>>4026109>Does the reciprocal rule still hold up at higher megapixelsNo
Everything gets more difficult/less forgiving at higher resolution
That "rule" is/was just a very general rule of thumb for getting "sharp enough" shots. It shouldn't be considered some magic threshold under/over which you're guaranteed blurry/sharp shots. It's a sliding scale. As you approach the rule you'll get a larger proportion of blurry shots, and the ones you think are sharp are going to be ever so slightly blurrier than if you'd just raised the SS. This is amplified with higher resolution because what would be <1px of blur (AKA no blur) is now visible because of the smaller pixels.
Lens diffraction has the same problem. For me on 24mp, f/11 is just as sharp as f/8, f/16 is noticeably but not terribly less sharp and f/22 very noticably less sharp. If I had a 50 or 100mp sensor then the described sharpness levels would happen at 8/11/16.
This need to stay at larger apertures causes problems. Lenses have to be built to be sharper at larger apertures which is more expensive and results in bigger, heavier lenses. It also reduces the DoF which is perversely itself a bigger problem with more mp. The effective DoF (that is, the bits that are physically OOF but have a circle smaller than 1 sensor pixel) shrinks because the sensor pixels get smaller, causing the circle of confusion (basically tiny bokeh balls) to spill out and look OOF.
Having said all that, you're really no worse off having a 100mp "slightly unsharp" photo than a 24mp "sharp enough" photo because you can just scale it down and get what you would have gotten with 24.
>>4026122Declaring one resolution the "perfect" resolution is a fallacy because you can always just scale down to hide the above issues. This applies to noise as well. There's no more noise on a scaled down 100mp image with its tiny pixels than a 100% 24mp image. I'm sure 10 years ago people were calling 16mp "perfect", and 10 years from now it will be 48.
