Pixel Peeper Anomalies - They're Totally Missing the Big Picture

On Thu, 27 Aug 2009 00:44:30 +0200, Wolfgang Weisselberg
wrote:

Brad Sanborne wrote:

room. However, on the downside, I found that as the camera locks in on
auto-focus, it *very* *slightly* defocuses the camera from its original
decision as it locks in the pre-determined focus point.

That may be an attempt to better serve all wavelengths rather
then using just the best setting for the wavelengths used for AF.

How big is the airy disk at maximum tele on your camera compared
to the pixel size anyway?

-Wolfgang

Holy **** are you ever a moron. Contrast detection focusing uses the same
wavelengths as imaged by the sensor. All visible wavelengths.

Go take another hit of LSD or whatever it is you're on. Maybe your keyboard
will turn into a snake next time.

eNo writes:
I totally agree with you that the noise (and sharpness, btw) thing is
over-done. Then again, when I try to get one of my high resolution
(10-12mp) images accepted at one of the stock agencies, and 9/10 times
the rejection points to noise and/or sharpness issues. And the only
way they can see this is in 100% pixel-peeping mode.

And what's with Leica not using a sensor AA filter? It looks like
Leica's software "solution" for the resulting aliasing doesn't actually
work (not very surprising, since the necessary information is _gone_ by
the time the software gets the image).

It seems a horrible tradeoff -- throwing out consistent image quality to
gain a slight bit of extra sharpness?! -- and gives the impression that
Leica doesn't actually understand digital image processing very well...

[Which seems an awful thing to say about such a revered company, but
that's the impression....]

-Miles

--
Un-American, adj. Wicked, intolerable, heathenish.

Brad Sanborne wrote:
On Thu, 27 Aug 2009 00:44:30 +0200, Wolfgang Weisselberg
wrote:

Brad Sanborne wrote:

room. However, on the downside, I found that as the camera locks in on
auto-focus, it *very* *slightly* defocuses the camera from its original
decision as it locks in the pre-determined focus point.

That may be an attempt to better serve all wavelengths rather
then using just the best setting for the wavelengths used for AF.


Holy **** are you ever a moron.

Do you need anger management, or do you just react that way when
you run out of arguments?

Contrast detection focusing uses the same
wavelengths as imaged by the sensor. All visible wavelengths.

Your sensor doesn't record all visible wavelengths equally ---
the filters are not flat passbands, for one. Nor does it record
all visible wavelengths at the same spatial resolution --- ever
heard of "Bayer"? --- and one would assume that the highest
usable resolution, the channel with the least noise and the
channel providing the best luminance information --- namely the
green channel --- would be preferably used for AF tasks.

So it's *obviously* *possible* for the sensor to focus using mainly
(or even only) the green pixels, which blocks most of red and blue,
hence a correction may be indictated.

To paraphrase you: "[swear word] you [indication of no brain
activity]". Or in civilized words: "Wouldn't you agree that that
was at least possible?"

Go take another hit of LSD or whatever it is you're on. Maybe your keyboard
will turn into a snake next time.

Talking about Brad's drug days, are we, Braddy boy?

How big is the airy disk at maximum tele on your camera compared
to the pixel size anyway?

No answer? Strange, I *thought* he might know ...

-Wolfgang

Brad Sanborne wrote:
On Thu, 27 Aug 2009 00:40:50 +0200, Wolfgang Weisselberg
Brad Sanborne wrote:
On Sat, 22 Aug 2009 04:44:31 -0500, Brad Sanborne
wrote:

Do the math on how much magnification you are
seeing by viewing a 3648 pixel-width image from a 10-megapixel camera on an
average 96-dpi LCD monitor at 1:1 resolution. That's like looking at a
negative with a 38x-power dissecting microscope.

Correction.

Let's for the sake of argument take a 10-megapixel P&S camera at 3648
pixels wide and a sensor width of 5.75mm. That's 634.5 pixels per mm.
That's 16,116 pixels per inch. On a 96-dpi monitor viewed at 1:1 that's
like looking at a negative with a 168x microscope.

