DSLR v Consumer Image quality

So here is an interesting test look at the linked photo, on screen, and
then print out the photo, at 300 dpi. On screen the bottom image has
much more detail but when printed the top looks sharper. Show the
printed photo to a number of people as ask which is sharper.

If there were printed at say 150 dpi then which looks sharper depends
on how far away you are. Step back about 10 feet and you will see the
same effect.

http://www.sewcon.com/300dpi_test/IMG_3302.jpg

The point is looking on screen can fool you, there is nothing wrong
with looking at the photo at 100% on the screen as long as you don't
use this as the only judgement for the quality of the photo.

Scott

Scott

Chris Brown wrote:

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.

Be sure to take the bellows extention into consideration when doing
macro work or everything will be underexposed (been there done that).

Check with your labs. I have a box of E6 unopened. (The owner of) the
only lab I found locally that knew what 4x5 was said, Oh, our JOBO has a
tank that can accept 4x5. I'm not sure how it works.

You can wrap it up and mail it off though.

If you buy a Polaroid back off of ebay, you can shoot Polaroid positives
right off, and you can also use it has a holder for ready load. If I
ever really decide to do chrome, I'm headed in that direction.

Bob

On Wed, 23 Feb 2005 11:49:04 -0500, rafeb wrote:


Owamanga wrote:

There's nothing overly wrong with occasionally critiquing at 100% or
more if the possible destination print dimensions require that, but in
many cases it's like a bank teller who uses an electron microscope to
check your signature. A fine waste of time.

For 4x6" prints, most definitely a waste of time.
Maybe even for 8x10"

Right, then in principle, we agree.

At this point I've sold a good number large-ish
prints (20x30", 24x36") that formerly I'd only
done at 8x10 or 12x18". I'm getting enough $
from these to where I think my on-screen viewing
time is well justified and properly compensated.

Although there is nothing wrong with the strive for perfection, I'd
add, that in my experience of what people have been able to sell in
the past, work that is now hanging in homes and work places, the
expectation of a high resolution print from 35mm format at these
dimensions just isn't there. I have a 48" by 36" B&W print of Paris
that hangs in the bathroom, possibly from a 35mm camera, definitely
film. Up close it's as grainy as hell, but realistically what would
you expect?

It's "atmospheric." :-)

--
Owamanga!

On Wed, 23 Feb 2005 17:00:59 GMT, "David J Taylor"
wrote:

Owamanga wrote:
[]
As far as I know, as yet, no CRT or LCD has a 6Mpix resolution. I
guess it won't be long, but right now that's a pipe-dream.

9MP monitors are already available...

http://www.pc.ibm.com/us/intellistation/t221/

Ah yes. At $7,000 every household will have one by the end of the
year.

;-)

That thing makes HD standards look positively shabby.

--
Owamanga!

rafe bustin wrote:

I disagree that you need a print to properly see or
judge a digital image. There have only been a few
situations (in my experience) where a print revealed
flaws that were not evident on-screen.

It is not so much that a print will reveal flaws that you can not see
on the screen as the other way around, flaws that you see on the screen
will often not be visible on the final print. The other part of this
is detail that you can see on the screen my not be visible on the
print. As I posted earlier a shot that looks better on the screen can
look worse when printed, again check out the link to the photo and see
which half looks better on the screen and then print at 300 dpi and see
which looks better printed.

http://www.sewcon.com/300dpi_test/IMG_3302.jpg

Scott

Owamanga wrote:

Although there is nothing wrong with the strive for perfection, I'd
add, that in my experience of what people have been able to sell in
the past, work that is now hanging in homes and work places, the
expectation of a high resolution print from 35mm format at these
dimensions just isn't there. I have a 48" by 36" B&W print of Paris
that hangs in the bathroom, possibly from a 35mm camera, definitely
film. Up close it's as grainy as hell, but realistically what would
you expect?


