Pixel shift resolution

[bronson]

It looks promising...

Don't know how it would be difficult to code this for magic lantern...
 

[ItsMeLenny]

Doesn't the actual sensor have to move? If that's the case it would be extremely difficult to program

Unless the stabiliser in lenses are electronically controlled by the camera, that'd be the only other way.

I think a lot of cameras have blurry imagery when zoomed in is because of chromatic aberration on a bad lens,
this might interest you: http://www.cs.ubc.ca/labs/imager/tr/2013/SimpleLensImaging/

[bronson]

Thanks for the link ItsMeLenny,
It's impressive..

Don't know how the clean sensor works.. it seems to me that it quickly shakes the sensor..
but turning around the bayer pattern requires to move the sensor in some micro-positions precisely..
 

[3DM1]

it could be that after a sensorclean is performed, the sensor isnt at the same subpixel position. if not, than you can align these shoots with PS manualy to create a superresolution picture.

to test this we would need a script that takes a picture, performs a sensor clean and another picture, because manualy start a sensor clean would shake the camera.