Strobe or flash for insect flight

[Graham Stabler]

I need to find a strobe of flash suitable for some stroboscopic imagery of an insect (bee of fly) flapping it's wings.

The plan is to attach the insect to a record player stylus and use the outputted sinusoid (I tried this and it works amazingly) to trigger the flash. By adjusting the phase shift I should be able to get a set of images that cover a complete flap cycle.

But I know to get good motion stopping I will need a lot of light and what is more I would like to get probably 2 flashes per second for 20 flashes or so.

Any ideas appreciated.

Cheers,

Graham

[Charles Krebs]

http://www.cookecorp.com/cooke/php/prod ... 31302.html

[Joseph S. Wisniewski]

Charles, you're a madman.

But you're my kind of madman. Do you own one of those, or have access?

Look at the specs on that thing, 24 J in a 750nS pulse. That's 32 megawatts.

This might be "overkill" for Graham's application. The almighty Google says a bee's wings beat 210-290 beats/sec. That's 3450 usec for a full cycle.

My normal high speed lighting setup consists of four recent Nikon flashes (SB-80 or SB-800) connected via PC cords or Nikon SC-18 cords. The flashes are manually set to very low power. The direct connections are necessary, the Nikon built in optical triggers are dog slow.

1/17,800 sec (56 usec) at 1/32 power
1/32,300 sec (31 usec) at 1/64 power
1/41,600 sec (24 usec) at 1/128 power

Considering that once the the insect is secured to the rapidograph stylus (I think that's what Graham meant) so the 24 usec is about 0.7% of a cycle. Some blur, but it shouldn't be too bad.

1/128 power is plenty, if the flashes can be positioned within a few inches of the bee. I routinely use 1/32 power with flashes a full foot away from exploding glass, with pretty good results.

[dmillard]

Graham -

Have you seen this site?

http://www.pbase.com/fotoopa

David

[g4lab]

Graham
The notion of using a phono stylus as a vibration sensor for a bee is positively brilliant.

[Graham Stabler]

Charles, looks awesome, will find out the price.

Joseph, a rapidograph is a pen but I think we are on the same wavelength. 200hz is quite typical and 20us is a perfectly good exposure time from my experience with high speed, the problem is that I need several flashes in relatively quick succession.

David, yes I have but he takes only one image.

Cheers,

Graham

[Graham Stabler]

g4lab, can't take credit for that idea, it's in some pretty old literature.

Graham

[Joseph S. Wisniewski]

Charles, looks awesome, will find out the price.

Be sitting down when you find the price...

Joseph, a rapidograph is a pen

So, what's this "rapido player" you mentioned earlier in the thread?

but I think we are on the same wavelength. 200hz is quite typical and 20us is a perfectly good exposure time from my experience with high speed, the problem is that I need several flashes in relatively quick succession.

At 1/128 power, the Nikon SB-80 or SB-800 (SB-80 are pretty cheap on the bay) can deliver 128 flashes in pretty much as "rapid succession" as you'd like. They're quite efficient devices, gating the flash tube on and off at high speed to get short durations, instead of shorting out the main flash capacitor and "dumping" power.

This is 6 shots at 2mS intervals (fired by a Mumford "time machine" controller). 1/32 power, four flashes connected together (2 on the left, one on the right, one behind and up high).

[Graham Stabler]

So, what's this "rapido player" you mentioned earlier in the thread?

I said record player. It plays records, 48's, 78's whatever. A a gramophone, a deck.

I suppose the key is the fact that I can get the flash really close, however I will be looking at dark objects at at least 1:1 so will probably need more light than your glass images. I'll try out the Nikon flashes, well worth a punt.

Graham

[DaveW]

There are some DIY flash triggers here:-

http://courses.ncssm.edu/hsi/pacsci/fpaper.html

I had read before, and the link below confirms, that camera flashguns are better for high speed work than studio flash. The studio flash on lower power settings simply reduces the power, but not duration of the flash. Whereas the camera type guns misnamed "power" settings do not reduce the power of the flash at all, but it's duration as you place them nearer the subject:-

http://www.scantips.com/speed.html

Some others on high speed flash that people may find interesting:-

http://chdk.wikia.com/wiki/Samples:_Hig ... Flash-Sync

http://www.highspeedflash.com/

http://www.naturescapes.net/022004/jb0204.htm

http://www.woodselec.com/HowTo.htm

DaveW

[Graham Stabler]

Thanks Dave, some excellent links as usual.

My trigger will have to be a custom job, it will take the sine wave from the stylus, do zero crossing detection and then produce triggers at various delays from the start of a cycle. Not totally trivial but my experiment controller is based on a micro controller that makes this stuff pretty easy. I have done crossed beam triggers for high speed video but that site gives me even more ideas like landing detection with a piezoelectric.

Things are looking good and if one flash won't do it, I suspect two will.

Graham

p.s. I moved from Nottingham last year, I did my degree, PhD and a couple of years post doc there.

[DaveW]

Nottingham's a good place to be moving from Graham! I am on the high ground up Mapperley.

DaveW

[Joseph S. Wisniewski]

My trigger will have to be a custom job, it will take the sine wave from the stylus,

Stylus? You just got through telling me it wasn't a pen.

do zero crossing detection and then produce triggers at various delays from the start of a cycle.

You may find you have better results if you take the zero crossings of the first differential. The beat is highly asymmetric, and shouldn't look much like a sine wave. The middle of the stroke, near the zero crossing of the AC coupled waveform, is chaotic and will jitter all over the place. Top of stroke it much more reliable.

Not totally trivial but my experiment controller is based on a micro controller that makes this stuff pretty easy. I have done crossed beam triggers for high speed video but that site gives me even more ideas like landing detection with a piezoelectric.

p.s. I moved from Nottingham last year, I did my degree, PhD and a couple of years post doc there.

But did you meet Robin Hood?

[Graham Stabler]

You may find you have better results if you take the zero crossings of the first differential. The beat is highly asymmetric, and shouldn't look much like a sine wave. The middle of the stroke, near the zero crossing of the AC coupled waveform, is chaotic and will jitter all over the place. Top of stroke it much more reliable.

The motion of the thorax is not especially asymmetrical (despite what the wings may do) and does give a sine wave, I have actually tried this. Even if the two half cycles are slightly different I can take care of this by looking at every other zero crossing.

Graham

[Graham Stabler]

I received a NikonSB-800 speed light today, unfortunately no cable came with the order as I had requested but it is a start.

I forgot all about this but I took a picture of my scope when I had the bee attached to the stylus.




Could do with a few less cycles to see the shape but gives some idea.

Graham