I don't want to represent myself as a pro
photographer -- I'm definitely not. On the
other hand, in recent years, I've made a
nice chunk of change selling prints. And
I've learned this much -- it's not my own
judgment of the print that matters, but
what my customer thinks. Customer being
defined as he or she who hands me the cash.
(Most often she than he, by the way.)

Were it entirely up to me, I'd be a bit
leery of making 24x26" prints from 35 mm
negatives, or 20x30" prints from 10D
captures. But I'm not going to tell a
paying customer that I won't do that,
nor will I feel the least bit guilty
when I do.



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

Scott W wrote:

The eye can see detail to 0.7 line pairs per inch for 100% contrast, at
a light level of 23 foot lamberts, this drops to 0.4 line pairs per
inch at 10% contrast. And you can see down to less then 2% contrast if
the line pairs per inch are at 0.1.

At what distance, and do you assume that viewing distances for monitors
and prints are proportional?

What all this means is that you will see detail on the screen that you
will not see in the print.

I follow the logic, but in practice I almost always find things in the
print that were not obvious on the monitor, and not the other way round.
For instance, if I use the clone tool to remove a neon orange "for sale"
sign on a dock, I might work at 300% or even 1000%.

At 200% on the monitor it looks nearly flawless, but in the print it's
not quite that good.

Bob

bob wrote:
Scott W wrote:

The eye can see detail to 0.7 line pairs per inch for 100%
contrast, at
a light level of 23 foot lamberts, this drops to 0.4 line pairs per
inch at 10% contrast. And you can see down to less then 2%
contrast if
the line pairs per inch are at 0.1.

At what distance, and do you assume that viewing distances for
monitors
and prints are proportional?

It depends of the size and resolution of your monitor, for mine I
figure I need to be back about 4 to 6 feet. It also depends on how
close you view your prints. BTW I am assuming 8 x 10 prints when I say
4 to 6 feet.

What all this means is that you will see detail on the screen that
you
will not see in the print.

I follow the logic, but in practice I almost always find things in
the
print that were not obvious on the monitor, and not the other way
round.
For instance, if I use the clone tool to remove a neon orange "for
sale"
sign on a dock, I might work at 300% or even 1000%.

At 200% on the monitor it looks nearly flawless, but in the print
it's
not quite that good.

The eye has can see the lowest contrast detail at at about 0.1 line
pairs per minute (or about 30 line pairs per inch when viewed at 12
inches). If you have a low contrast defect that is blown up large
enough it will be less visible then when it is seen smaller. My guess
is that something like this is going on.

Scott

wrote:

Considering only image quality, up to A4 prints. DSLR users talk about
their superior image quality, but when I go to say, Steves Digicams,
and compare on-screen a 200% enlargement of the same image, far greater
than real life, I see very little difference in quality between a D70
and a FZ20.

I think this says more for the FZ20 than about DSLR models.
Most consumer digicams have lots of image problems, including
purple fringing at wide angle, bothersome sharpening/JPEG artifacting,
and washed-out highlights. The FZ20 and Minolta A2 seem to transcend
most of these problems.

All(?) DSLR models perform better than the FZ20 at 800-6400 ISO,
but that might not be terribly important to you.

On the downside, lens selection for DSLR is much worse than
what you've got already on your FZ20. It's a pathetic situation
when Sigma makes is the best cost/performance DSLR lens (18-125).

Scott W wrote:
So here is an interesting test look at the linked photo, on screen, and
then print out the photo, at 300 dpi. On screen the bottom image has
much more detail but when printed the top looks sharper. Show the
printed photo to a number of people as ask which is sharper.

There's either something wrong with your equipment or with mine. On my
screen the top photo is clearer. On my printer, the top photo remains
clearer at both 150 and 300 pixels per inch (as set in Photoshop).

At roughly 400% it becomes evident that the top photo is a blurred
(Gaussian?) version of the bottom, and that the blurring masks the
hideous .jpg artifacts that make the bottom image appear hazy. Zooming
in on the tower at 1600% you can see that this is clearly the case --
The artifacts in the bottom image spray the highlights over everything.
Perhaps the top photo was sharpened after it was blurred.

Bob