Fuji F50fd

[tongodeon]

A while ago I wrote a glowing review of the Fuji F31fd. It's only 6 megapixels, but the super-sensitive, super-low-noise ISO3200 CCD gives it much better clarity than any higher-resolution camera I've seen. The only real down-sides are the size (a little bulky) and the proprietary xD memory card.

Fuji just announced the F50fd. It's got a 12

Comments

[mskala]

Yes, and even if the object reflected 100% of the light it received, a lot of that (assuming it's not a perfect mirror reflecting the light straight from the bulb to your camera) would be reflected in some other direction, not at you. I think that factor will be even more significant.

(Note I said 2% efficient light bulb and then calculated as 1%; correcting that below.)

Say your object (seen from the 40W bulb) covers 1/100 of a sphere - that'd be pi/25 steradians, and I think a reasonable estimate of how big a typical indoor object might be in relation to its distance to a lamp. I'm imagining let's say a copy of HPatDH sitting on a desk under a 40W lamp. Then it reflects 18% of the light equally in all directions. Then the object is radiating 18% of 0.8W/100 =0.00144W, which is less than the light bulb in my previous comment by a factor of 1736. Multiply the size of the light collector by that, it needs to subtend 5.99*10^-3 steradians, rough estimate (I'm not doing the sphere-cap formula again) 2.5 degrees radius. About the size your fist looks if you hold it at arm's length. That's actually pretty big.

Mind you, that's assuming low light, a darkish object, that it's required to appear #FFFFFF in the image taken at 1/30s exposure, and that the pixels are pure photon counters, as well as numerous approximations along the way.

I've heard it claimed that human eyes are sensitive enough to detect individual photons, but that the noise floor is too high for them to really be usable down to that level.

I think the bottom line is that digicams are probably not quite at the ultimate limits of the physics yet, but we're starting to get into that range and it may be an issue in the future. The astrophotography people know a lot about the quantum limits of CCDs. They're the ones who could really answer this properly.