"Less Megapixels = Better Color"...

stevo23 wrote:

57even wrote:

stevo23 wrote:

57even wrote:

stevo23 wrote:

57even wrote:

stevo23 wrote:

57even wrote:

tokumeino wrote:

57even wrote:

[...]

If you downsize the files to the same size (either in pixels or print size) the noise level will be more or less identical. It doesn't matter how big the pixels are, only how much light you collect over the whole sensor, or per sq m of the final image.

More small pixels collect the same amount of light as fewer large ones.

Bit-depth is only relevant to DR.

Just look at the SNR data on DXO.

Not beeing an expert, this makes sense to me, especially now that the circuitery between photosites doesn't eat as much sensor space as before, and now that there are microlenses to collect light all over the surface.

I can imagine that with a high density, more photosites will provide random values, but in that case, more accurate photosites will be in the closest neighborhood to provide an adequate value, provided the demosaicing/photosite-level-NR algorithm is not stupid and can more or less sort between random and accurate photosites.

The real problem when photosites get very small (ie Phone cameras) is that the photodiodes have to be shallower. Light absorption is highest at the surface, but longer wavelengths penetrate further. When you get down to depths like 2microns, you lose some of the red signal so colour accuracy is affected.

But visible light is in the sub-micron / nanometer range. Why wouldn't red be able to penetrate?

It does - it goes right through the active region of the photodiode without being detected. That's the problem.

I'm kind of missing the point then. How is shallower an issue?

Because if a large proportion of longer wavelength red photons are not detected, you get a less accurate red response.

Yes, that part is easy. But if the wavelength is considerably smaller than the pixel, why is it that smaller pixels miss some of the red? None of the visible spectrum is nea near 2 microns.

Actually red wavelengths are about .75 microns and some phone camera pixels are as small as 1.5 microns across. But its not just width, it is physical depth. Red photons can travel up to 20microns into the silicon surface before interacting with a silicon atom.

If the depth of the photodiode is only 2microns, quite a few will not be recorded.

So you're saying that they need to travel for longer before they will interact? How does one ever get any decent red response if it's up to 20 microns? I'm still clueless here. Sorry - I'm trying to understand why the wavelength of red is a problem, not being intentionally obtuse.

Absorption starts at the surface, but the probability of red interaction is lower.

So lets say that at a wavelength of 700nm, half are absorbed within 1micron, half of the rest within 2microns, half of the rest within 3microns and so on.

The absorption rate by depth is:

1 - 50%

2 - 75%

3 - 87.5%

4 - 93.75%

5 - 96.875%

And so on.

These are not correct numbers, but you get the idea.

By comparison, 99.99999% of blue photons are absorbed within about 0.2 microns.

A photosite with a depth of 4 microns absorbs nearly 94% of the red photons, but at 2 microns it's only 75%. That's enough to get a decent signal, but it's significantly lower than at 4 microns, where its nearly all of them.

Does that help?