5 ms·

If you zoom out and crop, you still lose the same angular resolution to diffraction as if you had just stopped down at the longer focal length. Object space len

by CarVac 8y ago

If you zoom out and crop, you still lose the same angular resolution to diffraction as if you had just stopped down at the longer focal length. Object space lens resolution is diffraction limited by the size of the entrance pupil, in exact inverse proportion to the depth of field. Plus when you zoom out and crop you waste sensor resolution and photon gathering power... Plus, even an expensive camera suffers at f/22. The only thing you can do is use tilt shift and an actual larger aperture.

BeetleB 8y ago
>If you zoom out and crop, you still lose the same angular resolution to diffraction as if you had just stopped down at the longer focal length. Object space lens resolution is diffraction limited by the size of the entrance pupil, in exact inverse proportion to the depth of field. Interesting - any real world tests to demonstrate this? When it comes to DOF, I'm not interested in logic or theory. I've seen too many flamewars that involve only theory and no actual field tests. >Plus when you zoom out and crop you waste sensor resolution and photon gathering power... I started digital photography on 2 MP. Then moved up to 8 MP. Then to 16 MP. From my experience, the difference between 8 and 16 MP is fairly small. The only benefit for me to go to higher resolution at this stage is the ability to crop and still get high resolution. Without cropping, I don't benefit from extra megapixels. I'll concede the light per pixel point, though.
- CarVac 8y ago
  I just tested 100mm at f/32 against 50mm at f/16, which have the same entrance pupil size, and when cropped to the same part of the subject, the longer focal length simply wipes the floor with the shorter focal length. https://i.imgur.com/GtljNmg.png https://i.imgur.com/GtljNmg.png top is 50mm, bottom is 100mm Why? The antialiasing filter and demosaicing destroy detail at a fixed scale relative to the captured image.
  - BeetleB 8y ago
    Wow - Thanks for actually making the effort. A few quibbles: 1. The goal is to maintain the DOF distance. For 100mm at f/32, if you go to 50mm, you need f/8. 2. The assumption is that the lens is as sharp at 100mm as it is at 50mm (on some lenses this is a terrible assumption). I tried a similar experiment comparing 25mm with 50mm (f/5.6 vs f/22). You can see the results at https://imgur.com/a/ik6qe5K https://imgur.com/a/ik6qe5K To me, they are almost identical - with 25mm seeming a tad bit sharper. In a way, though, you're right - the difference is so small that I might as well shoot at f/22.
    - CarVac 8y ago
      The goal is to maintain DOF and field of view; if you crop to the same field of view, and the entrance pupils are the same size, then you have the same DOF. 100mm at f/32 is the same as 50mm at f/8 only if you don't crop 50mm by 2x. These are both excellent lenses, and at f/32 and f/16, optical quality basically ceases to matter.

by CarVac 8y ago

BeetleB 8y ago

>If you zoom out and crop, you still lose the same angular resolution to diffraction as if you had just stopped down at the longer focal length. Object space lens resolution is diffraction limited by the size of the entrance pupil, in exact inverse proportion to the depth of field. Interesting - any real world tests to demonstrate this? When it comes to DOF, I'm not interested in logic or theory. I've seen too many flamewars that involve only theory and no actual field tests. >Plus when you zoom out and crop you waste sensor resolution and photon gathering power... I started digital photography on 2 MP. Then moved up to 8 MP. Then to 16 MP. From my experience, the difference between 8 and 16 MP is fairly small. The only benefit for me to go to higher resolution at this stage is the ability to crop and still get high resolution. Without cropping, I don't benefit from extra megapixels. I'll concede the light per pixel point, though.

CarVac 8y ago
I just tested 100mm at f/32 against 50mm at f/16, which have the same entrance pupil size, and when cropped to the same part of the subject, the longer focal length simply wipes the floor with the shorter focal length. https://i.imgur.com/GtljNmg.png https://i.imgur.com/GtljNmg.png top is 50mm, bottom is 100mm Why? The antialiasing filter and demosaicing destroy detail at a fixed scale relative to the captured image.
- BeetleB 8y ago
  Wow - Thanks for actually making the effort. A few quibbles: 1. The goal is to maintain the DOF distance. For 100mm at f/32, if you go to 50mm, you need f/8. 2. The assumption is that the lens is as sharp at 100mm as it is at 50mm (on some lenses this is a terrible assumption). I tried a similar experiment comparing 25mm with 50mm (f/5.6 vs f/22). You can see the results at https://imgur.com/a/ik6qe5K https://imgur.com/a/ik6qe5K To me, they are almost identical - with 25mm seeming a tad bit sharper. In a way, though, you're right - the difference is so small that I might as well shoot at f/22.
  - CarVac 8y ago
    The goal is to maintain DOF and field of view; if you crop to the same field of view, and the entrance pupils are the same size, then you have the same DOF. 100mm at f/32 is the same as 50mm at f/8 only if you don't crop 50mm by 2x. These are both excellent lenses, and at f/32 and f/16, optical quality basically ceases to matter.