LF+scan+print: Case study, with prints

G- Blank wrote:

In the grand scheme that's all that matters.
Yup

rafe b wrote:

Optical prints only, please. I'm quite familiar with the
nature of inkjet prints. What I would like from you
or anyone else is an example or specimen of what you
would consider an exceptionally sharp optical print, at
2x or more if from LF, or 4x or more if from MF.

[The scans I posted at the start of this thread were
of prints at 2x and 6x, from LF.]

Ok so this is a bit of a cheat since it is from 35mm and not MF but
here is a scan of a 4x print of 35mm
http://www.pbase.com/konascott/image/51936563/original
This is sharper then anything I get from Costco and their Noritsu model
3101.

For low enlargements (dare I use that word) an optical print can be
extremely sharp. As the enlargement gets larger the advantages of
scanning the film becomes apparent, IMO.

I don't have a top of the line inkjet to compare to.

Scott

"Scott W" wrote in message
oups.com...

Ok so this is a bit of a cheat since it is from 35mm and not MF but
here is a scan of a 4x print of 35mm
http://www.pbase.com/konascott/image/51936563/original
This is sharper then anything I get from Costco and their Noritsu model
3101.


Offhand it looks like a good sharp optical print
but I don't have anything to reference it to right
this moment. Later, maybe.

For low enlargements (dare I use that word) an optical print can be
extremely sharp. As the enlargement gets larger the advantages of
scanning the film becomes apparent, IMO.

I said exactly that in my very first post in this thread, BTW.
The tiny scan of the 8x10" print isn't all that impressive, but
the scan of a 24x30" print seems to capture most of what
was in the original film scan. And that was with an ancient
Epson 7000 printer, on matte paper.

I don't have a top of the line inkjet to compare to.

I'll see if I can come up with something later. Tell
me more about how the print was made, if you would.


rafe b
www.terrapinphoto.com

Scott W wrote:

Tom Phillips wrote:
G- Blank wrote:

Fair enough, though the goal should be -not to need burning and
dodging,....something I strive for and usually accomplish especially
when creating color imagery.

If I might opine...

Exposing a negative that makes a straight
print on G2 or 3 without any burning or
dodging or other contrast controls was the
first assignment I ever had in my college
ZS class (too many years ago to tell.) But
while a useful professional exercise in
learning how to produce a good negative and
print, it hardly satisfies personal creative
inclinations whether b&w or color

Burning and dodging is a natural extension of
negative exposure. So-called digital burning and
doging ain't, since there's no light involved,
and I've been using PS for years and years.
Digital allows for contrast and brightness
manipulations, but I'd say in PS it's esay to
overdo and usually appears little contrived,
meaning it's harder to blend (feather) pixels
so the resulting contrast looks as natural as a
traditional burned and dodged print. For one,
if you change the image gamma, the brightness and
contrast manipulations become readily apparent.

I assume you are using levels for burning and dodging since brightness
and contrast controls are about worthless.

I have done burning and dodging in both the darkroom and using
Photoshop and I find I have way more control in Photoshop.

I guess my point was the two aren't really comparable IMO.

When burning and dodging a negative you are making an
actual exposure, hence traditional burning and dodging
is, as I say, a natural extension of negative exposure.

While PS levels and other contrast adjustments allows extensive
digital image control, I wouldn't say it offers more control.
In fact with color it offers less since the color gamuts of
most output devices vs. the mode gamut always shrinks. You
can make all the color adjustments you want; you won't get a
full gamut of colors.

When printing b&w negatives you have vast printing controls
at one's disposal, not just burning and dodging. Paper grades,
various developers affecting print contrast and color, bleaching,
etc allow very fine contrast controls. I doubt many of the fine
prints I've made could be duplicated in photoshop..

Scott W wrote:

Tom Phillips wrote:

yeah, never mind Nyquist rafy boy, if you even
know what it is...which pragmatically only
applies to pixels and doesn't apply to film.
It is more complicated in film but it applies. Film is a sampled
system, a gain is
either exposed or it is not. Film does not sample in a uniform maner,
but this
is in no way a loop hole around the Nyquist limit.

Note I said "pragmatically." What this means is
when using high quality optics with film (expecially
high resolving power films) there's no need to reduce
scene signal frequencies as is a standard requirement
with most digital. There simnply aren't enough pixels
available to handle high signal input frequencies.

Silver halides are exposed and record tonal detail on
a molecular level. No photodector can ever do this.
There is an inherent size limitation due to the
electronics involved vs the exposure needed to produce
a good digital signal.

BTW the term pixel was being used in film long before digital camera
existed.

The term pixel applied to video, not film.
Photoscientists use comparitive pixels for
determining film pixels, but is not an exact
correlation. Basically it's the number of
photons absorbed in a given area of either
film or a photodetector. But it really only
relates to the number of such pixels needed
for nominal print comparisons, i.e., the point
at which a viewer can't tell the difference
between images from film or digital using
typical 4x6 prints. That threshold is generally
6 mp.

There's SIMPLY NO SUCH THING AS A BIGER SENSOR.
It's typical digital geek nonsense...no silicon
sensor will EVER be able to have the resolving
abilities of LF film since (factually, as opposed
to ignorant geek misinformation) 1) silicon wafers
cannot be made that large,
And just how large to you think a wafer is?

and 2) photodectors

Not as large as LF film. There are engineering
limitations.

