Bizarre pictures from cheap digital camera!!

Chris Beney said:

Jamie said:

Could it be that in addition to the possibilities proposed above,
the blades do bend somewhat when under stress? Were they moving
fast?

Click to expand...

The man said they were solid.

Click to expand...

I must have missed that.

Chris Beney said:

The C shape is only remarkable because of the position of the blade
and of the relative timing of camera electronics and rotational
speed.

Click to expand...

Would be a heck of a shot to try to reproduce if the speculation of scan rates, etc. is correct!

Chris Beney said:

If the camera scans from the top and the blades are turning
clockwise then if it starts scanning when the upper blade is at the
"north west" position, as it scans down the upper blade the blade
will turn a bit for each line scanned, making the top ha;lf of the
C. The hub will seem correct then the lower half of the blade
(still clockwise but now right to left) will be scanned as it turns
to the "south west" position, making the bottom half of the C
shape. You don'y need bent blades or other magic to explain that.

Click to expand...

No... that's true. I had not read that teh blades were "solid"... so I was speculating it was an "all of the above" scenario.

Jamie

I've got a theory about the secondary floating prop as seen in this picture.



A little background first.

The camera recorded the images on my Compact Flash card in the standard format so I can insert this compact flash card in my G3 and play the images on the G3's LCD screen. I did this, and set the LCD to display the histogram. The image above was taken at midday with a very bright Texas sun shining directly at the white-painted wooden propellor. Hence, the propellor was extremely bright. This showed up on the histogram as a very narrow spike on the right hand side that went "threw the roof". On the actual display, the prop is blinking indicating severe overexposure. (the darker areas of the prop are the shadow of tree leaves)

Now, here is my theory. I know in the world of audio, if you overload an amplifiier, expecially in severe cases, you generate a host of additional frequencies, the most prominent being overtones or harmonics, namely the 2nd harmonic, an octave above the original frequency.

Could there be a similar phenomenon with light and the "light amplifiers" inside a digital camera? With the severe overload of the white propellor, couild there be a harmonic generated which would in a sense duplicate the original but at a different part of the image? This might explain the "floating prop" and also would explain why ONLY the prop itself is floating and not the hub and the brake drum - because the hub and brake drum are painted black and would NOT have experienced the overload.

What do you guys think? It might make for some interesting understanding of the actual way that light is captured by these digital cameras.

I'll get up to the big city (Amarillo) in a day or two and check for one. I'm hoping to work with a ballet troup in the spring, maybe I'll try this beast on them.

I've always enjoyed fooling around with oddball cameras -- sometimes I can even pass it off as art.

Here's one from my first digital camera back around '97 or '98, a tethered device just half a step above a webcam. The pixel patrol won't like it, but several of my friends have prints on their walls.

Georg Steiner said:

I think I know the answer:
... a rather slow camera that reads line by line with
quite some delay!...

Click to expand...

You are correct, sir.

This also explains the "Floating Blade" problem. What you're seeing is the tip if the blade you're later seeing at the bottom of the image. When the scan started, the tip was caught at the top of the image, then outran the scan, leaving a "floating" blade.

It is becoming clearer to me that the deficiency is in the processor, most likely nothing to do with the CCD. Where's the money? Besides the CCD It's in the processor and that's what is cheap in this case. We don't know this wouldn't duplicate itself in a number of digicams, maybe even some higher quality ones. Also it is readily apparent that the shutters are electronic (scans). I think I knew this all along but have not actually studied the inner workings of digicams.

Jamie said:

Could it be that in addition to the possibilities proposed above,
the blades do bend somewhat when under stress? Were they moving
fast?

jamie

Click to expand...

--
http://www.DForbesRowanPhotos.OrangeCountyandSurrounding.PhotoShare.co.nz
D. F. R.

I suppose but that's getting pretty specific for a guessing game. One thing for sure, you should have kept the little homeless thing, what a conversation piece it is. Who knows what types of digital effects you would have been able to come up with.

But if bright objects created "harmonics" then many pictures would be loaded with them. I still think it's something to do with the motion vs the shutter and write speed/write direction of the CCD and processor. The blade curve is far easier to explain - I'm not even surprised to see it - than the floating prop, that one's the real trip. What a fascinating thread. Somewhere in the Forums there is an engineer who has a more precise answer. You could post a single link to several "hot" threads and see who stumbles in. This thread will probably live on for months.

