|
On 3/17/2016 8:27 PM, Ctein wrote:
Dear Mike,
I am CCing Moose and DDB in on this conversation, because Moose
and I have been discussing this camera and its performance and
looking at comparison photographs (and it's his fault I just
bought one).
I appreciate being copied, and accept the blame. :-)
And . . .
As mentioned, I don't like the physical changes to the camera
body-- it's too bulgy and angular for my rather delicate and
thinly-padded fingers. It's gonna hurt holding that camera for
long periods of time. I'll put up with it.
Different bodies, hands, fingers . . . I HATED the on/off switch on
the original. I'm comfy with the Mark II body.
One of my semi-irrational reasons for not buying the camera --
having all the controls rearranged between the Mark I and II --
turns out to be unfounded. I can reconfigure the controls so that
they are pretty close match to what I have on the Mark I. Better,
actually, since I have more buttons to play with. That isn't going
to bother me at all it turns out.
I'm still figuring some of it outŠ like what to do with (and how
to program) the flippy switch below and to the right of the
viewfinder.
It's in the Buttons and Dials section, and has 6 options, plus OFF.
See below.
It's nice that there is a dedicated button for HDR.
Which you may recall I jumped through some major programming hoops
to implement on the Mark I. And, happily, it's naturally assigned
to a button that doesn't exist on the Mark I, so it won't get me
confused. The built-in HDR works very well, even with long night
exposures (up to 10-15 seconds, anyway). My only complaint about
it is that I can't customize the interval; in my opinion, a two
stop gap is usually too large, And I don't really need +/- 4 to 6
stops. I wish I could tune it to my preferences. Maybe in a
future firmware update.
As previously mentioned, the image stabilization is amazingly
good. Consistently handholdable with the 45 mm lens down to
between 1/5 and 1/10 of a second. I set the first curtain release
at a 0.2 second delay, which isn't really noticeable to me, and it
works really nicely in combination with the IS.
This is where things get tricky/confusing for me. Oly µ4/3 bodies
from the E-P1 have included an Anti-Shock setting in the Menus, to
wait briefly after the first curtain closes before reopening it to
start the exposure. 1/8 sec. got rid of most shutter shock. By 0.2
sec., I assume you mean the 1/4 sec. setting. Sensible and
understandable, and makes a huge difference in tele and macro. But,
by 1/4 sec. the bird is out of the frame. I don't do a lot of that
sort of shooting, but enough that I don't like much delay.
Oly first introduced an electronic first curtain in the E-M10, then
in the E-M1 via firmware update. The E-M5 II manual says "To prevent
camera shake caused by the small vibrations that occur during
shutter operations, shooting is performed using an electronic
front-curtain shutter." In the camera Menu, however, it says that
is AS mode, EFC is only used at 1/320 sec. and below. That's fine,
as that's the highest speed at which I found shock blurring on the
E-M5, but somewhat contradictory.
Then Silent mode "is performed using electronic shutters for both
the front and rear curtains so that the miniscule camera shaking
caused by shutter movements can be reduced," In their test of the
new 300/4, IR publishes some test shots that show Silent Mode to be
slightly, but significantly, superior to AS Mode @ 1/80 & 1/125
sec. in reducing body movement induced
blur. After one pays for the 300/4, Silent Mode is obviously
needed to get all the resolution one paid for.
First, I wonder if you have checked whether the 1/4 sec. delay you
are using makes any difference on the Mark II. With EFC, I don't see
how it would make a difference. If it's a setting you've carried
over from prior bodies, you may be able to change to zero sec.
delay, which just enables EFC.
Second, both because you are using the camera and mention you may
review it, is the weird interaction between A exposure Mode and
Silent shutter Mode. When I got my first E-M5 II, I cheerily did
what all the above suggests, set it to Silent with zero delay. The
problem of rolling shutter geometric distortion of fast moving
objects is just not an issue with my subjects. Then I found myself
getting some way too long exposures of tele shots. (The updide of
this was discovering that the IBIS is WAY better, really amazing, at
focal lengths and shutter speeds I would otherwise never have
tried.)
What I discovered is that the exposure algorithm for adjusting Auto
ISO in A Mode was way off and way different from the Mark I. In
bright and dim light, they were about the same, but in the in
between area, wildly different. It would hold the ISO down until
shutter speed got very low. For example, a shot that would be 1/120
@ ISO 1600 on the Mark I would be 1/20 @ ISO 200 on the Mark II.
After an exchange with a helpful tech at Oly support, I knew no
more, except that it was all news to them - and they couldn't
duplicate it. Sometime later, for whatever reason, I switched to
Anti-Shock Mode - and the problem was gone.