Let's for the sake of the argument not forget that the standard
size one looks at through the loupe is not a 4.31x5.75mm,
but a 24x36mm frame (for common compact P&S cameras), if not
larger (for less P&Ssy cameras). So it's 101.3 pixels per mm
equivalent and thus 2574 pixels per inch equivalent. On a
96-ppi monitor: 27x equivalent.

Or how about a 15.1 megapixel dSLR at 4752 pixels wide and a sensor width
of 22.3mm. That's 213 pixels per mm. That's 5,410 pixels per inch. On a
96-dpi monitor viewed at 1:1 that's like looking at a negative with a 56x
microscope.

Nope: 132 p/mm = 3.353 p/in = 35x --- assuming a 35mm would
have held 15MPix of data and little enough grain for a matching
enlargement.


Now why on earth did you do all that math over when the same calculations
were already done in the last paragraph?

Because your's were wrong.
29x != 27x (actually 26.8111...x) and 36x != 35x (actually 34.952x).

Only you did them in error.

I did? Really?

The width of a 35mm film frame is 36mm.

That must be the reason why I wrote "24x36mm frame".
That must be the reason why I calculated
3648 p / 36 mm ~= 101.3 p/mm
^^^^^
4752 p / 36 mm = 132 p/mm
.. ^^^^^

Oh, did you think "assuming a 35mm [negative] would have held
15MPix of data [...]" had to be "assuming a 35mm [width] would
have held 15MPix [...]".
Why on earth would one think a one-dimensional *length* could
hold 15MPix?
Why on earth didn't even check the math?
How on earth did you come up with the wrong answers --- and
differently wrong answers for each calculation?

Do you always like making a fool of yourself so publicly?

Yes, I love doing that, especially as it turns out you're
wrong and I am right.

Maybe you just have an OCD involving mathematical masturbation.

Maybe you just don't grasp math.

On Sat, 22 Aug 2009 05:13:32 -0500, Brad Sanborne wrote:

Even if we take a 36mm width as a standard 35mm-film frame for a virtual
equivalent negative-size for both, then the P&S image is being viewed with
a 29x magnifier and the dSLR image is being viewed with a 36x magnifier
when viewed at 1:1 on a 96-dpi monitor.

Yep, wrong as it comes.

-Wolfgang

Miles Bader wrote:

And what's with Leica not using a sensor AA filter?

It seems a horrible tradeoff -- throwing out consistent image quality to
gain a slight bit of extra sharpness?! -- and gives the impression that
Leica doesn't actually understand digital image processing very well...

Maybe they think their lenses are so unsharp that no AA blur
filter is needed anyway. ;-
Or maybe sharp-but-imaginary is better than slightly-less-sharp-
but-truthful.

-Wolfgang

Alan Browne wrote:
Brad Sanborne wrote:
I started to wonder how all this brouhaha over sensor-noise and
resolution
came to be the determining factor in quality photography. Then it
dawned on
me.

Before digital cameras only people working in their own darkrooms studied
their negatives and slides with a good powered loupe. Usually only 5x,
8x,
or at the most 10x power. In fact I have my old 8x loupe sitting
beside me
right here, a little desktop reminder of my darkroom days. The average
photographer and snapshooter, of which there are millions today, used
to be
happy with recovering their pack of prints or 8x10s from the local lab.
crud snipped

They have totally missed the big picture .... and probably always will.

And you're missing a much more important point.

High quality gear is expected to provide high quality results. This was
true in 1900, 1925, 1950, 1975, 2000 and now. However, dollar for
dollar we are getting ever increasing quality and capability not to
mention immense time/cost savings from avoiding film and unneeded prints.

This means that the average snap shooter is getting commendable
technical (and often aesthetic) results with far less than would have
been neccessary in the past. Further Mr. Snappy takes more photos (no
cost/convenience issues) and gets better at it and gets more useful
results more often. The future bodes well for all.