Gus Stangeland said:

I've got a theory about the secondary floating prop as seen in this
picture.



A little background first.
The camera recorded the images on my Compact Flash card in the
standard format so I can insert this compact flash card in my G3
and play the images on the G3's LCD screen. I did this, and set the
LCD to display the histogram. The image above was taken at midday
with a very bright Texas sun shining directly at the white-painted
wooden propellor. Hence, the propellor was extremely bright. This
showed up on the histogram as a very narrow spike on the right hand
side that went "threw the roof". On the actual display, the prop is
blinking indicating severe overexposure. (the darker areas of the
prop are the shadow of tree leaves)
Now, here is my theory. I know in the world of audio, if you
overload an amplifiier, expecially in severe cases, you generate a
host of additional frequencies, the most prominent being overtones
or harmonics, namely the 2nd harmonic, an octave above the original
frequency.
Could there be a similar phenomenon with light and the "light
amplifiers" inside a digital camera? With the severe overload of
the white propellor, couild there be a harmonic generated which
would in a sense duplicate the original but at a different part of
the image? This might explain the "floating prop" and also would
explain why ONLY the prop itself is floating and not the hub and
the brake drum - because the hub and brake drum are painted black
and would NOT have experienced the overload.
What do you guys think? It might make for some interesting
understanding of the actual way that light is captured by these
digital cameras.

Click to expand...

--
http://www.DForbesRowanPhotos.OrangeCountyandSurrounding.PhotoShare.co.nz
D. F. R.

I think you have discovered that the Space Time continuum does exist.

To prove this theory:

1.) Strap your self the propeller and see if you visit another time, Oh and take your camera along to take pictures. Of course if you capture images of tomorrow you could create a time paradox, Or if they are from yesterday how will you know?

Just funning everyone,

2.) Most likely AS already stated a combination of effects, shutter speed, the tips of the blades travel at a different speed then the root of the blades, and the speed at which the processor captures the image to memory.

3.) My first theory is right after all…

HAPPY NEW YEAR EVERYONE.

DirkvdB said:

This also explains the "Floating Blade" problem. What you're seeing
is the tip if the blade you're later seeing at the bottom of the
image. When the scan started, the tip was caught at the top of the
image, then outran the scan, leaving a "floating" blade.

Click to expand...

Could you explain more precicely the mechanism of this? If it is a single scan that makes the C shape I can't see how it can make an S out of just the tip. If it is multiple scans that make the C it is very difficult to understand why the C is so smooth, wouldn't the synch have to be near perfect, and wouldn't it be extraordinary if that occurred in several photos with varying wind? And also please explain the issue of the similar light and dark patches on the blades, which I have taken to be shadows of leaves.
Chris Beney

Gus Stangeland said:

I've seen some strange pictures from cameras, but NEVER anything
like this. This wincharger (wind-powered generator) was spinning in
the wind when I took this picture. The propellor on this wincharger
is 6 feet in diameter and STRAIGHT! Look what the camera did!!



Here's another picture, showing 2 winchargers. The one on the left
was NOT spinning and thus looks normal.

Click to expand...

Gus Stangeland said:

How can this possibly be?? I have NOT altered the photos - just
downsized them. The camera was a Digitrex with 1600x1200 pixel
resolution (1.9 megapixels). It had fixed focus, fixed aperture,
auto exposure. Can anyone explain how in the world it came up with
these pictures?

Click to expand...

--
JWP

DirkvdB said:

Georg Steiner said:

I think I know the answer:
... a rather slow camera that reads line by line with
quite some delay!...

Click to expand...

You are correct, sir.

This also explains the "Floating Blade" problem. What you're seeing
is the tip if the blade you're later seeing at the bottom of the
image. When the scan started, the tip was caught at the top of the
image, then outran the scan, leaving a "floating" blade.

Click to expand...

