[UnknownVT]

Nope, I have to stick with my first observation. Mine are on the sensor, and they do vary depending on aperture setting.

Sorry I am still dubious -
however I'm not going to put dust on my sensor to try it out.

It doesn't make sense to me - I have many years with film photography and if there was dirt etc in the film gate - ie: dirt obstructing the film, on the film - it shows up regardless of the aperture set since it is blocking light at that point/area
- it ought to be the same for a digital sensor - any dirt/dust on the sensor blocks the light getting to the sensor at the dirt's location area/spot.

Perhaps you can please try to explain why aperture should affect the visibility of dust on the sensor?

Thanks,

[magnum1]

Perhaps you can please try to explain why aperture should affect the visibility of dust on the sensor?

Thanks,

I honestly can't as I am not knowledgeable enough. All I know is if I take two pictures, one at f2.8 and one at f11, spots will show up much more at f11. Have no idea why????

[UnknownVT]

I honestly can't as I am not knowledgeable enough. All I know is if I take two pictures, one at f2.8 and one at f11, spots will show up much more at f11. Have no idea why????

OK here's my attempt at an explanation -

Preamble -

First, I have noticed that dirt on film seems to show up sharper than dirt on digital sensors. On digital sensors they kind of look blurry.

I am speculating that dirt/dust on the sensor is on a plane other than the actual plane of detection/recording -

eg: the Bayer filters/lenses must have a finite thickness - which covers the actual pixel site.

Then there is the optical anti-aliasing filter which again has a thickness -

Plus there may be some anti-static anti-dust coating(s) that may again have some thickness.

So between the two optical systems and any coatings - there is a finite amount of thickness/space that separates the dust from the actual pixel detecting site/plane -
so it's not quite like film which has the emulsion side facing the lens so any dirt/dust on it is on the actual capture/detecting plane.

Hopefully people are still following me?

There is depth of focus (Wikipedia link) not to be confused with depth of field - this is the tolerance/depth that a projected image from the lens is actually in focus on the detection plane - eg: sensor or film.

The depth of focus is dependent on the aperture - so the smaller the aperture the greater the depth of focus - so with a smaller aperture the dirt will tend to be more in the focus part of the projected image - whereas with a larger aperture the dirt is more in the out of depth of focus part of the projected image.

This may explain your observation?

Anyone else care to comment, or correct this, please?

[Marc Sabatella]

I've never understood the reasons, but it's certainly a well-known fact and easily verified next time you get dust on your sensor: the spot sharpens as you stop down, every time.

BTW, a couple of differences between digital and film worth mentioning:L

With film, you never have the opportunity to find out if the spot gets sharper when you stop down, because you never get to see two shots with the same dust spot.

With digital, the spot might cover a few pixels but due to demosaicing, have a larger effect than that - but conversely, the pixels covered by the spot might not be completely black.

[magnum1]

I've never understood the reasons, but it's certainly a well-known fact and easily verified next time you get dust on your sensor: the spot sharpens as you stop down, every time.

I'm just glad I wasn't imagining this....

[leadfoot]

The reason that sensor dust becomes more apparent is because of the "virtual image", as you stop down a lens.

A pinhole filter, ie an infinitely small, circular hole, is a perfect lens. Everything beyond the pinhole is in perfect focus. The "virtual image" is the exact inverse, everything behind the pinhole filter is also perfectly in focus (if you look backwards through the pinhole). It does not matter where the imaging sensor is in relation to the pinhole - everything is in perfect focus.

The "star effect" of a lens is a diffraction effect from the aperture blades of the lens, and is not a perfect pinhole. If the aperture blades formed a perfect circle as you stop down, there would be no star effect, but would just form a more in focus object.

The net result is that the more you stop down, the more in-focus anything in the focal plane is, and as you approach a perfect lens (a pinhole), everything is in focus. So, if you set your camera and lens to F22, base ISO, manual focus, and shoot a blank wall, anything you see will be on the focal plane, ie the sensor, the virtual image.

Sorry for the lame, geek explaination - it's been too long since college optics. The pinhole filter is the 2 dimensional equivalent of the impulse response of an electrical circuit. Impossible to achieve, but it clearly shows the result of dust on the sensor.

[UnknownVT]

The reason that sensor dust becomes more apparent is because of the "virtual image", as you stop down a lens.

Yes , this is true if the dust is not actually on the focal plane.

If the dust is actually on the focal plane - then it ought to be in focus no matter waht the aperture set - that was the whole crux of the previous discussion.

If the dust is not actually on the focal plane - then aperture comes into play -
(our lenses are NOT pinholes) - the depth of focus is the operative effect.

[leadfoot]

Yes , this is true if the dust is not actually on the focal plane.

