Does Autoclean on the 5D2 do anything?

[Loswr]

[Dust on the sensor, much like the annoying dust on film, especially large format film, was not apperture specific... if the dust was there, it would be seen whether shooting 1.4 or f22... Couldn't it suggest the F22 dust is more dust in the lens because I would think dust on the sensor would be same hard outline of light not getting to the sensor either way...

Nope. Sensor dust is aperture sensitive - it becomes progressively more evident at narrower apertures. Lens dust, for all intents and purposes, is invisible in the final image. The difference from film is where the dust is in relation to the light recording medium. With film, the emulsion is right out there for the dust to settle on. But when we say 'dust on the sensor' we really mean is 'dust on the dichroic filter on top of the low pass filter on top of the IR cut filter on top of the phase conversion filter on top of another low pass filter on top of microlenses on top of the actual photosites of the sensor'.

In fewer words, unlike film, dust 'on the sensor' is not at the image-forming plane, but some distance in front of that plane, so diffraction partly determines the size and darkness of the blur spot.

Or, in as few words as I can manage: sensors aren't film.

 

[awinphoto]

[Dust on the sensor, much like the annoying dust on film, especially large format film, was not apperture specific... if the dust was there, it would be seen whether shooting 1.4 or f22... Couldn't it suggest the F22 dust is more dust in the lens because I would think dust on the sensor would be same hard outline of light not getting to the sensor either way...

Nope. Sensor dust is aperture sensitive - it becomes progressively more evident at narrower apertures. Lens dust, for all intents and purposes, is invisible in the final image. The difference from film is where the dust is in relation to the light recording medium. With film, the emulsion is right out there for the dust to settle on. But when we say 'dust on the sensor' we really mean is 'dust on the dichroic filter on top of the low pass filter on top of the IR cut filter on top of the phase conversion filter on top of another low pass filter on top of microlenses on top of the actual photosites of the sensor'.

In fewer words, unlike film, dust 'on the sensor' is not at the image-forming plane, but some distance in front of that plane, so diffraction partly determines the size and darkness of the blur spot.

Or, in as few words as I can manage: sensors aren't film.

I'd have to take your word, but just thinking of it logically, the dust on the film would be in rough proportion to the dust on the sensor... it's not like the dust on the sensor is floating in front of the sensor and a narrow beam of light would make it in focus, plus if i remember right, the light leaving the lens elements (lens assembly) to the image plane (sensor/film) is channeled and designed to bend the rays to meet at a specific point to obtain focus... It shouldn't be scattered light bouncing all over the place inside the camera so the channeled light, to me, should show the dust at any aperture, but I haven't spent much time looking into it and thinking out loud... Anyways as long as i dont see dust in my images, i'm good.

 

[awinphoto]

neuro, after re-reading your post again, you're saying the camera sensor dust settles on the filters infront of the sensor and not the sensor itself, am I getting this right? How much distance is inbetween the area the dust on the filter in relation to the sensor? thinner than a fingernail? still seems awful close to be similar to dust on film, but I'll take your word and go from there...

 

[Loswr]

Actually, it is sort of like the dust is floating in front of the sensor, except rather than floating, it's stuck what amounts to a piece of glass covering the sensor. You're not seeing the dust, you're seeing the shadow of the dust. Try this - take a light source a good distance from a surface (light over a dining table, for example). Put your hand right next to, but not touching, the table. That dark, hard-edged shadow that's about the same size as your hand is dust on film. Now, lift your hand a few inches. That lighter, softer-edged shadow is dust on a sensor. That's a simplified case, obviously, but similar in concept. That extra space between the dust and the photosites, unlike film where there's no space, means diffraction operates on sensor dust, and wider apertures literally allow the light waves to bend around the dust speck and fill in the shadow cast by the dust.

 

[Loswr]

neuro, after re-reading your post again, you're saying the camera sensor dust settles on the filters infront of the sensor and not the sensor itself, am I getting this right? How much distance is inbetween the area the dust on the filter in relation to the sensor? thinner than a fingernail?

Exactly. Depends on the sensor and the design of the filters, but in general the stack of filters over the sensor is ~2.5mm thick (about as thick as 6 or 7 fingernails stacked up). In fact, there's even more space in cameras with the self-cleaning system. The sensor itself isn't shaken to remove the dust, it's the first/horizontal low-pass filter that vibrates. So, that would need to be separated a bit from the remainder of the filters (IR cut, quarter-wave plate, and second/vertical low pass filter) to allow it room to vibrate.

 

[awinphoto]

Actually, it is sort of like the dust is floating in front of the sensor, except rather than floating, it's stuck what amounts to a piece of glass covering the sensor. You're not seeing the dust, you're seeing the shadow of the dust. Try this - take a light source a good distance from a surface (light over a dining table, for example). Put your hand right next to, but not touching, the table. That dark, hard-edged shadow that's about the same size as your hand is dust on film. Now, lift your hand a few inches. That lighter, softer-edged shadow is dust on a sensor. That's a simplified case, obviously, but similar in concept. That extra space between the dust and the photosites, unlike film where there's no space, means diffraction operates on sensor dust, and wider apertures literally allow the light waves to bend around the dust speck and fill in the shadow cast by the dust.

I guess I always assumed for the most part the sensor I see with the mirror locked up and in the self cleaning mode was the actual sensor itself but if the filter is separating the dust from the filter, then it makes sense I suppose... As I've always said I'm not too geeky into pretending to know every nuance of the makings of the camera, just how to shoot manually and have it work when i need it how i need it... thanks for the insight.

 

[epsiloneri]

awinphoto, just try the experiment I suggested (taking empty sky with the smallest aperture). And prepare for a shock Then take the same image with fully open aperture and watch the dust disappear...

 

[katwil]

awinphoto, just try the experiment I suggested (taking empty sky with the smallest aperture). And prepare for a shock Then take the same image with fully open aperture and watch the dust disappear...

I can vouch for this. Be sure to switch off the auto focus on your lens.

 

[TexPhoto]

One thing Canon could do is make auto clean work a little more often. How about after every lens change, and every couple hundred photos if there has been no change.

How about every time you switch your camera on/off? That's the default on the cameras I've used.

Why would I switch my camera off? It won't take pictures that way, and does not extend battery life. The only reason I do shut it off is to run the cleaning cycle, but that's not really automatic.

 

[epsiloneri]

Why would I switch my camera off? It won't take pictures that way, and does not extend battery life.

That's interesting, how habits are different. I always switch a camera off when I'm not using it, not so much for autoclean, but to lock buttons and avoid accidently activating various functions. I also would have thought that it does extend the battery to not be in stand-by mode, but that's nothing I have tested so if you say differently, I'd just have to take your word for it.