So, if that is really the case, then what would happen when photographing some other moving objects? Let's take the example of a racecar moving from left to right across your field of view (perpendicular to you). Would this 'slow scanning' camera take a picture that makes the car look like it is leaning backwards, with the top of the car more to the left than the bottom of the car? How about something moving upwards in front of the camera? Would there be a point where the object would disappear from the photo because the top of the image had already been scanned but the object was gone from the bottom before the camera got to that part?

Just trying to understand what's going on behind the scenes. In any case this seems very odd.

Rob

Check this link. Another camera made from a flatbed scanner.
http://www.sentex.net/~mwandel/tech/scanner.html

Scroll down and look at the picture of the moving garage door. This will convince you.

Bill

Rob Bell said:

DirkvdB said:

Georg Steiner said:

I think I know the answer:
... a rather slow camera that reads line by line with
quite some delay!...

Click to expand...

You are correct, sir.

This also explains the "Floating Blade" problem. What you're seeing
is the tip if the blade you're later seeing at the bottom of the
image. When the scan started, the tip was caught at the top of the
image, then outran the scan, leaving a "floating" blade.

Click to expand...

So, if that is really the case, then what would happen when
photographing some other moving objects? Let's take the example of
a racecar moving from left to right across your field of view
(perpendicular to you). Would this 'slow scanning' camera take a
picture that makes the car look like it is leaning backwards, with
the top of the car more to the left than the bottom of the car?
How about something moving upwards in front of the camera? Would
there be a point where the object would disappear from the photo
because the top of the image had already been scanned but the
object was gone from the bottom before the camera got to that part?

Just trying to understand what's going on behind the scenes. In
any case this seems very odd.

Rob

Click to expand...

BillS said:

Check this link. Another camera made from a flatbed scanner.
http://www.sentex.net/~mwandel/tech/scanner.html
Scroll down and look at the picture of the moving garage door.
This will convince you.

Click to expand...

I see no second garage door anywhere, where should I be looking?
Chris Beney

I was only offering an example of distortion caused by a relatively slow image scan. The image of the garage door, opening and closing clearly demonstrates this. I understand this does not explain the second prop blade.

Bill

Chris Beney said:

BillS said:

Check this link. Another camera made from a flatbed scanner.
http://www.sentex.net/~mwandel/tech/scanner.html
Scroll down and look at the picture of the moving garage door.
This will convince you.

Click to expand...

I see no second garage door anywhere, where should I be looking?
Chris Beney

Click to expand...

I'd love to see them. Care to post, or email?

LizL said:

Those are amazing! I wonder if you can duplicate the distortion by
shooting other moving objects? Whether or not, it is still
interesting.

I got some very strange results once when I was taking photos
inside unlit slave quarters in the basement of a Confederate era
mansion. I never posted the photos. I figure people would just
accuse me of doctoring them, too!
--

Happy with G 2 since Dec., 2001; also own Canon S 900 printer; p
Base believer (Spaces in names are there to protect the search
process)
Advice for Beginners (tons of links and info):

http://forums.dpreview.com/forums/read.asp?forum=1010&message=3697387

LizL

Click to expand...

--
http://www.goff.tk

DirkvdB said:

This also explains the "Floating Blade" problem. What you're seeing
is the tip if the blade you're later seeing at the bottom of the
image. When the scan started, the tip was caught at the top of the
image, then outran the scan, leaving a "floating" blade.

Click to expand...

This is the best (and true, IMHO) explanation for the floating blade. I am absolutely convinced that this effect is also created by the line-wise scanning of the sensor. If you imagine a clockwise rotating propellor and a up-to-down CCD scanning it is not so difficult to understand the effect (in all three pics with floating blades, by the way).
-Frank

Indeed it is very likely that the effect is created in some way or other by the line-wise scanning of the sensor. But as to that being an 'explanation' and 'not so difficult to understand' I fear I lose you there.

A valid explanation for the C shape of the main blade has been spelled out carefully on this thread, but no-one has yet explained the C and S on the same picture other than simply saying 'it's the scanning'.
Chris Beney

Frank Berger said:

DirkvdB said:

This also explains the "Floating Blade" problem. What you're seeing
is the tip if the blade you're later seeing at the bottom of the
image. When the scan started, the tip was caught at the top of the
image, then outran the scan, leaving a "floating" blade.

Click to expand...

