Here is what I expect from what I've learned about it so far:
The Wide DR mode reads half of the pixels early (after only 1/8 of
the time). This makes them 3 stops less sensitive, giving three stops
extra highlight headroom. But since 44% of the photons are wasted
this way, the cost is more photon noise.
But on the other hand since those pixels are going to be the ones to
capture burn, you have a very strong signal, so noise is far less of
an issue than a normal exposure. I consider this pretty much a non
issue.
What he's saying is that, if you shut half the pixels off early, you
lose signal in the shadow areas at the expense of extending the
highlight range.
You don't, that's why you get 1/2 the photosites exposed regularly/more to capture the shadow detail. The tradeoff in this sensor is you shoot at 6 megapixels and not 12, in that mode.
It's still a good idea, though, since you're gaining 3 stops at the
highlight range and losing -log_2 (.44) in the shadows.
You don't lose the shadows in this sensor.
The other two modes won't waste any photons, but the gain from
binning is yet to be proven. It could be anything from zero to a
factor sqrt(2) gain depending on the particular implementation.
Nothing to threaten the low light capability of the LX3.
How does binning ever gain you anything at high ISO? The fractional
shot noise in a given region of the image from gathering N photons is
still sqrt(N), no matter how you slice that region up into pixels.
You get two photosites to "vote" on the color, you get a better average, more surface area aimed at the "same position."
I disagree. Fuji is already really good to begin with, with ISO/noise
management.
They have good NR algorithms that preserve the crispness of edges,
but no NR algorithm no matter how good can recover detail below the
noise floor.
That's true but that wasn't my point. My point is that Fuji has better than usual ISO at the hardware sensor level, not just the software post recovery.
The real obstacle here is Lens speed, but it could be that the ISO
gain over the LX3 offsets this to a point of cancel it completely (or
maybe even give 1/3rd to 1 stop of an advantage anyway, but Ill play
conservative and say cancel out).
Its sensor has to be three times better to make that ground up. Can
Fuji's magic sensor somehow collect three times as many photons as
the LX3's? Probably not. Their tricks to improve dynamic range have
real potential, but I doubt they'll improve high ISO.
Considering that previous Fuji cameras have been doing pretty good, it is certainly possible. And you don't need 3 times. I don't know where you go that.
In theory you need 2 times (at telephoto, which is the worst case, it's from F2.8 on the LX3 to 5.3 on the Fuji). That's about 1 stop + 2/3rds EV's. That's less than two stops. Count about a 1/3rd under rated ISO (note: assuming the Fuji ISO rating is spot on) and now you have 1.5 times better to catch up, not three.
The DR is untouched. This camera most likely has more DR than a Nikon
D3, judging by what I know Fuji did with the SR sensor of the
F700/F710 back in time.
I find that hard to believe. At best the trick mentioned above (turn
off some pixels 1/8 of the way through the exposure) expands DR by 2
1/3 stops or so (see math above); there's no way that can make up for
the absurd amount that you can pull out of the shadows in a
full-frame DSLR.
Check out my posted shot. If the D3 can recover that.. btw, that's of a camera that is 5 years old!
Interesting innovation anyway, could be very good for high DR.
It is.
I imagine it will be great for high DR but significantly worse than
the LX3 for low-light. No amount of electronic wizardry will make
Planck's constant less.
You can say that but the thing is the F30 already did better at the same ISO than the LX3. The reason the LX3 catches up is mainly the lens speed. The EXR is supposed to be better than F30 so...
--
Raist3d (Photographer & Tools/Systems/Gui Games Developer)
Andreas Feininger (1906-1999) 'Photographers — idiots, of which there are
so many — say, “Oh, if only I had a Nikon or a Leica, I could make great
photographs.” That’s the dumbest thing I ever heard in my life. It’s
nothing but a matter of seeing, and thinking, and interest. That’s what
makes a good photograph.'
