DSLR v Consumer Image quality

The ONLY thing I need to improve is "shutter lag" (even with the Sony
828,
once aware of its foibles, you can avoid them)

Larry:

You prove two important points, and that is to get to know your camera
and then push it to the max, and to work at taking good photos. Over
the weekend I stood on a train platform with my 5060 and did incredibly
fast firing in the HQ mode at fast oncoming trains. I shut down all
the auto exposure features and bang, bang, bang. I suppose a dslr
would have been faster, but not so much so that it would have made a
difference. And, even though it takes a second or two for my camera to
'warm up' after it decides to rest I've learned over the years that an
occasional tap on the zoom control (which is now done almost
sub-consciously) suffices. As for good photos, I suspect that a very
high percentage of problems with small prosumer cameras can be
attributed to things like miserable knowledge of backlighting, camera
shake, and, sad to say, poor optics.

Incidentally, I think we do need to make some distictions between those
who are first time investors and those who have already invested in
very good prosumers. But by and large I want to use my money to travel,
get extras, upgrade software, etc. When Pentax gets a good dust
cleaning system and goes back to CF cards I may take the plunge (I have
some beautiful top of the line K mount lenses which I can still use on
my SLR), but right now I can think of very few reasons to do so.

The ONLY thing I need to improve is "shutter lag" (even with the Sony
828,
once aware of its foibles, you can avoid them)

Larry:

You prove two important points, and that is to get to know your camera
and then push it to the max, and to work at taking good photos. Over
the weekend I stood on a train platform with my 5060 and did incredibly
fast firing in the HQ mode at fast oncoming trains. I shut down all
the auto exposure features and bang, bang, bang. I suppose a dslr
would have been faster, but not so much so that it would have made a
difference. And, even though it takes a second or two for my camera to
'warm up' after it decides to rest I've learned over the years that an
occasional tap on the zoom control (which is now done almost
sub-consciously) suffices. As for good photos, I suspect that a very
high percentage of problems with small prosumer cameras can be
attributed to things like miserable knowledge of backlighting, camera
shake, and, sad to say, poor optics.

Incidentally, I think we do need to make some distictions between those
who are first time investors and those who have already invested in
very good prosumers. But by and large I want to use my money to travel,
get extras, upgrade software, etc. When Pentax gets a good dust
cleaning system and goes back to CF cards I may take the plunge (I have
some beautiful top of the line K mount lenses which I can still use on
my SLR), but right now I can think of very few reasons to do so.

rafe bustin wrote:

I disagree that you need a print to properly see or
judge a digital image.

Indeed not. To print, you almost always have to compress the gamut
and/or the dynamic rnage.

Andrew.

rafe bustin wrote:

I disagree that you need a print to properly see or
judge a digital image.

Indeed not. To print, you almost always have to compress the gamut
and/or the dynamic rnage.

Andrew.

On Wed, 23 Feb 2005 13:41:27 GMT, Owamanga wrote:


At the very least, you should be using a large format film camera.


I do that as well. It puts these puny
6 Mpixel DSLR images in perspective.


rafe b.
http://www.terrapinphoto.com

lid wrote:

"Roger N. Clark (change username to rnclark)" wrote:


Modern sensors are photon noise limited. The marketing departments
don't even know this, nor hype it. It is a demonstrated fact. See:
http://clarkvision.com/imagedetail/d...el.size.matter
http://clarkvision.com/imagedetail/d...ignal.to.noise


It says "The maximum signal-to-noise is then (52000/(square root
52000) = 228... The only way to improve on the signal-to-noise is to
acquire more photons."

But surely the relevant signal-to-noise ratio is the maximum signal
(for white) divided by the dark noise level (for black). What people
notice is noise in the shadows, whereas noise in the highlights is far
less interesting. There is much to be gained by reducing read noise
even without increasing full well capacity.

What you are saying is the dynamic range of the sensor,
not the signal-to-noise ratio of a single pixel.
They two specifications are different.

Roger

On Wed, 23 Feb 2005 13:41:27 GMT, Owamanga wrote:

On Wed, 23 Feb 2005 08:03:52 -0500, rafe bustin
wrote:

But many where a screen preview can reveal an irrelevant quality
difference because it's one that *won't* show on the print.

That's the point. The destination of any good photo is to be printed,
surely? So, that's *all* that matters.

When you enlarge a 3000x2000 pixel image at 200% on your monitor, you
are comparing it in size to a 60" x 40" print. This is *insane*. So
what if you can see an imperfection at that magnification? most people
are never going to *dream* about ever printing a 35mm image image that
size, so imperfections that only show up now are *irrelevant*.


Print size matters, of course. I have an
Epson 7000 at home, so I can do 24" wide
prints.

I believe you're saying the monitor is
perhaps "overly critical" and there is
much truth to that -- many flaws visible
on the monitor will be invisible in print.


But that does presume a decent monitor that's
properly calibrated. You've done that, right?

Not relevant.


Hogwash. It's entirely relevant. If the
monitor lacks focus or is misadjusted for
gamma, you will lose detail, and will be
surprised by what you see in print. If
you're doing digital darkroom work on a
cheap uncalibrated monitor at 600 x 800,
you're in for a rough ride.


