Jump to content
UltravioletPhotography

Pre-Flash Discussion


JCDowdy

Recommended Posts

EDITOR'S NOTE: There was an interesting discussion about Pre-Flash in a formal post. So I have split it off from the original thread it was posted in and put it here. (Original Thread)

**********************************

 

 

 

I have used an on-board flash when shooting a Long-legged Fly. When the flash went off, the fly jumped before it could be recorded. I tried again later - same thing. That is one very quickly reacting little fly!

 

Given that the fly is not faster than light, what was it reacting to?

Link to comment

Thanks Andrea.

If you consider the construction of their eyes, there are always some looking directly into the sun (if the sun is shining).

If we look into the sun, we directly know it is not safe and our eyelids close, we look away. These little creatures have no chance to do this. So I assume, their single eyes can withstand more light, perhaps even a flash is no problem.

Link to comment

Given that the fly is not faster than light, what was it reacting to?

 

:lol: :lol: :lol:

Well, if that fly was faster than the light, then we should notify Einstein. He would certainly want to know!

 

The fly was faster than the shutter I suppose. I tried and tried to get a flash foto of it. No dice. I watched it jump. It was all very cute. I've never had any other insect jump when a flash went off.

Link to comment

The Sourakov paper only states "...photographed with a Canon EOS camera the built-in flash at a shutter speed of 1/200 s = 5 ms" which seems to be the basis for the reflex time. I doubt this is correct, obviously the insect must be reacting to something it sensed prior to the flash.

 

My browsing found one insect photographer advising to use the flash in manual mode which suggests the pre-flash question is pertinent. Another source mentioned using silent shutter and diffuse flash to avoid the startle reflex.

 

These Long-Legged flies are unquestionably highly visually adapted, relying on courtship display and exaggerated morphological features more than chemical pheromone based signals for reproduction. Rapid movement is the stimulus for the startle reflex, if the timing of the preflash does not compute then perhaps are they seeing into the lens and reacting to fast aperture or shutter movements?

Link to comment

JD, I can't believe you actually found a paper about this exact thing !!!

It is always refreshing - and reassuring - to know that someone else has observed the same unusual thing one has noticed.

 

There was no pre-flash.

ADDED 14 Oct 2015: I think I am totally wrong!

 

I have many fotos of the Long-legged Fly made without flash. So I don't think they are reacting to shutter movements or fast apertures.

Link to comment

This comment is made in my usual total ignorance but perhaps the fly sees the leading edge of the flash but the camera needs to collect all the reflected light from the total flash to record an image. I imagine that the rise time for the flash must be pretty small but the total flash duration might be around 1/600 sec based on the reported one for a Nissin Di866.

Dave

Link to comment
A Stranger In The Wind
Very interesting Andrea. Often wondered what our winged friends look like in the UV spectrum. Excellent shots.
Link to comment

Excellent indeed and spawned an intriguing discussion, thank you!

 

Dave, absent pre-flash stimuli it seems we are down to flash duration. The paper reports a latency period for the startle reflex of ~2-5 ms or 1/500 to 1/200 sec. I don't know what is typical so I looked around and found interesting this posting: Actual Measured Flash Durations of Small Speedlight Strobes which shows the slowest flash durations are on the order of ~4 ms for the units tested with most of the emission in the front of the pulse. On lower power settings most of the speedlights tested were more than an order of magnitude faster.

 

So, I am trying to wrap my brain around how the fly gets away. Even at the longest durations most of the strobe should have reflected off the fly during the 2-5 ms latency period before it moves. If the fly is reacting to the light, it seems in the worst case one would have an image of a still fly with some blur as it reacts. So I am still stuck on the fly is reacting to something preceding it's visual perception of, and reaction to, the light from the Xe tube discharge.

Link to comment

So, I am trying to wrap my brain around how the fly gets away.

 

I do not know what happens when Andrea uses her camera, but I would first check if the pre-flash can even be turned off for the built-in flash for that unidentified Canon EOS. I know in some models of DSLR it is impossible to do.

Link to comment

I am back from China, then Canada. Did more vegetables there.

 

Strangely, I could not visit this site in China. I got an error. I did a screen catch on my laptop.

 

I have shot about 20 frames of this fly when I was in the Phillippines, sitting on a van and a fly landed on my leg. I kept taking photos of him/her... that was using a Coolpix 990 (3.2 megpix).

 

I contacted the author of the fly-flash paper today and he said he was not aware of the pre-flash and told me the model was EOS T1i. Anyone has data on how much time ahead is the pre-flash? I could not find anything in the manual.

 

Zach

Link to comment
  • 1 month later...

I made a note up there in Post #7 and said that I think I was completely wrong. It was me who was not aware of the Pre-Flash. That Long-Legged Fly, while very very fast, had to be reacting to a Pre-Flash. Nothing else makes sense !!!

 

I stand happily corrected. :lol: :D :D

 

Note that if you set TTL flash to a rear curtain setting, then pre-flash is more obvious because there is a larger gap between pre-flash and flash. That's how I realized there really was a pre-flash.

 

Duh !!!!

Oh well. If I already knew everything then it would be time for me to go. And I'm not ready to do that yet.

 

Somewhere there is probably some technical data about pre-flash duration and how long the gap is between pre-flash and flash.

Link to comment

Here is the technical data. I've been at my workshop.

 

First a Canon EOS1100D (outside Europe rebel?) built in flash, kit lens. Delay between pre-flash and main-flash is 77ms (1/12s):

 

http://up.picr.de/23451118ql.jpg

 

Then Pentax K-x, built in flash, kit lens. Delay between pre-flash and main-flash is 100ms (1/10s)::

 

http://up.picr.de/23451119rn.jpg

 

So ignition on leading edge or trailing edge of shutter doesn't change very much on 1/100s exposure time and fastest sync speed of cam around that. I can remember of a EOS1D Mk II that has about 64ms Delay (only first flash!) with external flash.

 

The homemade sensor, 4 photodiodes from a Thoma TM3 head, 50 Ohms shunt, obviously much faster than the flashes.

 

http://up.picr.de/23451120ec.jpg

 

Here type and settings of the gould oszi:

http://up.picr.de/23451121wd.jpg

Link to comment

Very nice toys Baffle!

 

Seems like the fly has between ~65-75 ms to get out of the frame.

 

That still seems pretty quick but much more believable.

Link to comment

If the conditions are not clear (distance unkown, aperture closed, lens slow) then a dslr from Canon or Pentax fires (in TTL mode) a burst of flashes. Possibly what Andrea saw..

 

Up to 100ms long. The time between last pre-flash and main flash remains the same.

Link to comment

If the frame is 0,1m and the time is 60ms then average speed is 0,1m/0,06s 1.6 m/s about 6km/h

 

That is not sooo fast, it is like a walking person.

Link to comment

Baffe, thank you for this measurement experiment. Great work!

 

And so -- when we know something is "there" and we look for it, then it becomes easy to detect -- as pre-flash is detectable for me now. I think in the past I was simply not thinking about it even though I knew very well that TTL flash has to perform a "measurement" prior to shutter-up. I would have sworn to you that there was no pre-flash. Well, such dimwitted thinking happens from time-to-time even when we know the facts. :lol: :D ;) B)

 

I'd say the Long-legged Fly is still pretty quick on its long-legs. :D

Link to comment

It is still quick if you consider the scale factor. I do not know how big the Long-legged fly is but it looks to be on the order of 5-10mm long nose to tail.

That 0.1m distance to escape the frame is on the order of 10-20 times it's body length in ~60ms, including whatever time the reaction delay actually takes.

I'll wager they are hard to swat!

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...