I have to disagree here. Being able to "take more photos" doesn't =
higher quality. It makes people lazy at composition/metering/focusing. I
highly doubt digital increases the amout of keepers. I'm TOTALLY
including myself and is one reason I have gone back to shooting
landscapes with my 4X5. With film I got at least a 50% "hit" ratio, with
digital I probably throw out 90% of that I shoot.

Also the total automation make MOST new photographers clueless about
the basics. I wonder how many have a clue what "hyperfocal distance"
even means?

As far as the OP, pixel peeping -is- just silly when you throw out color
satuation/balance and the other factors that give "feel" to an image.
IMHO THAT is what people should be looking at, not stuff you will NEVER
see in a final printed image. That's just being a techie nerd about
photography.

Stephanie

On Wed, 02 Sep 2009 04:27:33 -0400, "
wrote:

Alan Browne wrote:
Brad Sanborne wrote:
I started to wonder how all this brouhaha over sensor-noise and
resolution
came to be the determining factor in quality photography. Then it
dawned on
me.

Before digital cameras only people working in their own darkrooms studied
their negatives and slides with a good powered loupe. Usually only 5x,
8x,
or at the most 10x power. In fact I have my old 8x loupe sitting
beside me
right here, a little desktop reminder of my darkroom days. The average
photographer and snapshooter, of which there are millions today, used
to be
happy with recovering their pack of prints or 8x10s from the local lab.
crud snipped

They have totally missed the big picture .... and probably always will.

And you're missing a much more important point.

High quality gear is expected to provide high quality results. This was
true in 1900, 1925, 1950, 1975, 2000 and now. However, dollar for
dollar we are getting ever increasing quality and capability not to
mention immense time/cost savings from avoiding film and unneeded prints.

This means that the average snap shooter is getting commendable
technical (and often aesthetic) results with far less than would have
been neccessary in the past. Further Mr. Snappy takes more photos (no
cost/convenience issues) and gets better at it and gets more useful
results more often. The future bodes well for all.



I have to disagree here. Being able to "take more photos" doesn't =
higher quality. It makes people lazy at composition/metering/focusing. I
highly doubt digital increases the amout of keepers. I'm TOTALLY
including myself and is one reason I have gone back to shooting
landscapes with my 4X5. With film I got at least a 50% "hit" ratio, with
digital I probably throw out 90% of that I shoot.


I've noticed this myself recently, taking a lot of tight macro shots of
dragonflies and bees in flight hovering just above purple thistle blossoms
for the last few days. Good color compliments for composition. (Plus I
learned about a new species of leaf-cutter bees, they have a unique
pollen-gathering behavior.) Anyway... With film I might have shot half a
roll and been satisfied with most of them. With digital its too easy to not
pay close enough attention to composition. Knowing I can always take more,
I can always easily crop and reframe on the computer. I'm just not trying
as much as I used to, to get it right the first time. I used to like the
challenge. I may have to rethink this and consider overcoming the lack of
challenge as the new challenge.

I still get lots of great photos but none that I'm really excited about.
Those seem to be happening fewer and further apart. Perhaps a simple
solution is to use memory card capable of holding only 36 images. Use those
up for the day and I have to call it quits. Also making a deal with myself
that I can't delete any to make room for further attempts.

I too used to think that the instant-feedback aspect to digital would make
people better photographers faster. I'm strongly reconsidering this. With
film if you took a once-in-a-lifetime shot and got your negatives back days
later finding that shot ruined, you agonized over what went wrong so you'll
never make that mistake again. You'll never forget that painful lesson.
Digital brings a huge laissez-faire attitude to the plate. So what if you
missed that once-in-a-lifetime moment? You can always shoot more images.

No, you can't. That moment in time is gone forever. The feeling that you
can shoot more images and try again won't make you realize that. All your
once-in-a-lifetime shots might turn out like crap forever more just because
you think you can reshoot them anytime you want. The follow-up frames might
be technically correct now but they're insipid and boring. You missed the
moment. The discipline of getting it right the first time is disappearing
from photography. The world is becoming flooded with automatically focused
banality.