I let Oly support know about this. was thanked and was told a fix
might or might not show up in a firmware update, but certainly not
if it was due to the hardware. The latest big update has not changed
it. Soooo . . . For someone like me, a dyed in the wool A mode
shooter and long tele lover, it's a problem. (The Panny 100-400
arrives Wednesday. We put back our SoCal wildflower trip a day to
wait for it.)
I've been using AS mode, but with an even longer lens coming, I'd
really like to use Silent Mode. It appears I may be experimenting
with S and M modes (oops, make the M and S Modes). First, though -
AND Here's where the Lever setting comes in - With the Lever in Mode
2, the rear wheel becomes a direct ISO dial. So I can set the
aperture I want and adjust ISO with my thumb to get a usable shutter
speed.
Since I never do video, I assigned the video button to a
customized Hi Resolution mode. You knew I was going to get to
that. Moose and I have been comparing photos and discussing it at
some length. The new firmware speeds things up-- all the exposes
are taken within less than half a second (and Moose said, I think,
that there is a firmware in the works that will cut it down to
1/60 of a second,
Oly said early on that this was in the works, but not, I think, as a
firmware update. I'm guessing different hardware, faster reading
sensor?
Assuming short enough exposure times, of course). The
processing time is only 9 seconds. That is not excessively long to
wait.
Something that Moose and I found out-- color rendition with
bizarre light sources, like Christmas LEDs, is LOTS better in
HiRes mode. Normally, Bayer array cameras go somewhat crazy with
the narrow spectrum LEDs, especially the blue-violet ones, which
exhibit a weird tone inversion-- they end up with dark blue halos
around them files rather than in the rendered light blue glows.
The HiRes mode doesn't do that. It has a few weirdnesses, but
they're much smaller and less annoying.
One thing that turned out not to be true-- there's no difference
in the captured exposure range between normal and HiRes mode:
it's exactly the same. Not sure, Moose, why it looked different in
your quick and dirty photos, but that was clearly an anomaly. My
careful control tests show that.
Really what I expected.
There is one really big difference, though, which is that the
HiRes are essentially noise free. Not that the E-M5 is all that
noisy at ISO 200Š except to the fetishistsŠ It's already very,
very good. In HiRes mode, even that miniscule amount of noise is
entirely gone; the tonality is absolutely smooth with no hint of
variation from pixel to pixel. It's amazing. And it extends all
the way down to the shadows, which never get grainy. If you feel
the need to open up the shadows, they stay nice and creamy. It is
very, very much a large-format look, almost a large-format contact
print look.
At the highest ISO allowed for HiRes, ISO 1600, the noise level is
the same as what you'd see in a normal exposure at ISO 200.
Yes, that part is awesome.
Okay, but what about resolution?! You know what, it's not a big
deal! By which I mean that in a full frame 17 x 22" print from my
P 800 printer, I can just barely see any sharpness difference
between normal and HiRes loads. I really have to put my nose up
to the print to distinguish the two. Oh yes, I can easily see a
big sharpness and resolution difference on the screen at 100%/200%
enlargements. But in a decent sized print? Nuh uh.
Both normal and HiRes photographs benefit from some deconvolution
sharpening, but it works much better on the HiRes photos. Edge
acutance jumps considerably without getting crunchy or, well,
"edgy." It just looks naturally sharper.
Yup, yup.
The effect is a little cruftier on a normal
photograph. It doesn't terribly bother me, but it's the kind of
thing that gives Oren Grad fits. Well, I think he'd really like
the look of the HiRes prints, and I suspect for him there'd be a
huge visual difference between normal and HiRes in 17 x 22" size,
because the HiRes edge detail just rolls off smoothly, like
large-format film with a decent lens. For us mere mortals, though,
it's a subtle thing.
Go to a bigger print, 24 x 32", (or crop significantly) and you
can clearly see the difference between the formats in print
sharpness and resolution. But, short version? The real import of
this mode is not the extra resolution, but the creation of a
large-format freedom from noise and natural-looking smoothness to
edges and tonal transitions.
And there's the fun of real life, vs. advertising and expectations.
The "BIG thing" is the rez, but what really matter more, for most
uses, are color accuracy and noise.
There are two other things that interest me about it. First, we have
lots of lenses that considerably out resolve the sensors.
Second, the whole vast number of serious digital shooters have
calibrated their idea of the detail a sensor can render based on the
limitations of Bayer array sensors. (Except, of course for Ctein,
Foveon shooters and other paragons.) Downsample an HR image file to
the size of the sensor, and it resolves a great deal more detail at
the pixel level than the otherwise identical shot taken a moment
before or after in normal rez mode.
Moose
|
|