> Oly and Panny's latest, the Pen-F and GX8, are 20 MP, and the Pen-F HR mode
> is proportionally larger. This amounts to nothing, about a 10% nominal
> increase in linear resolution. Differences in sensor system, Bayer
> interpretation and/or AA filter spec. can make as much or more visible
> change than this with identical pixel counts. My guess is that this will be
> the sensor for the E-M1 II. But even if they go to 24 MP (but see below), it
> wouldn't amount to much, a 22-23% increase in nominal rez, visible at 100%,
> but not much practical change.
For me, real-world practicality says that 18-24MP is the sweet spot.
16MP is definitely on the short end of it. Beyond 24 MP, it buys you
additional flexibility in cropping and processing. Too few pixels and
you have to tread carefully with cranking the volume to 11. And with a
camera like the original E-1, you NEVER cropped other than for 8x10 or
5x7 prints.
While I understand Chuck's hesitation for higher density sensors, I'm
perfectly fine with it as long as the gains are real and not a matter
of sales pitch.
While I hate referencing ancient cameras, the lessons learned are
still applicable and quite illustrative as they represent three very
distinct sensor designs and philosophies. I will make a comparison
between three high-quality, but ancient cameras in my possession.
Olympus E-1, 5MP
Olympus E-3, 10MP
Panasonic L1, 7.5MP
>From a dynamic range perspective, there is no contest. The E-1 rules
the day, closely followed by the L1. From a noise perspective at
base-ISO, the L1 blows the other two cameras out of the water. From a
noise perspective at boosted ISOs, the E-1 is glob central, but it has
a "natural" look to it. The E-3 is unusable because of the banding,
and the L1 is great up to a point where it falls over the edge in a
rather abrupt and obscene way.
>From a sharpness perspective, this is where things get interesting.
The E-1 is ruthless when it comes to resolution. It neither has enough
pixels, nor is the AA filter very friendly. But worse of all, is the
E-1's four-pixel mix. That is what really cuts down on resolution. The
L1, at only 7.5MP is able to resolve anything the 10MP E-3 can do as
the AA filter is almost non-existent and the images respond extremely
well to pixel-level sharpening. (expect for the dot artifacts in ACR).
Like the E-1, the E-3 has a stronger AA filter and really needs the
4-pixel merge in order to keep colors from flying the coop.
Doing downward resizing works great on the E-1 and L1, but not so much
on the E-3. The E-1 needs it for sharpness, but it also makes the
exceptional tonal separation even better. The L1 is still the best
"landscape" camera of the bunch, and pixel reduction wakes those files
up like nobody's business. The E-3? No, not so much. I haven't figured
out how to get those files to shrink without artifacts.
What about the modern cameras that you all have? Here is the bugaboo.
I see certain artifacts and in-camera processing that makes the images
clean, sharp, etc., but at the expense of tonal separation and
pixel-level goopyness. The pixels look like plastic. Is plastic a bad
thing? ABSOLUTELY!!! But only if you don't have enough pixels. When
you get enough pixels, all is forgiven. The Nikon D810 is a good
example of this. When you look at the entire picture, it's lovely. But
when you look at the files at 100%, they look like cellphone pictures.
The Sony A7___ cameras are even worse when you dive down to pixel
level.
So, if you have enough "plastic" pixels, they become "organic".
"Organic" is good. Bring on the pixels, I say.
AG Schnozz
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|