edform
Veteran Member
This has been an interesting thread, made more so by the quite graphic demonstration in Jeff's images that 1" sensor cameras need wide aperture lenses if they are to minimize the effects of diffraction.
It should also be born in mind that wide aperture lenses have problems with optical aberrations to a greater degree than small aperture designs, a fundamental difficulty that can only be defeated with complex, sometimes very complex, optical designs that cost scads of money to make. This is particularly a problem for wide range zoom lenses where aberration correction has to be maintained over the whole of the range. As a result, the lenses on modern super-bridge cameras are astonishingly complex devices. The lens on the RX10-III, as an example, is not only a seriously complicated physical assembly of glass types and shapes but is made hugely more complicated by the fact that the lens on its own would not be acceptable on any camera. Certain areas of aberration are not corrected in the lens, instead they are corrected with complex processor mathematics - the optical system is actually the lens plus the processor.
Returning to the main subject of image quality: it is a fact that as aperture grows smaller, diffraction increasingly degrades image quality but it is equally a fact that as aperture grows larger, lens aberrations increasingly degrade image quality and there will be a short compromise zone of aperture size only in which neither of these deleterious effects acts to push image quality below an acceptable level. It is only possible to widen this zone at the big aperture end by pouring money into a big funnel leading to a development department staffed with really, really clever people and sailing right along the bleeding edge of what is possible in glass, electro-mechanics, and processor technology to arrive at the kind of lenses we are now seeing on our 1" sensor cameras.
To widen the zone of usable aperture size at the small end is also possible because diffraction effects can be mathematically predicted and therefore, can also be corrected with complex mathematics - over a small range only.
These warring requirements for image quality maximisation in small sensor cameras, and the inescapable need for lots of money to push the opposing forces apart, are the reasons we are being asked to pay the best part of £1500 for an RX10-III and why we may have to pay £1700 for an RX10-IV.
This page discusses diffraction in a very informative way...
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm
The mathematical tools on the page reveal some fascinating insights - for example: to push the size of the diffraction disc, which is a measure of the fuzziness of the image of a fine point as it appears at the sensor, down to the size of the sensor pixels, the condition required to just produce images with full sensor resolution, needs f/2.0 on an RX100 at the shortest focal length of the lens. At all apertures smaller than that the image is degraded by diffraction. Of course, achieving this performance would also require a lens sufficiently aberration free that it could resolve the sensor pixels at f/2.0. It's difficult to justify big numbers of megapixels on camera with such sensors when this stuff is weighed into the argument.
--
Ed Form
It should also be born in mind that wide aperture lenses have problems with optical aberrations to a greater degree than small aperture designs, a fundamental difficulty that can only be defeated with complex, sometimes very complex, optical designs that cost scads of money to make. This is particularly a problem for wide range zoom lenses where aberration correction has to be maintained over the whole of the range. As a result, the lenses on modern super-bridge cameras are astonishingly complex devices. The lens on the RX10-III, as an example, is not only a seriously complicated physical assembly of glass types and shapes but is made hugely more complicated by the fact that the lens on its own would not be acceptable on any camera. Certain areas of aberration are not corrected in the lens, instead they are corrected with complex processor mathematics - the optical system is actually the lens plus the processor.
Returning to the main subject of image quality: it is a fact that as aperture grows smaller, diffraction increasingly degrades image quality but it is equally a fact that as aperture grows larger, lens aberrations increasingly degrade image quality and there will be a short compromise zone of aperture size only in which neither of these deleterious effects acts to push image quality below an acceptable level. It is only possible to widen this zone at the big aperture end by pouring money into a big funnel leading to a development department staffed with really, really clever people and sailing right along the bleeding edge of what is possible in glass, electro-mechanics, and processor technology to arrive at the kind of lenses we are now seeing on our 1" sensor cameras.
To widen the zone of usable aperture size at the small end is also possible because diffraction effects can be mathematically predicted and therefore, can also be corrected with complex mathematics - over a small range only.
These warring requirements for image quality maximisation in small sensor cameras, and the inescapable need for lots of money to push the opposing forces apart, are the reasons we are being asked to pay the best part of £1500 for an RX10-III and why we may have to pay £1700 for an RX10-IV.
This page discusses diffraction in a very informative way...
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm
The mathematical tools on the page reveal some fascinating insights - for example: to push the size of the diffraction disc, which is a measure of the fuzziness of the image of a fine point as it appears at the sensor, down to the size of the sensor pixels, the condition required to just produce images with full sensor resolution, needs f/2.0 on an RX100 at the shortest focal length of the lens. At all apertures smaller than that the image is degraded by diffraction. Of course, achieving this performance would also require a lens sufficiently aberration free that it could resolve the sensor pixels at f/2.0. It's difficult to justify big numbers of megapixels on camera with such sensors when this stuff is weighed into the argument.
--
Ed Form
Last edited:
