О сенсорах. На аглицком правда. Ежели кому интересно

The article
http://www.luminous-landscape.com/essays/sensor-design.shtml

leads me to think if, as the author puts forward, 6.8 micron is the optimum size of a photosite and 5 micron is the limit before quality suffers, then the math is:
for aps-c sensors the optimum is about 7.3MP and the limit is 13.5MP
for FF sensors the optimum is about 18.7MP and the limit is 34.6MP

(e.g. FF 36 x 24mm accomodates 5294 x 3529 pixels at 6.8 microns per pixel or 18,682,526 pixels in total)

If you buy the idea, it seems that aps-c sensors are near the limit but FF sensors still have room to develop as the current reduced-frame sensors are at 10MP but FF are only at 17MP. The 5D and 1ds-II are both quite close to the optimum. Of course, you have to buy the author's idea in the first place about the way the optimum size of photosites relates to the wavelength of visible light.

Just for sharing. Absolutely NO intention to start an aps-c vs FF fight whatsoever.


И один из ответов:

The problem with MR's conjecture is that it assumes that current photosites are completely photo receptive (i.e. the entire face of the photosite is used as the collection area). In practice, this isn't close to true.

Fujifilm, for example, makes a marketing point of the shape and size of their photoreceptors being larger than would be the case with square photosites (has to do with the way the power and data lines track around the photo sensitive areas). Whether or not that is a valid claim or not is difficult to verify, as not enough information has been released by any sensor maker to be able to do a side-by-side comparision.

Kodak (and FillFactory) made a big deal of the larger photo receptive area on the sensor that went into the Pro 14n/SLRn. Indeed, it was huge compared to any other CMOS I know of, but it introduced some additional issues that weren't well dealt with, especially in the initial iteration.

My personal assessment is that each individual sensor technology currently being used hasn't changed much in terms of photo sensitivite area, though using other techniques the SN of an individual photosite data has been improved somewhat.

The original Nikon D1 sensor is a good case in point for illustration of how some things have changed and some things haven't. Most people don't realize that the sensor in that camera was essentially a 10mp sensor that was "binned." The D200 apparently has pretty much the same photo sensitive area, but does not have to bin (combine) the photosites in order to achieve the same type of noise production as the original D1 sensor did. Some of that is post processing. Some probably is better integrity in the substrates. Some may be less signal spill between photosites and transfer mechanisms. I don't believe any of it is larger photo sensitive areas (though I'm speculating here--I have no way of assessing that by physical measurement or published specifications).

Thus, I think MR's theory is probably valid for any individual technology going through slow and constant evolution (e.g. Sony's CCDs), but may not be correct for any clear change in technology. I'll cite one possible example: Canon's CMOS uses MOSFET and Nikon's CMOS (LBCAST) uses JFET transisters. What little Nikon published about the change from MOSFET to JFET (and which can also be found in a few academic papers) is that, all else equal, the underlying noise structure of the sensor would be significantly lower. Unfortunately, there's no easy way to compare the two in actual use, as there's much more to a sensor than the underlying transister noise (that pesky photosensitive area being used, for one ;~).

Thus, one might say that for the current Sony CCD technology the limit for APS might be in the 12-15mp range. But I'm a bit uncomfortable saying that's THE limit.
--
Thom Hogan
author, Nikon Field Guide & Nikon Flash Guide
editor, Nikon DSLR Report
author, Complete Guides: D50, D70, D100, D200, D1 series, D2h, D2x, S2 Pro
http://www.bythom.com

http://www.cambridgeincolour.com/tutorials/digital-camera-sensor-size.htm

Тоже местами полезная инфа. Например поглядеть на каком сенсоре начиная с какой диафрагмы начинается влияние дифракции на резкость.