Sometimes a naive question on birdforum can provoke an interesting discussion and I stumbled on a recent one on "sharpness" rather late as the OP is on my ignore list. There were nevertheless some misconceptions on that thread and it has since gone OT, so I thought I would start anew.
The potential resolution of an objective lens is defined by its diameter, as light gets diffracted at its edges.
William Dawes' criterion was 4,56/objective dia. in inches or 116/objective diameter in mm. Accordingly, an excellent 116 mm telescope would be able to resolve 1 arcsecond (1") and a 58 mm telescope would be able to resolve 2".
Dawes' observations were based on the ability to separate double stars with a minimum loss of brightness of 5% at their merger but today we can use a 1951 USAF chart with ever decreasing separation of line pairs (increasing frequency of line pairs/mm) to measure resolution. With increasing frequency the contrast diminishes until the lines merge into a grey blur. At higher spacings (lower frequency) the contrast will always be higher than at the resolution limit and it is nonsense to suggest that a lens with high resolution might show inferior contrast at higher spacings than another lens with lower resolution.
There is often a tendency to make comparisons between binocular or telescope lenses and camera objectives. The latter are focal devices and project an image on to a film or sensor. Binoculars are afocal devices and the image is focussed on the observer's retina. To achieve this the focal planes of the objective lens and eyepiece are brought to, or near to coincidence, so that the observed image for normal vision appears at infinity or, in the case of corrected vision, at the eye testing distance of 20 ft/6 m.
Binoculars seldom have a FoV exceeding 9° but even a full format 50 mm camera lens has a 46° FoV and the requirement is not only to provide acceptable centre sharpness but also acceptable edge sharpness. If this were achieved at f/5,6 then the aperture would be a mere 9 mm, offering a maximum potential resolution of 13", a miserable value for a binocular.
"Binastro" on this forum has often stated that adding a prism and eyepiece (lens2scope) to even the best telephoto camera lens makes a very poor scope.
I measured the resolution of my little 65 mm Swarovski scope at 1,78", which is diffraction limited. To overcome the limitations of my own eyes I had to boost the magnification to 130x using a 3,5 mm astronomical eyepiece and was able to resolve 5 line pairs/mm or a line width of 1/10th of a mm at a distance of 23 m! I doubt that there is a commercially available camera lens capable of that, maybe though on a spy satellite.
MTF measurements on camera lenses show loss of contrast from the field centre to the field edge of specific frequencies of sagittal and tangential line pairs (10, 20, 30 lp/mm) at the focal plane (sensor) and do so for various apertures. One could not conduct these measurements on a binocular or scope because they are afocal instruments and edge performance could not be boosted due to the limitations of the booster or interchangeable eyepiece, which would itself introduce field curvature or astigmatism. Edge performance of binoculars and scopes is a matter of personal preference and will have to remain subjective.
John
PS:- Low frequency contrast can, of course, be bosted by digital processing and can occasionally be seen in the birdforum gallery when it has been used to excess. It can also be achieved in analogue photography by using a diluted negative developer so that boundaries are starved of developer and appear sharper.
The potential resolution of an objective lens is defined by its diameter, as light gets diffracted at its edges.
William Dawes' criterion was 4,56/objective dia. in inches or 116/objective diameter in mm. Accordingly, an excellent 116 mm telescope would be able to resolve 1 arcsecond (1") and a 58 mm telescope would be able to resolve 2".
Dawes' observations were based on the ability to separate double stars with a minimum loss of brightness of 5% at their merger but today we can use a 1951 USAF chart with ever decreasing separation of line pairs (increasing frequency of line pairs/mm) to measure resolution. With increasing frequency the contrast diminishes until the lines merge into a grey blur. At higher spacings (lower frequency) the contrast will always be higher than at the resolution limit and it is nonsense to suggest that a lens with high resolution might show inferior contrast at higher spacings than another lens with lower resolution.
There is often a tendency to make comparisons between binocular or telescope lenses and camera objectives. The latter are focal devices and project an image on to a film or sensor. Binoculars are afocal devices and the image is focussed on the observer's retina. To achieve this the focal planes of the objective lens and eyepiece are brought to, or near to coincidence, so that the observed image for normal vision appears at infinity or, in the case of corrected vision, at the eye testing distance of 20 ft/6 m.
Binoculars seldom have a FoV exceeding 9° but even a full format 50 mm camera lens has a 46° FoV and the requirement is not only to provide acceptable centre sharpness but also acceptable edge sharpness. If this were achieved at f/5,6 then the aperture would be a mere 9 mm, offering a maximum potential resolution of 13", a miserable value for a binocular.
"Binastro" on this forum has often stated that adding a prism and eyepiece (lens2scope) to even the best telephoto camera lens makes a very poor scope.
I measured the resolution of my little 65 mm Swarovski scope at 1,78", which is diffraction limited. To overcome the limitations of my own eyes I had to boost the magnification to 130x using a 3,5 mm astronomical eyepiece and was able to resolve 5 line pairs/mm or a line width of 1/10th of a mm at a distance of 23 m! I doubt that there is a commercially available camera lens capable of that, maybe though on a spy satellite.
MTF measurements on camera lenses show loss of contrast from the field centre to the field edge of specific frequencies of sagittal and tangential line pairs (10, 20, 30 lp/mm) at the focal plane (sensor) and do so for various apertures. One could not conduct these measurements on a binocular or scope because they are afocal instruments and edge performance could not be boosted due to the limitations of the booster or interchangeable eyepiece, which would itself introduce field curvature or astigmatism. Edge performance of binoculars and scopes is a matter of personal preference and will have to remain subjective.
John
PS:- Low frequency contrast can, of course, be bosted by digital processing and can occasionally be seen in the birdforum gallery when it has been used to excess. It can also be achieved in analogue photography by using a diluted negative developer so that boundaries are starved of developer and appear sharper.
Last edited: