Weird Stobe Effect, 5D MII

Why does some of the water in the photo look like it was taken with a strobe light? It was taken in available light at the settings you see with a Canon 5D Mark II. 100% Crop. Any ideas?

I see nothing remarkable or unusual. The appearance of some drops being "frozen" and shouldn't be with .6 sec, could very well be that, similar to a flash effect, the drops were only illuminated by the natural light at one brief position or a series of positions. Without being there to see lighting conditions,it's impossible to explain fully. Probably not much help but that's my take.

Certain water droplets are illuminated many times at regular intervals along parabolic trajectories just as if a strobe light was used. I don't see how that can happen.

This doesn't look right to me. I do quite a bit of stop action photography and I've never seen anything like this. ... And what's with the red droplets? Interesting.

I've not seen this either and couldn't make a guess as to what's causing the effect.

The different colored droplets a most likely due diffraction of the sunlight hitting the droplets at different angles, so that we're only seeing the most strongly diffracted color.

I am interested to know how the discontinuous arcs occurred.

--P

Right! The arcs should be continuous! I wonder if there is something wrong with my camera or if there was some bizarre light conditions I was unaware of!

Maybe I was moving through a time-warp??

Check your Flux Capacitor. I think it's broken.

Ha!

Maybe the droplets hit 88mph...

Man, that is interesting. I was thinking maybe a malfunction on the system board that made the sensor go into a sort a video mode where it turns on and off 24 or 30 frames a second, but there are too many droplets. In 0.6 seconds you'd have either 14 or 18 in a row and the rest of the image would look like it does (of course) because the ground isn;t moving laterally. But then I remembered that the 5D3 does 60i FPS, so in .6 seconds you'd have 36 drops flying through the arc. Some of the arcs have about that number. I know this is a crazy idea, but if the system board did glitch a bit and the logic got messed up and combined the 5D3's multi-exposure capability with the 60 FPS capability and thus fired off a 36-image multiexposure (.6 sec x 60 FPS), you'd still have the right exposure adding the fast frames up, plus your terra firma would appear as one, and only laterally-moving objects would get caught in the 36 exposures (or less if the water arc started partway into the exposure).

I know, far fetched, but if that glitch did happen it would at least explain how you got the shot. Maybe check settings and ensure multi exposure is turned off? Maybe send that thing is and have it checked.

By any chance were you shooting with Live View which is the gateway to the video feature?

Ack. I just looked again and saw it's a 5D2 which I don't think does multi-exposure. Well, there goes my theory!

I say you just run 1.21 gigawatts through it and see if it clears the glitch.

What the hell is a gigawatt???

No live view. I almost never use it. Too used to looking through the viewfinder! It didn't do this on all of my shots but I have a few like this. I have no explanation.

Did you have an LED flashlight on? Many LED flashlights pulse the light. For available light, would expect soft diffuse shadows, but I see strong hard shadows on the leaves, thus I suspect a light of some sort.

Roger

No artificial lights. Only natural light.

What light is causing the sharp shadows on the leaves?

The only source of light is sunlight.

Here is the full image for reference. Area shown above is left-center.

This was taken in a spruce-fir-aspen forest with filtered sunlight. I wonder if movement of tree branches or leaves in the wind could have created some sort of strobe effect in certain areas of the image? When I examine the photo (and others like it) at 100% there are areas with the strobe effect and other areas where water droplet movement and/or bright reflection trails are continuous. In the same image! So it seems that only certain light rays have the strobe effect and not the entire image. Imagine an aspen leaf fluttering in the wind rapidly blocking and unblocking the light. Whaddaya think?

I wonder if it could be related to this:http://petapixel.com/2011/02/11/strange-effect-when-shooting-bass-player-with-a-canon-5d-mark-ii/

On one water drop trajectory, I counted about 30 drops, so in a 0.6 second exposure, that implies about a 1/50 second period, and the rate looks quite regular after accounting for the parabolic trajectory. Seems kind of difficult to get that high of a rate with fluttering leaves.