(that's pixels for rafe's pitiful education...)
No, a photodector and pixel are not always the same thing (for you
rather limited education)
can't get that small.
And just how small do you think a photodector can get?

For a decent signal, about a 5 microns pitch is average.
Most are larger, because the larger the pixel the better
signal to noise ratio. 50 - 80 square microns is common

With film it's just the opposite: the smaller the
silver halide crystal the better the resulting
signal to noise ratio is (which in film has to do
with grain and enlargement, so is really a different
animal than electronic noise.) And silver halides are
typically 1 micron .

G- Blank wrote:

In article .com,
"Scott W" wrote:

And just how small do you think a photodector can get?

The electrochemical "photo detectors" of silver based films are
molecular in size. The film itself is the actual detector
therefore a sheet of film can be created to any size and has the
potential to have limitless information stored.

Good point. 20x24 inch film, anyone?

The light sensitive sites of electric pulsed signal convertors
"detectors" are not nearly that small.

If you want to speak purely on an impulse level the ultimate final
answer is atomic- but then man hasn't conclusively created anything new.

And you can't make a photodetector that small since
the signal would then be useless. Photodetectors have
to be a certain size (area) since the exposure threshold
is vastly greater for a pixel than for a silver halide
crystal.

As you note halide exposure is molecular. Meaning it
only takes a minimum of 3 photons striking any given
molecule in any given crystal to expose it and initiate
photolysis (chemical decomposition which results in
silver thus forming an image.)

rafe b wrote:

"G- Blank" wrote in message
...
In article .com,
"Scott W" wrote:

And just how small do you think a photodector can get?

The electrochemical "photo detectors" of silver based films are
molecular in size. The film itself is the actual detector
therefore a sheet of film can be created to any size and has the
potential to have limitless information stored.

Limitless information? Not on your life.
Scalability of film has its limits, as you know.

By comparison with silicon wafers, it doesn't.
The greater limitation would carrying around
cameras and film holder larger than 24 inches...

Mr. Stafford is still looking for just the right
sort of wheelbarrow with which to haul his
24-pound behemoth.

The light sensitive sites of electric pulsed signal convertors
"detectors" are not nearly that small.

No, the best are around 5 to 7 microns on a side.

But individual grains of silver don't resolve anything
either and are much less predictable than a single
CCD or CMOS sensel.

rafe b
www.terrapinphoto.com

rafe b wrote:
"Scott W" wrote in message
oups.com...

Ok so this is a bit of a cheat since it is from 35mm and not MF but
here is a scan of a 4x print of 35mm
http://www.pbase.com/konascott/image/51936563/original
This is sharper then anything I get from Costco and their Noritsu model
3101.


Offhand it looks like a good sharp optical print
but I don't have anything to reference it to right
this moment. Later, maybe.

For low enlargements (dare I use that word) an optical print can be
extremely sharp. As the enlargement gets larger the advantages of
scanning the film becomes apparent, IMO.

I said exactly that in my very first post in this thread, BTW.
The tiny scan of the 8x10" print isn't all that impressive, but
the scan of a 24x30" print seems to capture most of what
was in the original film scan. And that was with an ancient
Epson 7000 printer, on matte paper.

I don't have a top of the line inkjet to compare to.

I'll see if I can come up with something later. Tell
me more about how the print was made, if you would.
It is a print from a friend so I don't know much other then he was
shooting with a Nikon 35mm film camera at the time. The print is only
a 4 x 6, so about a 4x enlargement. From the look of it I don't think
it can be anything but an optical print, which is kind of odd because
it was made in 2003. He use to do his own printing I don't think he
has for years. It is however one of the sharper prints I have on hand.
I'll email him as ask how he had it printed.

Scott
..

Scott W wrote:

Tom Phillips wrote:
Simply and factually stated, the scan resolution
determines the digital image size. One can upsample
or downsample (which greatly degrades image quality
because you either lose image detail or interpolate
image detail), but one cannot enlarge that detail.
The detail in a digital image is maximized at the
scanned resolution and is all there is.

I had lost track of this thread for a while so I apologize for
responding to several day old posts.

Scanning is of course a sampling system, and like any sampling system
you at some point reach a limit where more samples give no more
information.

The sampling is limited by the scanner's resolution.
Thus if you have a 10 mp camera you cannot sample at
10 mp, but about 1/3 to 1/4 of that (nyquist states
1/2 minimum but in actual practice it/s 1/3 to 1/4.)

If you sample past this limit, the Nyquist limit, then
down sampling does not entail information lost (read detail). Whereas
up sampling does not add any information it does not lose any either.

If you sample betyond the limit aliasing and artifacting
result. Plus you always lose information when upsampling
or downsampling. It's inherent in the process...

Tom Phillips wrote:
While PS levels and other contrast adjustments allows extensive
digital image control, I wouldn't say it offers more control.
In fact with color it offers less since the color gamuts of
most output devices vs. the mode gamut always shrinks. You
can make all the color adjustments you want; you won't get a
full gamut of colors.
Since I do my prints on photographic paper I don't see how this could
be an issue.

Scott