing for easier aim), the LF cutoff frequency goes higher and higher.

I suspect the combination of these two factors is one reason that most reflectors for microphone use tend to max out in the range of 24" to 32". This seems to be a good tradeoff. If you make them bigger for better LF performance, they become difficult to focus. If you make them smaller for wider beam and easier focus, the LF cutoff rises up into the voice range (or even higher).

A parabolic mic seems to be good for specific uses (where you can adjust dimensions to suit your needs), but is not a good general purpose long-range "miracle mic."

[John Willett]

Originally Posted by Bob Hart

An interesting notion which came up was that of the WW2 and prior long-range acoustic detection devices which preceded early VHF ground-to-air radars.

These acoustic systems consisted either of huge squarely constructed conical feedhorns or large shallow parabolic reflectors. Neither of these structures were orientable or portable.

By the nature of its ability to focus off-axis sources to focal points in a limited area around the natural point of focus, the parabolic listening station was able to plot to a limited ability, off-axis direction and altitude of sources. The parabolic listening station reflector was uncannily similar to the cropped parabola of the modern Foxtel style dish, except on a much larger scale.

It would be technically feasable but highly impractical to construct a parabolic dish of concrete in a dug pit under the aerobatic display box at the airfield. Likewise, an aerobatic display might be scheduled to occur within the focal zone of a radio-telescope dish or a decommissioned older generation large communications dish in order to make a one-off specimen of a unique sound. Most of those dishes are now long gone from here.

Using a communications dish which is still in service would be hazardous to human health and recording equipment due to high-energy RF at the focal point where the microphone would be placed.

This is all of course swerving well off the course of practical application into speculative theory but an interesting excursion nevertheless.

Here are some pictures I took of the "sound mirrors" for this purpose in Kent (UK) - just to show you how large they are.

They are open to the public one Sunday a year in July.

.

[Bob Hart]

John.


Thanks for that. If must have been a fair old nightmare, drawing, making and setting up the formwork for the concrete dishes.

It seems a pity there is no apparent conservation work being done to the deteriorating concrete.

[Bob Hart]

A footnote to my endeavours above, as luck would have it, on the roadside today I found an old 48" or 4ft diameter aluminium satellite dish. There are no brand names on it, only a stylised blue lightning logo.

It was around two inches too broad to go in the back of the small hatchback so it had to ride upon cushions on the roof, tied down with strips of nylon spiral wire wrap. It was a fraught but successful journey.

This is a centre-fed dish. Initial tests with the Sony ECM-55B mike hand-held in front of the temporarily aimed dish, suggest it is far more critical in its acoustic "focus" than the smaller steel dish which is fed off-axis.

With the audio gains turned up, outdoors, it will pick up the ticking clock at 18 metres. The clock comes in faintly but distinctly over the environmental noise floor.

There is less forgiveness for off-axis aim. The distance from the dish centre towards the feed horn, seems to be about twice as critical as with the offset dish. The "best" zone is about a 25mm area along that centre radius line.

The audio seems more mellow than with the smaller dish. There does not seem to be any more actual audio gain as such but discrimination between on-axis and off-axis sound seems improved.

The aluminium dish is quite light but much more awkward for agile use on a tripod head. It does not have adjustments built in for aiming so I will have to invent these unless I use a heavy camera tripod.