But if there were a reflection (e.g. off of leaves) and a resonance were set up, then the angles would not have to change much for the intensity to change.
It could also be a high level atmospheric turbulence effect but if some parts of the image show continuous streaks, then it is a more local phenomenon.
Can you reproduce it? Was that the only image of the scene you did?

Roger

That was pretty freaky Charlie!! I'm thinking there is some kind of sampling problem with some 5DM2 cameras. The shutter may have been open for 0.6 sec on my photo but how often is the information the sensor is receiving sent to the imaging chip? Would a sampling problem also cause the bizarre string movement in the videos Charlie posted?

I do have a couple other photos that show the strobe effect. Haven't had a chance to try to reproduce it. I'll post here if I do.

"how often is the information the sensor is receiving sent to the imaging chip?"

Light is collected in each pixel and only once, at the end of the exposure, are the data read off the sensor and sent to downstream electronics.
I have never seen an effect like this on other fast moving subjects, like a meteor in a night photo.

The bass player video is unrelated; that is due to rolling shutter in video mode at near frequencies of the vibrations of the string.

Roger

Phillip, when water drops exceed a few hundred microns in diameter they are no longer spheres, but spheroids and begin to undergo shape oscillations, changing from prolate spheroids to oblate spheroids and back as they fall. (This was first measured by atmospheric scientists in the 60s and the shape oscillation rate depends on the drop size.) How much light they scatter towards the camera depends on that shape. Thus, the series of bright spots is telling you that you've got large drops flying through the air and oscillating in shape. The effect gets stronger as the drops get larger up to about 10mm in diameter (at that size, they break apart as the fly). It's been long enough that I don't remember if it's the oblate or prolate shape that scatters light the strongest.

Mark, I think you're on to something. It would seem that if light was less strongly refracted from the droplet in the invisible part of the path, I wonder if at very high magnification of Phill's image, the droplet would be visible. If there is no light at all, I wonder if this is some cross-polarization effect. If Phillip was using a polarizer, and the droplet rotated to an angle that polarized the light 90 degrees to the polarizer on the lens, then no light would be detected be detected by the sensor.

I find this to be very interesting, and would love to see the reason for it.

--P

Preston, after further consideration, there's another complication, but let me try to explain with getting too long winded. (BTW, the first half of my PhD thesis was light scattering from single spheres so you're sending me back 40 years ) There may well be two effects:

1) the shape oscillations mentioned in my previous post. The amount of light scattered from a drop scales with the diameter at its equator. A prolate sphere is shaped like a frisbee and an oblate sphere is shaped like a football (falling point down), so for the same drop volume the diameter of the oblate shape is much smaller and you get much less light scattered. (As a rough guess, the difference could easily be a factor of 4 or more).

2) The brightness of the scattered light varies strongly with the angle between the drop and the detector (camera) where the line from the light source (sun in this case) determines the angle (0 would be direct back lighting). For very small collection angles (think lens aperture) this can easily change by 1,000x - 10,000x. That change gets smaller as the collection angle gets larger and I don't know what the collection angle for a single pixel, imaged through Phillip's lens, at his standoff distance is likely to be. However, I wouldn't be surprised to find changes in brightness of 10x to 100x.

Based on this further examination, it seems that the changing angle is more likely to provide the results that Phillip sees than the shape oscillation.

I see this effect regularly in my images, so it's quite common, just needs the right combination of water splashing and shutterspeed.

note: I went web searching and found this site that shows something like (but not exactly) what I'm talking about in item 2...see figure 3. here

Thanks, Mark. The more I think about it, a polarization effect is less likely than your explanation above. I think you've got it, old chap!

--P

Here's a 3:1 enlargement of the droplets.

Interesting assessment Mark. Here is another portion of the same image that shows unbroken light trails (just above the yellow marks I made) from reflections of the sun on the surface of the stream. Seems this would lend credence to your oscillating water droplet explanation.

On further reflection, I expect that the drop oscillation is most likely responsible, not the changing angle. I'd need to do a changing angle estimation to support idea 2 and those are not simple.