3 ms·

What’s the size of a photon? You seem to be suggesting that the limiting factor is the physical size of the sensor, which is totally absurd. Photography optics

What’s the size of a photon? You seem to be suggesting that the limiting factor is the physical size of the sensor, which is totally absurd. Photography optics are not even close to approaching the limits of physics.

nayuki 3y ago
Wrong. Silicon image sensors are benchmarked by their "quantum efficiency" - the number of electrons generated by a photon. That number is something like 0.5, which isn't far from 1.0. There is very finite room for improvement in sensors. https://en.wikipedia.org/wiki/Quantum_efficiency https://en.wikipedia.org/wiki/Quantum_efficiency Now, the following point is not well-known, but the most important thing that determines image quality is simply the frontmost piece of glass in a camera system. Not the sensor size, not the ISO. The front glass determines how many photons you receive per second, how much the depth of field is, and how noisy the image is. For example, shoot with a certain lens on a full-frame sensor. Now take that lens, put a "speed booster" behind it, and shoot with a crop sensor. Both images will come out essentially the same in terms of field of view, depth of field, and noise. Again, the front glass is what matters, and everything behind it in the optical system are just technical details. Also as a reminder, shot noise is proportional to 1/sqrt(n), where is n is the number of photons. There are a very finite number of photos you can collect per pixel per aperture per second in real-world shooting situations, and that number is something like 10000. https://en.wikipedia.org/wiki/Shot_noise https://en.wikipedia.org/wiki/Shot_noise
- yunohn 3y ago
  > "quantum efficiency" - the number of electrons generated by a photon. That number is something like 0.5, which isn't far from 1.0. QE is a ratio as described by your Wikipedia link. There is a massive difference between a ratio of 0.5 and 1.0!
  - nayuki 3y ago
    0.5 to 1.0 is just one F-stop in photography. If you make your sensor 2× better, that's not going to overcome a 100× reduction in front glass area when moving from a DSLR to a phone.
    - yunohn 3y ago
      No, I’m saying you have completely misunderstood QE. 1 is the upper limit - 100% efficiency. You can’t just say 50% is very close to 100%!
throw0101b 3y ago
> Photography optics are not even close to approaching the limits of physics. The relationship between aperture sizes, pixel sizes, and diffraction sizes certainly do effect the quality of images: * https://www.cambridgeincolour.com/tutorials/diffraction-photography.htm https://www.cambridgeincolour.com/tutorials/diffraction-phot... * https://fstoppers.com/science/closer-look-lens-diffraction-566322 https://fstoppers.com/science/closer-look-lens-diffraction-5... * https://www.edmundoptics.ca/knowledge-center/application-notes/imaging/limitations-on-resolution-and-contrast-the-airy-disk/ https://www.edmundoptics.ca/knowledge-center/application-not... * https://en.wikipedia.org/wiki/Airy_disk https://en.wikipedia.org/wiki/Airy_disk

nayuki 3y ago

Wrong. Silicon image sensors are benchmarked by their "quantum efficiency" - the number of electrons generated by a photon. That number is something like 0.5, which isn't far from 1.0. There is very finite room for improvement in sensors. https://en.wikipedia.org/wiki/Quantum_efficiency https://en.wikipedia.org/wiki/Quantum_efficiency Now, the following point is not well-known, but the most important thing that determines image quality is simply the frontmost piece of glass in a camera system. Not the sensor size, not the ISO. The front glass determines how many photons you receive per second, how much the depth of field is, and how noisy the image is. For example, shoot with a certain lens on a full-frame sensor. Now take that lens, put a "speed booster" behind it, and shoot with a crop sensor. Both images will come out essentially the same in terms of field of view, depth of field, and noise. Again, the front glass is what matters, and everything behind it in the optical system are just technical details. Also as a reminder, shot noise is proportional to 1/sqrt(n), where is n is the number of photons. There are a very finite number of photos you can collect per pixel per aperture per second in real-world shooting situations, and that number is something like 10000. https://en.wikipedia.org/wiki/Shot_noise https://en.wikipedia.org/wiki/Shot_noise

yunohn 3y ago
> "quantum efficiency" - the number of electrons generated by a photon. That number is something like 0.5, which isn't far from 1.0. QE is a ratio as described by your Wikipedia link. There is a massive difference between a ratio of 0.5 and 1.0!
- nayuki 3y ago
  0.5 to 1.0 is just one F-stop in photography. If you make your sensor 2× better, that's not going to overcome a 100× reduction in front glass area when moving from a DSLR to a phone.
  - yunohn 3y ago
    No, I’m saying you have completely misunderstood QE. 1 is the upper limit - 100% efficiency. You can’t just say 50% is very close to 100%!

throw0101b 3y ago

> Photography optics are not even close to approaching the limits of physics. The relationship between aperture sizes, pixel sizes, and diffraction sizes certainly do effect the quality of images: * https://www.cambridgeincolour.com/tutorials/diffraction-photography.htm https://www.cambridgeincolour.com/tutorials/diffraction-phot... * https://fstoppers.com/science/closer-look-lens-diffraction-566322 https://fstoppers.com/science/closer-look-lens-diffraction-5... * https://www.edmundoptics.ca/knowledge-center/application-notes/imaging/limitations-on-resolution-and-contrast-the-airy-disk/ https://www.edmundoptics.ca/knowledge-center/application-not... * https://en.wikipedia.org/wiki/Airy_disk https://en.wikipedia.org/wiki/Airy_disk