Unusual photo

[dave_seagren]

<p>I need to know where I can send a copy of a photo I took to find out how it occurred. It was a shot out the front of an open air bi-plane through the propeller and I got 7 stopped images of the propeller edge with no blurring and the overall view image.</p><div></div>

[JDMvW]

<p>I was going to say that Your shutter speed was fast enough to stop the motion, until I downloaded your image which shows no propeller at all. I don't know what those bars across the screen are, but they aren't the propeller.</p>

[dave_seagren]

<p>If it was a fast shutter speed would I not just get 1 stopped image of the prop not 8?</p>

[seismiccwave]

<p>The shutter curtain has two pieces. As the first curtain opens the second curtain closes right behind. So you really get a thin slit that moves from top to bottom or bottom to top which let the light in to expose your media be it film or digital. As the shutter slit is traveling the propeller is spinning at the same time. While the shutter curtain is traveling the propeller blades (probably a two blade propeller) passed over the curtain multiple times.</p>

[JDMvW]

<p>Here's the above picture resized so it can be shown in-line. Note that whatever the lines are, they are over the fuselage of the plane, not in front of the plane.</p><div></div>

[dave_seagren]

<p>They definitely are 8 images of the propeller edge. The prop was wooden painted red on the tips. I can post a copy of the plane and prop if needed.</p>

 

[dave_seagren]

<p>The picture was tken through a plexiglass windscreen. Are the images on the fuselage not reflections?</p>

[JDMvW]

<p>If it's the prop, then how did it get between you and the front of the plane?</p>

[seismiccwave]

<p>The lines on the cowl is a reflection of the propeller. It is definitely a two bladed propeller from a Waco bipe.</p>

[dave_seagren]

<p>Did my photo come from a reflection off the windscreen?<br>

I believe it was a Waco bi-plane.</p>

 

[conwaygroup]

<p>Hansen has it right. The props RPM was most probably very close to being sync'd with your shutter speed. If the prop had been going slightly slower it should have appeared as a single blade. If it was going faster it would have turned into a blur.</p>

[patrick j dempsey]

<p>Now, if you remember how fast the shutter was set, you can approximate how fast the prop was moving! </p>

[david_klaffenbach]

<p>I think Hansen has the right principle but the wrong technology.</p>

<p>If the prop was turning at 1800 rpm (just a guess which makes for easy math), it would make one revolution every 1/30s or one blade-pass every 1/60s. Focal-plane shutters in SLRs tend to have sync speeds around 1/250 or so these days, and so they would be completely open for longer durations such as 1/30s.</p>

<p>I think this may be due to use of an electronic "rolling-shutter" in which data is captured row-by-row off the sensor as it read out. Also, your EXIF data is weird: it says ISO100, f/3, 1/30s which would be overexposed by many many stops if correct and it was just after noon on a sunny day.</p>

<p>Were you using a camcorder with a still capture feature or is this a capture from a video stream?</p>

[phule]

<p>David,</p>

<p>Opanda IExif returns: f/5.6 at 1/125 and ISO 200. This is a Nikon D100</p>

[david_klaffenbach]

<p>Rob,</p>

<p>I don't see that (D100, 1/125 or f/5.6) at all in the EXIF of either image. I'm using the EXIF option in IrfanView. Which image did you look at?</p>

[phule]

<p>I'm such an idiot. You're absolutely correct. I checked the wrong image in Opanda. (Similar file names on the desktop from a different thread). Sorry.</p>

[oskar_ojala]

<p>David got it right, this is very common with cheap cameras such as cellphone cams; the sensor is scanned row by row, which means that the whole image is not captured instantaneously, rather, it's capture piece-by-piece. The effect is somewhat similar to really old cloth shutters.</p>

[geoffs1]

<p>Oskar has it right. It's a very well-known artifact in shutterless digital cameras:<br>

http://www.photo.net/digital-camera-forum/00CKgJ<br>

http://people.rit.edu/andpph/text-focal-plane-artifacts-in-digital-cameras.html</p>

<p> </p>

[peter_mounier]

<p>I think the answer is that the plane dropped a few feet during the exposure. The view in the distance didn't change because it's too far away to have affected the perspective, and the plane's fuselage is stationary to the camera, so it didn't change. The prop was exposed multiple times as the plane dropped, changing the relative position to the plane.</p>

<p>Peter</p><div></div>

[conwaygroup]

<p>Maybe this link will help. The slit of light that Hansen refers to moves from top down across the sensor. Imagine the prop spinning as the shutter drops. The shutter drops and as each new segment of the sensor is exposed, half the prop will have rotated out of frame and the other half into frame but in a slightly different position. In other words, every other line in the photo would represent the same half of the propeller blade.<br>

<a href="http://regex.info/blog/2008-09-04/925">http://regex.info/blog/2008-09-04/925</a></p>

[peter_mounier]

<p>If the shutter was rolling, wouldn't there be vertical blurs?</p>

<p>Peter</p>

[conwaygroup]

<p>The rolling shutter and a change in altitude are probably both at play. Good point Peter. An altitude change would explain why the lines don't radiate from a central hub.</p>

[stp]

<p>You guys are better than CSI. Case solved in about an hour just by casually dropping in.</p>

[dave_seagren]

<p>The picture was taken with an inexpensive video camera as a still image.</p>

[geoffs1]

<p>I've laten literally thousands of slow shutter-speed photos of prop planes with a standard DSLR (Nikon D70, Canon 20D, 40D, 1D2IIn, and 5D) and I've never seen anything like the OT's photo, or the ones in the links I posted a couple of items up. When the sensor readout occurs with the shutter closed, there are no artifacts. It's only with "electronic shutters" where the sensor readout is occurring along with subject motion that you get artifacts like these.</p><div></div>