I recall a moment I recently missed from having this digital laissez-faire
attitude. During sunset I was passing by my picture-windows and noticed one
lone white tree-trunk in woods being lit up by a spear of intense red light
from the setting sun. The rest of the woods dark and silhouetted. The
alignment of the trees and sun being just right to allow that one spotlight
of red to sneak by all the other trees in the woods to hit just that one
white trunk. I grabbed my camera, stepped outside, and started to fire off
a few shots. I wasn't paying that close enough attention to the first few
frames. Quickly reviewing them the exposure was off, the red being blown
out and too light. So what, I can shoot more. Right? By the time I looked
back that intensely red spear of light was gone. The moment lost in less
than 30 seconds, never to be recreated. The trees' branches will have
changed by this day next year, the sun may not set with an intense red on
that day, the field across the dirt road might be planted with corn next
year (planted with low soy plants this year) which will prevent that
shallow angle of light from even getting here. Yep, chances of it
recreating itself are gone, missed it.


Also the total automation make MOST new photographers clueless about
the basics. I wonder how many have a clue what "hyperfocal distance"
even means?

If they did they'd have far fewer complaints about focusing speeds and back
and front focusing issues. Further adding to the "so what, I can just shoot
more and get it right next time." Leading directly back to the "no you
can't" closed-loop of half-assed photography.


As far as the OP, pixel peeping -is- just silly when you throw out color
satuation/balance and the other factors that give "feel" to an image.
IMHO THAT is what people should be looking at, not stuff you will NEVER
see in a final printed image. That's just being a techie nerd about
photography.

Stephanie

wrote:
Alan Browne wrote:

This means that the average snap shooter is getting commendable
technical (and often aesthetic) results with far less than would have
been neccessary in the past. Further Mr. Snappy takes more photos (no
cost/convenience issues) and gets better at it and gets more useful
results more often. The future bodes well for all.

I have to disagree here. Being able to "take more photos" doesn't =
higher quality.

Does *not* taking photos somehow teach photography, or even
successful snapshooting?

Or are you saying that instant feedback and the ability to quickly
try different things and see what happens, as digital provides,
will not improve quality? That is what you imply!

It makes people lazy at composition/metering/focusing.

Maybe true for you. But for everyone?

I highly doubt digital increases the amout of keepers.

So you are saying the amount of nonkeepers multiplied by millions.
That's a pretty tall claim you'd need to prove.

I'm TOTALLY
including myself and is one reason I have gone back to shooting
landscapes with my 4X5. With film I got at least a 50% "hit" ratio, with
digital I probably throw out 90% of that I shoot.

Hmhm. So you are not experimenting with film, else you wouldn't
get 50+% keepers, or you have a much lower standard when it comes
to film ...

.... and you shoot just 5 times more with digital than with film?
And you expect anybody to do that?
Especially those who have a digital camera, but no film camera?

Also the total automation make MOST new photographers clueless about
the basics. I wonder how many have a clue what "hyperfocal distance"
even means?

Does it matter? No, not at all.
What matters is that tons of people have cheap access to basic
photography, thanks to digital. If even *one* of them creates
*one* keeper, the amount of keepers has *increased*.

Let me repeat that, slowly: If even one single keeper is created by
someone who would not have taken the shot with film (for whatever
reason: too expensive, too complicated, wouldn't have brought
the gear along, ...) then the amount of keepers has increased
and thus digital has increased the amount of keepers.

You need to prove that that has not and will not happen to support
your claim.

And then, most people don't even look for keepers, they look
for *memories*[1], thus the demand "you *must* produce keepers"
is elitist and completely off base from reality.

As far as the OP, pixel peeping -is- just silly when you throw out color
satuation/balance and the other factors that give "feel" to an image.

Color saturation and balance can be adjusted in post processing.

IMHO THAT is what people should be looking at, not stuff you will NEVER
see in a final printed image. That's just being a techie nerd about
photography.

Sharpness cannot be increased (only apparent sharpness, and that
only to a small degree, though you can blur). How does one learn
about critical focus and how good or bad the AF works in different
situations when one does not check such stuff?