If the dust is actually on the focal plane - then it ought to be in focus no matter waht the aperture set - that was the whole crux of the previous discussion.

If the dust is not actually on the focal plane - then aperture comes into play -
(our lenses are NOT pinholes) - the depth of focus is the operative effect.

Yup. However, dust/dirt is never directly on the raw silicon (the focal plane) - if it was there would be a big divot in every image.

The raw silicon is coated with a number of things, like the baysian color mask, anti-aliasing filter, etc, and dust sits on top of these coatings and coverings. The thickness of these coatings is many times the wavelength of visible light. As the lens is stopped down, the DOF increases, for both the image, and dust near the raw silicon.

In any case, it's an interesting effect, how dust and dirt on the sensor becomes more in focus as the lens is stopped down, and the DOF increases for both the image and dust.

[Unr3a1]

I have a similar problem with my new K-x, except mine has a single pixel in the lower right quandrant that shows up as a red light. When I look at the pictures on my computer, it appears to be a single pixel. If I go black and white with the pics, it shows up as white. It's a single pixel and I am not sure what is causing it.

[magnum1]

I have a similar problem with my new K-x, except mine has a single pixel in the lower right quandrant that shows up as a red light. When I look at the pictures on my computer, it appears to be a single pixel. If I go black and white with the pics, it shows up as white. It's a single pixel and I am not sure what is causing it.

Sounds like a dead sensor pixel and not dirt.

[UnknownVT]

The raw silicon is coated with a number of things, like the baysian color mask, anti-aliasing filter, etc, and dust sits on top of these coatings and coverings. The thickness of these coatings is many times the wavelength of visible light. As the lens is stopped down, the DOF increases, for both the image, and dust near the raw silicon.

Hello? ..... didn't I just say that previously in post #18 ? -

OK here's my attempt at an explanation -

I am speculating that dirt/dust on the sensor is on a plane other than the actual plane of detection/recording -

eg: the Bayer filters/lenses must have a finite thickness - which covers the actual pixel site.

Then there is the optical anti-aliasing filter which again has a thickness -

Plus there may be some anti-static anti-dust coating(s) that may again have some thickness.

So between the two optical systems and any coatings - there is a finite amount of thickness/space that separates the dust from the actual pixel detecting site/plane -
so it's not quite like film which has the emulsion side facing the lens so any dirt/dust on it is on the actual capture/detecting plane.

Hopefully people are still following me?

There is depth of focus (Wikipedia link) not to be confused with depth of field - this is the tolerance/depth that a projected image from the lens is actually in focus on the detection plane - eg: sensor or film.

The depth of focus is dependent on the aperture - so the smaller the aperture the greater the depth of focus - so with a smaller aperture the dirt will tend to be more in the focus part of the projected image - whereas with a larger aperture the dirt is more in the out of depth of focus part of the projected image.

Joking aside - thank you for the confirmation -
that dust on the surface of the sensor is not actually on the pixel site detection/recording plane -
so this is exactly the reason why dust on sensors are normally out of focus, and why stopping down the lens, as observed by some, seems to make it clearer.

In any case, it's an interesting effect, how dust and dirt on the sensor becomes more in focus as the lens is stopped down, and the DOF increases for both the image and dust.

Sorry have to be careful (and pedantic?) here -
DoF normally means Depth of Field - which is the parts of the image deemed in focus in front of the lens (ie: in the actual physical scene) -
whereas depth of focus (Wikipedia link) is different - it's the parts of the projected image in focus behind the lens on the film or sensor.

[bobore]

I have a similar dust spot on the upper left corner of my pictures. Lifting the mirror and looking straight on at the sensor, where would this speck of dust actually be located? Should be simple to figure out, but I'm barely awake!

[UnknownVT]

I have a similar dust spot on the upper left corner of my pictures. Lifting the mirror and looking straight on at the sensor, where would this speck of dust actually be located? Should be simple to figure out, but I'm barely awake!

It ought to be the diametrically opposite corner

The lens projected images are supposed to be upside down and mirror imaged.

[Unr3a1]

Sounds like a dead sensor pixel and not dirt.

Sorry if this is a stupid question, but is there anything that can be done to fix this, or do I just need to edit it out in pp? Thanks.

[JeffJS]

Sorry if this is a stupid question, but is there anything that can be done to fix this, or do I just need to edit it out in pp? Thanks.

A pixel mapping function is built into most cameras. It is either in the setup or custom menu. I don't remember which at the moment. Just run it, dead/hot pixel should be gone then. Dirt will always show up black.

https://www.pentaxforums.com/forums/pentax-beginners-corner-q/109456-dirty-sensor.html