When you introduce a printer -- even a Fuji or
a LightJet -- you introduce another huge set
of variables, issues, and technologies.

So use the same equipment to print the comparisons. That way you keep
it a constant.

Constant, yes. But having the effect of
masking or obscuring and generally degrading
the original image. For example, there are
very few print technologies that can match
the Dmax or gamut of a good monitor, and
inkjet prints are far from being "continuous
tone." The ubiquitous "banding" seen on
many inkjet and dye-sub printers has no
parallel on a monitor.


For me, the prints are the ultimate goal, but
not every image makes it to print. And not
every print makes it past the scrap bin.

If you are binning stuff just because it won't look good at 5ft by 3ft
when viewed from 15inches then you need to get your head tested.


I bin stuff for any number of reasons,
some technical, some not. A large part
of art is knowing what to keep and what
to toss.


rafe b.
http://www.terrapinphoto.com

wrote:

Roger, while I normally agree with your work, I'm not at all sure about
this bit..


The human eye with normal vision has a image
resolution equivalent over 500 megapixels.


Your website seems to argue that you simply multiply the ability of the
eye to resolve detail, by its angle of view.

But hang on a tick!!! The eye is only sharp in a *very* small angle -
I would guess less than a 5 degree cone! It fills in the rest with
memory and 'false' resolution by assumption.

By your argument, one only needs a small patch of high resolution
data in the center inch or so of an 8x10 print (e.g. about 1 inch
in diameter) and the rest can be fuzzy. Obviously the
eye+brain see more than that. Your eye wanders around
even if you are not aware of it. As you view a scene,
you move your eyes around to see all the detail.
This is true whether looking at a photographic print,
or looking at a real scene. Stand outdoors and examine
a real scene and the wealth of detail you can see all around you.
Now produce a photograph with that same detail: the detail
you can see with your eyes in the real world. It can't be
done with a digital camera, 35mm film, or medium format
film. One needs large format film, and 4x5 fine grained
film is about the minimum, and even then it does not have the
detail you experience with one's full field of view.
If you've never seen a big large-format print, at the next
opportunity, do so. It can be like looking
out a window to the real scene.

Roger

In article ,
rafe bustin wrote:
On Wed, 23 Feb 2005 13:41:27 GMT, Owamanga wrote:


At the very least, you should be using a large format film camera.


I do that as well. It puts these puny
6 Mpixel DSLR images in perspective.

I gave it and bought one last weekend. It's a great old thing, MPP VI,
upgraded to VII spec with the rotating international back. Two days ago, I
shut myself in my bathroom, blocked the gap under the door, turned out all
the lights, and loaded my first sheet of film. That was a bit of an
intimidating experience - I knew vaugely what to expect after unsealing the
box, but having to do it all in total darkness by feel alone was kind of
daunting. I'm hoping it'll get easier with experience. I also got a rollfilm
back for 6*7, for when I'm feeling too wimpy to use sheet film.

So far, I have spent hours playing with the camera, but have yet to take a
single exposure. How many digital cameras can you say that about? ;-)

I expect I'll be taking a picture this weekend - probably a macro of a dried
flower using the triple extension. I'm hoping my friendly local lab won't
mind me just dropping the dark slide off with them, otherwise I have no idea
how I'm going to get this processed. I don't especially fancy setting up my
own E6 darkroom.

My wife has decided that we are calling the camera, "Snapper", because
something with so much character needs a name. This is not because we expect
to use it to make expensive snapshots, but because the thing essentially
seems to be a load of springs held in formation by willpower, which leap out
and "snap" at the unwary.

Not entirely sure what I've got myself into here...

On Wed, 23 Feb 2005 07:44:37 -0700, "Roger N. Clark (change username
to rnclark)" wrote:

wrote:

Roger, while I normally agree with your work, I'm not at all sure about
this bit..


The human eye with normal vision has a image
resolution equivalent over 500 megapixels.


Your website seems to argue that you simply multiply the ability of the
eye to resolve detail, by its angle of view.

But hang on a tick!!! The eye is only sharp in a *very* small angle -
I would guess less than a 5 degree cone! It fills in the rest with
memory and 'false' resolution by assumption.

By your argument, one only needs a small patch of high resolution
data in the center inch or so of an 8x10 print (e.g. about 1 inch
in diameter) and the rest can be fuzzy. Obviously the
eye+brain see more than that. Your eye wanders around
even if you are not aware of it. As you view a scene,
you move your eyes around to see all the detail.
This is true whether looking at a photographic print,
or looking at a real scene. Stand outdoors and examine
a real scene and the wealth of detail you can see all around you.
Now produce a photograph with that same detail: the detail
you can see with your eyes in the real world. It can't be
done with a digital camera, 35mm film, or medium format
film. One needs large format film, and 4x5 fine grained
film is about the minimum, and even then it does not have the
detail you experience with one's full field of view.
If you've never seen a big large-format print, at the next
opportunity, do so. It can be like looking
out a window to the real scene.

Except by these standard of arguments, the print will only have about
1/5th of the dynamic range that an eye can handle, plus information
for only one eye (eg, no 3D data). By this argument, it's hardly
convincing - of course in reality our brains ignore all that stuff and
those prints do look amazing.

--
Owamanga!