-Wolfgang

[1] just have someone close to you die and look at the photos taken
from that person for instant education.

On Wed, 02 Sep 2009 08:57:04 -0500, Jake Nielson
wrote:

I still get lots of great photos but none that I'm really excited about.
Those seem to be happening fewer and further apart. Perhaps a simple
solution is to use memory card capable of holding only 36 images.

I remember the days when I was out for the afternoon with my 3
double-sided 4" x 5" plate holders I would get quite annoyed with
those people who in the same time could take 36 photographs with their
busy little 35mm cameras. What amused me is that at the end of the day
we would be lucky if we had one good shot each.

I still tend to take as much trouble with my D300 as I used to take
with the plate camera. I won't comment on the results.



Eric Stevens

Brad Sanborne wrote:
I started to wonder how all this brouhaha over sensor-noise and resolution
came to be the determining factor in quality photography. Then it dawned on
me.

Before digital cameras only people working in their own darkrooms studied
their negatives and slides with a good powered loupe. Usually only 5x, 8x,
or at the most 10x power. In fact I have my old 8x loupe sitting beside me
right here, a little desktop reminder of my darkroom days. The average
photographer and snapshooter, of which there are millions today, used to be
happy with recovering their pack of prints or 8x10s from the local lab.
Never once looking at the quality of the negatives or slides beforehand--no
more than looking at it to see if it was a pleasing image and there wasn't
too much blur to allow appreciation of the subject. The subject being the
whole point of photography. Today they all easily take a high-powered loupe
to every photo ever taken. Do the math on how much magnification you are
seeing by viewing a 3648 pixel-width image from a 10-megapixel camera on an
average 96-dpi LCD monitor at 1:1 resolution. That's like looking at a
negative with a 38x-power dissecting microscope. Zooming into the negative
or slide with magnifications far beyond what the professional darkroom
photographers even had with their loupes, just to see what is there. To see
if their image is worth appreciating or worth printing.

In film photography days, sure, sharpness and resolution was important, but
the experienced photographer knew that the subject of the photograph itself
was far more important than any technical quality of that image. Without a
decent subject and composition then all the technical quality in the
universe was meaningless, and still is. If a subject and composition was
good it would even withstand a lot of grain (noise) in the image when
printed. In fact grain was often incorporated to give certain photographs
the right mood and feeling. Heavens forbid that any pixel-level noise
should be in any photo today, noise of such small nature that it literally
disappears when printed. But not so to the pixel-peeper beginner
photographer who wants to pretend they are an instant pro. Does everyone
here forgot how many ways we used soft-focus filters, put meshes in front
of our lenses, or even smeared vaseline (even noise-oil in an emergency) on
filters just to reduce the contrast (dynamic range) and resolution of a
photo to obtain the style and mood that we needed in order to obtain a
useful and marketable image?

Today, we have millions of "Insta-Pro Snapshooters" who incessantly believe
and promote their inane belief that technical quality will always
compensate for their snapshot subjects and compositions. Just because it's
so easy for them to take that high-powered 38x loupe to their digital
negative and desperately look for something that might be worthwhile in
their snapshots at a pixel level. Not finding it, then they think there's
something wrong with their camera instead of themselves.

They have totally missed the big picture .... and probably always will.

Whilst agreeing with you that I can (and do) examine pics far more
critically since I went digital it's not completely without merit.

Whilst on holiday I only had my Sigma 17-70mm f2.8-4.5 with me and the
picture of my son on a fairground ride really needed something in the
range of 2-300mm. I've got a 1400x900 crop from the 10mp RAW file that
is still very sharp, it makes a great desktop and could be printed
still if req'd.

Had I used the sony kit lens I doubt I'd have got anything usable. I
took some test shots with my Tamron 70-200mm f2.8 and compared them to
test shots taken with my Sigma 135-400mm APO and cropping the Tamron's
images to match the view of the sigma at 400mm produces a picture only
slightly less sharp, though a lot less pixels.

The sigma is now redundant and going for sale on ebay shortly.

Mike