This is the best (and true, IMHO) explanation for the floating
blade. I am absolutely convinced that this effect is also created
by the line-wise scanning of the sensor. If you imagine a clockwise
rotating propellor and a up-to-down CCD scanning it is not so
difficult to understand the effect (in all three pics with floating
blades, by the way).
-Frank

Click to expand...

no text

Here are some more pictures from this camera. Maybe these will clear things up - or - make things even more complex!

This first one is similar to the previous pictures except there are no leaf shadows on the propellor. Also, as you can see, there is no additional floating prop.



This camera takes a LONG time to store each picture (which I was not used to). So I snapped this next picture at the propellor, and thinking I took the picture, I lowered the camera. The camera finally got around to snapping the picture right as I was lowering the camera. So the camera was moving in a downward direction. Note the pronounced curvature of the tower and the building.



In the picture below, here we see another one of those mysterious floating props. Now here are some interesting observations. Note that the leading edge of the propellor has a plastic protection strip - seen as a greenish-grey strip on the leading edge of the prop. It appears on the right side of the upper half of the prop and on the left side of the lower half of the prop (it is on the leading edge - the prop is rotating clockwise). Now take a close look at the floating prop. If you look carefully, you will see that the plastic protection strip is ALL on one side, namely the right side. This says that this floating prop is really only ONE HALF of the prop - apparently the upper half. How bizarre is that?? But wait, there's more! Notice that the upper half of the full prop is shaded by the tree - almost the entire half-prop is shaded. But when you look at the floating prop, it is NOT shaded! This would imply that this floating image is actually of the lower half of the prop and not the upper half as deduced by the leading edge protection. How do you explain that?



This next picture below shows what the unit looks like with no wind. The propellor is not moving.



And finally, below, you can see that the tower and the building are straight and not bent as in one of the pictures above/

Here is my explanation about the S-shaped blade:

Pretend that the scanning goes from top to bottom. The scanning reaches the blade that points to about 12 o'clock. As the scanning moves down, the blade moves from 12 o'clock to 3 o'clock and at 3 o'clock the vertical speed of the blade is the biggest (and bigger than the vertical speed of the scanning) wich results in the fact that the blade leaves the scanner.

Why is it S-shaped? Why not linear? Well the answer is in the proportion between the speed of the blade and the scanning. I guess it has to do with sinus/cosinus waves when the blade moves in the shape of a circle. There are actually TWO parameters that influences the different angles inside the "painted" S-blade. That is the horisontal speed of the blade and the vertical speed of the blade. Between 1 o'clock and 2 o'clock none of these speeds are big. Because of this the painted angle is another one (steeper) in the middle of the painted S-blade.

So, what you actually see when you see the S-shape is not a whole blade but rather the tip or at least an outer part of the blade, scanned at different moments. It's a little bit like those photos that you may have seen taken at the finish line in the Olympic games.





Also remember that the propeller is probably so fast that it may move one or two times around in the same time as the scanning has scanned the circle shaped area where the propeller has moved. Because of this you can see more than two blades.

I hope that my teaching skills is good enough. Otherwise I should change my career.

Chris Beney said:

Frank Berger said:

DirkvdB said:

This also explains the "Floating Blade" problem. What you're seeing
is the tip if the blade you're later seeing at the bottom of the
image. When the scan started, the tip was caught at the top of the
image, then outran the scan, leaving a "floating" blade.

Click to expand...

This is the best (and true, IMHO) explanation for the floating
blade. I am absolutely convinced that this effect is also created
by the line-wise scanning of the sensor. If you imagine a clockwise
rotating propellor and a up-to-down CCD scanning it is not so
difficult to understand the effect (in all three pics with floating
blades, by the way).
-Frank

Click to expand...

Click to expand...

--
Looking for my first Digicam (for about 500 €)

Gus Stangeland said:

5. I cannot run any more tests because I returned the camera the
next day to get my refund. I did not want to wait until after
Christmas to hear "sorry, that was a special deal, you can't return
it". (As it was I had to plead with the manager).

Click to expand...

After stirring up this hornet's nest, would you consider going back to Office Max and buying the camera back again so that you can do more tests? Inquiring minds want to know.

Wayne Larmon