Know your light source - LED's.

10-10-20

Some of you may think that this is the oddest thread in the world, and that I'm madder than a box of frogs, but bear with me.

Inspired by some photos in pix.ie, I tried cracking off a few shots of dripping water. Knowing that I didn't have sufficient light for my S7000 to capture high speed shots I employed the use of an LED powered headlamp.

So, strictly messing around, I shone the lamp at the water 90 degrees to the camera and started shooting.
I noticed this effect in the photos...



See the segmented lines, almost like train carriages?

Well, I was able to work out what caused them and it's down to the modulation of the light source I was using.
The LED lamp was a Pezel http://en.petzl.com/petzl/LampesAccueil with 4 modes, 1. On full brightness, 2. 75% brightness, 3. 50% brightness, 4. Flashing.

I had the lamp on full brightness, but it seems that the output is still modulated (meaning for example - on for 95% of time, off for 5%) somewhat and generates these segmented lines where a normal solid-state lamp would create continuous lines.

So, to prove this further, I retained the aperture and shutter settings and set the lamp to 50% brightness (modulating at 50:50 duty) and I got this stuttering effect...



So, it all goes to show that your light source isn't always what you think it is.
If I had been using a large/bright setup where all of the light source was derived from LED's and had been using very fast shutter speeds, I'm guessing above 1/500, I could conceivably open and close the shutter during the LED lamp's off period and received no light from the subject.

I'm laughing at the moment because somebody just voted 4/5 for the stuttering effect...

magicbastarder

i noticed something similar with the remote controls for the projectors in work - they have laser pointers built in, and if you sweep the spot quickly back and forth along a wall, it becomes apparent that the beam is not continuous.

DaireQuinlan

Reminds me of that perplexing thing that fajitas (? could be wrong) posted. Taking shots indoors under flourescent lamp and there was a darkish band along the bottom or top. I initially thought it was a shutter problem, but it actually turned out that the shutter speed was almost exactly the same as the frequency of the light fixture so half the frame was getting full illumination and then as the slot in the shutter curtain moved down the frame the light was switching off. V. wierd initially. If the shutter speed had been any faster then the entire frame would have been underexposed or normally exposed, if it had been any slower then it wouldn't have shown up at all.

Fajitas!

It was someone else, but I had seen it happen with a few of mine too!

10-10-20

Yes Daire, you get that with CRT TV/Monitor screens because of the line scanning method used to excite the pixels.
Slower speeds mean you get chopped displays, interlacing gives you a saw-tooth look to the screen and moire gets emphasized with the Hz.

Typhoon.

Thats something like this effect....http://www.metacafe.com/watch/598291/can_you_explane_this/

or the same reason that wheels on cars appear to be going backwards...where the fps matchs that of your eye (or video)

I know its a little off topic but could it not be the case so if you had a fast enough shutter speed you could capture in darkness even though the light is on...if you took the photo during one of these "dark pulses"....Theoretically

DaireQuinlan

Typhoon. wrote: »

Thats something like this effect....http://www.metacafe.com/watch/598291/can_you_explane_this/

or the same reason that wheels on cars appear to be going backwards...where the fps matchs that of your eye (or video)

I know its a little off topic but could it not be the case so if you had a fast enough shutter speed you could capture in darkness even though the light is on...if you took the photo during one of these "dark pulses"....Theoretically

Hmm, probably not completely dark, I think you'd be much more likely to get that effect where some portion of the field is dark. Thinking of it, the fastest that 35mm sutters actually travel is generally 1/250 (or thereabouts, check your fastest flash sync speed). Anything faster than that is caused by the top shutter "following" the bottom shutter leaf down before the bottom shutter leaf has reached the bottom. This means that the film/sensor will ALWAYS be exposed over the course of 1/250th of a second, even though each individual bit of it will be only exposed for 1/2000th of a second or whatever. So its probably more down to bad timing, or the actual frequency of the light fitting, rather than the shutter speed.

10-10-20

It's probably the same technology behind the Eircom and UK Lottery advert where people in a room are playing around and toss items into the air, only for the screen to pause and then pan sideways showing a 3D view of the paused scene. Very realistic.
That's similar to high speed photography where multiple cameras are setup on a linear plane and timed to expose in sequence...

Ah--- it was used in The Matrix. (Time slice (or bullet time))
http://en.wikipedia.org/wiki/Bullet_time

5uspect

I use something similar for my PhD work.

My experiments involve measuring fluid flow optically by tracking lots of tiny neutrally bouyant particles suspended in the flow. The problem is that a camera cannot capture a properly exposed image of the particles as they streak with the relatively long exposuse.
(The fluid velocity can be anywhere from 1 m/s up to transonic or supersonic speeds)

So I use a high powered pulsed Nd YAG laser that emits a 1mm thick light sheet. The camera is focused on a small region of this slice of the flow. The laser pulse only lasts about 8 nanoseconds where as the camera exposure is of order 100 micro seconds. Since the only light the sensor sees (I test in the dark because the light in the wind tunnel lab interfere :eek!) is from the 8 nanosecond pulse the particles are "frozen in time" but to get a proper exposure the laser must be very, very powerful.

Usually with this method (Particle Image Velocimetry or PIV) you take two very rapid exposures separated by a time difference that allows the particles to move about 8 pixels. The images are them cross correlated to get a 2D vector field. Back before digital they used to double expose film with two pulses, quite similar to the water drops. It was quite difficult do figure out the flow velocity. With the double exposure method we use you eliminate that problem.

I've attached a snazzy video of the results you get. It shows the velocity of the flow over a surface. The velocity begins to slow near the wall due to viscosity.

Spyral

Yes Daire, you get that with CRT TV/Monitor screens because of the line scanning method used to excite the pixels.
Slower speeds mean you get chopped displays, interlacing gives you a saw-tooth look to the screen and moire gets emphasized with the Hz.

forensic manual sez : use 1/15 shutter speed for a film monitor

10-10-20

Being pedantic - I reply: Is that a US or European manual?