Interesting example of aliasing...

On Mon, 04 Jun 2007 09:54:38 +0800, Steve Underwood <steveu@dis.org>
wrote:

>Clay wrote:
>> On Jun 3, 8:35 pm, Steve Underwood <ste...@dis.org> wrote:
>>> Eric Jacobsen wrote:
>>>> Got this from a friend, who was mostly interested in it because it
>>>> looks cool, and we both play an on-line game that includes this type
>>>> of helicopter.
>>>> http://www.break.com/index/helicopter-blades-stop-moving.html
>>>> It's a pretty interesting example of visual aliasing.   I wonder what
>>>> the rotation would have looked like if one of the rotors had been
>>>> painted a different color.
>>> That's fun. I wonder if it was staged, by getting the rates tuned up for
>>> it, or just fortuitous? That tail rotor is pretty slow rotating too,
>>> which adds nicely to the effect. The way the rotor stays still for so
>>> long is a good illustration of how a helicopter really locks down the
>>> speed of the rotor.
>>>
>>> Steve
>> 
>> 
>> I think Randy is on to the right idea. I bet somebody used the audio
>> to create a shutter trigger for a ccd camera. I recall an audio
>> triggered stroblight that was used for diagnosing large mechanical
>> transmission systems. The device had a microphone and the audio was
>> FFTed and displayed on a screen. The user could move the cursor over
>> to the desired frequency of interest and this selected a frequency
>> that was used to trigger the strobe light. A phase adjustment was also
>> available. It was really cool! The transmission case was opened and
>> the gears were run. The strobe was synched to the sound of interest
>> and the phase was adjusted to effectively rotate the bad spot on a
>> single gear into sight. This made finding bad cags and bearings rather
>> simple in an operating system.
>
>Taking a second look, I guess its too steady to be anything but hard 
>locked, probably acoustically. Its probably someone trying to demo their 
>autosyncing camera. You might be surprised how little the rotational 
>speed of a helicopter blade varies over a range of conditions, though. 
>I've developed electronics for helicopters and had very steady sustained 
>patterns in signals which turned out to be effects from the main rotor.
>
>Steve

I think you're right.   Looking at it again there's essentially zero
drift between the blade phase angle and the view point relative to the
helicopter.   The only time the blade rotates is when the helicopter
turns.

Off the top of my head I think you'd expect a fair amount of drift
between the rotor and the frame rate even if both used high quality
oscillators as references.  I doubt there's a rubidium clock in the
helo and the camera, so something else must be happening.   Your
acoustic-lock theory sounds likely to me.

And it's a slick trick!

I'd not heard of using this technique like you described previously to
find the source of mechanical noises.   That's a neat application as
well...I can think of a lot of places where that would be handy.

Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.ericjacobsen.org