Pixel density revisited

Started Oct 22, 2008 | Discussions thread
bobn2
bobn2 Forum Pro • Posts: 61,171
Re: Pixel density revisited

natureman wrote:

bobn2 wrote:

natureman wrote:

bobn2 wrote:

Phil Askey wrote:

I think a 30+ page review correctly qualifies any statement about
pixel density and noise. Where is this reverse evidence?

It could be any length, but if it didn't contain any statements
relating noise content to final output size, it wouldn't be
qualified, would it? I've read through it quite carefully, and I
can't find any pointer there that would help me understand the
relative noise between the cameras for usual output sizes. There are
a lot of statements about how noisy the A900 is, but all related to
per pixel noise.

There's nothing "absolute" about "final output sizes". 100%
magnification is "absolute" and "qualified" and so is "per pixel
noise".

I agree with the first part of the statement - of course, since final
output sizes vary, there is nothing absolute. I disagree strongly
with the second, there is nothing 'absolute' either about 100% or per
pixel noise. For a start, the 'pixel' you are looking at on the
screen is an amalgam of data from a number of sensor elements,
combined together using an algorithm you know nothing of, with
unknown effects on the noise. As for qualified, that depends on the
qualifications, and 'per pixel noise' is not qualified in an easily
understandable way to the level or quality of noise apparent in a
viewed image.
Really, you are playing around with semantics, both in your use of
'absolute' here, and 'fair' in previous posts. Big words like
'absolute' and 'fair' sound fine and dandy, but, as I've tried to
point out before, for comparative reviews what's important is
usefulness and objectivity. Given that most people (admittedly not
all) buy cameras as devices for taking photographs, a useful test is
one that gives a clear indication of the likely quality of those
photographs and an objective test is one that produces an unbiased
assessment of the likely quality of those photographs. The measuring
and emphasising of tests that do not directly relate to output image
quality and then asserting that they do fails on both counts.

In your previous post you said "It could be any length, but if it
didn't contain any statements relating noise content to final output
size, it wouldn't be qualified, would it?"

Loom at the context, Phil Askey suggested that 30+pages was enough to qualify any statement about noise (a bit of a bogus claim, since only 3 or 4 pages are anything to do with noise) which raises the question, what qualifications would you want to make about noise. As I've said (repeatedly) what matters to most photograhers is final image quality, so if you don't qualify your statements with respect to that, they're not really qualified, in the sense of the discussion to that point. Of course, you can make any irrelevant qualifications you like, like the results of the tests apply only to rainy Tuesday mornings, but they're hardly hepful qualifications in the context of a review to assist purchase decisions.

What I'm saying is that dpreview does qualify it by showing 100%
comparison images from every camera in their reviews. In other words,
all cameras are compared at equal magnification and the amount of
magnification is specified,

Incorrect, cameras are compared at different magnifications, that is exactly the problem.

and that qualifies their analysis and
comparisons.

But not in a useful way.

In even more other words, the qualification is the fact
that all cameras are compared at an equal, specified magnification of
their images (100%). The results (noise, etc.) are shown on the basis
of that qualification.

You must have a different definition of 'magnification' from me. What is it that has been equally magnified? It's not the image frame, because that gets magnified differently according to the pixel count. It's not the individual pixel, because that gets magnified differently according to its size. There is no 'equal magnification' at all.

They also show some images at other magnifications.

Unspecified onces.

Regardless of pixel density, dpreview is showing what each camera
produces at its' maximum native output size,

No it doesn't. The native output size of a FF camera is 24x36mm, of an APS-C camera 23x14 or so, of a P&S less than a fingernail. DPR shows nothing at native output size, nor would it be useful to do so.

when they show
comparisons at 100%, and that is as objective as anyone can get.

What's not objective is the drawing of conclusions from tests which don't support them. Any test of anything is objective, it's the narrative that isn't. Unfortunately, the narrative also encourages the uninformed to interpret the test results in ways that don't hold water.

I
don't know about the rest of you but if I want to see what a camera
will do, noise-wise or anything else, at a particular print size, I
can just resize the 100% images myself.

Not in the way that DPR presents them, other sites provide better data for that purpose. But surely, the purpose of a review site is to present results in a suitably digested, easy to use form?

There are other things in the reviews that to me are not
objective/unbiased (like the lenses used) but comparing images at
100% is proper.

And I can't for the life of me see what 'proper' has to do with it.

-- hide signature --

Bob

Post (hide subjects) Posted by
igb
igb
igb
igb
igb
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow