LF+scan+print: Case study, with prints

In article ,
"rafe b" wrote:


If we accept that scanners can capture the tonality
of film, then they should have no problem with prints,
which have much lower dynamic range and Dmax.

IIRC, chromes have a DR in the high-3 range, C41 negs
in the high-2 range, and the best inkjet prints in the low-
to mid-2 range. I'd expect optical prints to be in the
mid- to high-2's range.

Have you ever used a reflection densitometer?


As usual, given the limitations of the medium, I don't draw
conclusions about tonality, only about sharpness and detail.

Limitations of the media? More likely the limitations of
operators-photographers, printing people and a host of
other little issues related to how something can get fobar'd.

You actually should be drawing conclusions with your eye balls
and less with the math. Because eyes especially the ones that creates the
stuff should be the deciding and not based on sets or formula.

Math however like anything is somewhat
subject to interpretation (Like in where the decimal point is supposed
to go to say its good enough .

rafe b
www.terrapinphoto.com

We can pretty much eliminate anything done with transparencies
because only Cibachrome prints might be available, since I don't
do them may as well forget them.


--
"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918

greg_____photo(dot)com

"G- Blank" wrote in message
...
In article ,
"rafe b" wrote:


If we accept that scanners can capture the tonality
of film, then they should have no problem with prints,
which have much lower dynamic range and Dmax.

IIRC, chromes have a DR in the high-3 range, C41 negs
in the high-2 range, and the best inkjet prints in the low-
to mid-2 range. I'd expect optical prints to be in the
mid- to high-2's range.

Have you ever used a reflection densitometer?


Nope. Does that matter? Are the numbers wrong?

In any case, I cite these numbers only to make the
point (which you did not refute) that a decent scanner
ought to have little trouble getting the tonality off a print.

I am prepared to believe that optical prints have
a wider DR and higher Dmax than inkjet prints.
But see no reason to think that that a good scanner
will fail in either case (optical print or inkjet print.)

When I spoke of "limitations of the media" I was
refering only to this idea of comparing scans-of-prints
over the web.


rafe b
www.terrapinphoto.com

In article ,
"rafe b" wrote:

"G- Blank" wrote in message
...
In article ,
"rafe b" wrote:


If we accept that scanners can capture the tonality
of film, then they should have no problem with prints,
which have much lower dynamic range and Dmax.

IIRC, chromes have a DR in the high-3 range, C41 negs
in the high-2 range, and the best inkjet prints in the low-
to mid-2 range. I'd expect optical prints to be in the
mid- to high-2's range.

Have you ever used a reflection densitometer?

Nope. Does that matter? Are the numbers wrong?

The problem is the difference in terminology.
In darkroom work we might say that a color optical print
can represent 4-5 stops of scene information even when
the negative can hold as many as perhaps seven or more.

Personally I like seven because it means burning is not a huge
problem depending on what is contained on the film. 3 stops
can be too much if the there is a lot of dark and light and not much
between them.

In any case, I cite these numbers only to make the
point (which you did not refute) that a decent scanner
ought to have little trouble getting the tonality off a print.

No reason to refute it. Try it with that new 4990 (How are you
doing with that?) One thing those Perfection do- do well
is scan flat media.


I am prepared to believe that optical prints have
a wider DR and higher Dmax than inkjet prints.
But see no reason to think that that a good scanner
will fail in either case (optical print or inkjet print.)

When I spoke of "limitations of the media" I was
refering only to this idea of comparing scans-of-prints
over the web.

Ok- gotcha.


rafe b
www.terrapinphoto.com



--
"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918

greg_____photo(dot)com

On Tue, 10 Jan 2006 18:08:15 -0500, G- Blank
wrote:

In article ,
"rafe b" wrote:

"G- Blank" wrote in message
...
In article ,
"rafe b" wrote:


If we accept that scanners can capture the tonality
of film, then they should have no problem with prints,
which have much lower dynamic range and Dmax.

IIRC, chromes have a DR in the high-3 range, C41 negs
in the high-2 range, and the best inkjet prints in the low-
to mid-2 range. I'd expect optical prints to be in the
mid- to high-2's range.

Have you ever used a reflection densitometer?

Nope. Does that matter? Are the numbers wrong?

The problem is the difference in terminology.
In darkroom work we might say that a color optical print
can represent 4-5 stops of scene information even when
the negative can hold as many as perhaps seven or more.

I believe my terminology is fairly standard; it's
expressed on a log scale. In round numbers,
prints are around 2.0 -- meaning a dynamic
range of 100:1, which translates to just under
7 stops (2^7).


In any case, I cite these numbers only to make the
point (which you did not refute) that a decent scanner
ought to have little trouble getting the tonality off a print.

No reason to refute it. Try it with that new 4990 (How are you
doing with that?) One thing those Perfection do- do well
is scan flat media.


Getting some decent results now, but not with
the Epson driver or with Silverfast, but with Vuescan,
in raw mode, scanning negatives as chromes.

It's hardly the sharpest scanner I've owned, but
it will have to do, at least for the LF stuff.


rafe b
www.terrapinphoto.com

In article ,
rafe b rafebATspeakeasy.net wrote:

I believe my terminology is fairly standard; it's
expressed on a log scale. In round numbers,
prints are around 2.0 -- meaning a dynamic
range of 100:1, which translates to just under
7 stops (2^7).

Ok didn't realize it was log scale you were stating
yes- that makes perfect sense.

Getting some decent results now, but not with
the Epson driver or with Silverfast, but with Vuescan,
in raw mode, scanning negatives as chromes.

Eh? W-T-Heck why doesn't it work well if you scan them as negatives?
Of course I don't know the 4990 but I have done Ok scanning negatives
with the Epson driver importing to PS.

Here's an example-

http://www.gregblankphoto.com/GregsI...aries/pages/We
stBranch.html


It's hardly the sharpest scanner I've owned, but
it will have to do, at least for the LF stuff.

Yes well- "we" are well aware that flatbeds are big compromise
and best option unless will to plop 9-10k down.

rafe b
www.terrapinphoto.com

--
"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918

greg_____photo(dot)com

On Tue, 10 Jan 2006 19:55:43 -0500, G- Blank
wrote:

In article ,
rafe b rafebATspeakeasy.net wrote:

I believe my terminology is fairly standard; it's
expressed on a log scale. In round numbers,
prints are around 2.0 -- meaning a dynamic
range of 100:1, which translates to just under
7 stops (2^7).

Ok didn't realize it was log scale you were stating
yes- that makes perfect sense.

Getting some decent results now, but not with
the Epson driver or with Silverfast, but with Vuescan,
in raw mode, scanning negatives as chromes.

Eh? W-T-Heck why doesn't it work well if you scan them as negatives?
Of course I don't know the 4990 but I have done Ok scanning negatives
with the Epson driver importing to PS.


I dunno why the Epson driver is behaving this way.
It could be a hardware issue with the scanner.
What I noticed, with the Epson driver, is that when
the file arrives in Photoshop, one or two of the
channels (usually green) has a histogram that's
bottomed out.

This doesn't happen with Silverfast or with
Vuescan, but Silverfast has its own problems.


Here's an example-

http://www.gregblankphoto.com/GregsImages/Gallery%20Tributaries/pages/WestBranch.html

Nice. I like it. It's fun to compare real work.
It would be even better if we could post large
images without fear of piracy.

This scan was done yesterday on the 4990,
using VueScan. It's been downsampled to
1200x958. It looks quite stunning when printed
on the R1800. If you send me your street
addresss (privately) I'll send you a print,
my compliments.

http://www.terrapinphoto.com/backyard_trees_1_A.jpg

This is one of eight images that lay latent
since last spring, awaiting resolution of
"the scanner problem."

Here's a 1" x 1" square from the negative,
scanned @ 2400 and downsampled 3:1.

http://www.terrapinphoto.com/backyard_trees_1_A_snip.jpg


rafe b

In article ,
rafe b rafebATspeakeasy.net wrote:

On Tue, 10 Jan 2006 19:55:43 -0500, G- Blank
wrote:

In article ,
rafe b rafebATspeakeasy.net wrote:

I believe my terminology is fairly standard; it's
expressed on a log scale. In round numbers,
prints are around 2.0 -- meaning a dynamic
range of 100:1, which translates to just under
7 stops (2^7).

Ok didn't realize it was log scale you were stating
yes- that makes perfect sense.

Getting some decent results now, but not with
the Epson driver or with Silverfast, but with Vuescan,
in raw mode, scanning negatives as chromes.

Eh? W-T-Heck why doesn't it work well if you scan them as negatives?
Of course I don't know the 4990 but I have done Ok scanning negatives
with the Epson driver importing to PS.


I dunno why the Epson driver is behaving this way.
It could be a hardware issue with the scanner.
What I noticed, with the Epson driver, is that when
the file arrives in Photoshop, one or two of the
channels (usually green) has a histogram that's
bottomed out.

This doesn't happen with Silverfast or with
Vuescan, but Silverfast has its own problems.


Here's an example-

http://www.gregblankphoto.com/GregsI...s/pages/WestBr
anch.html

Nice. I like it. It's fun to compare real work.
It would be even better if we could post large
images without fear of piracy.

This scan was done yesterday on the 4990,
using VueScan. It's been downsampled to
1200x958. It looks quite stunning when printed
on the R1800. If you send me your street
addresss (privately) I'll send you a print,
my compliments.

http://www.terrapinphoto.com/backyard_trees_1_A.jpg

Thats cool & I'll see what I can round up

Send it here, don't have mail service to my physical address.
Make sure you mark the package do not bend.

http://www.gregblankphoto.com/Biography.html



--
"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918

greg_____photo(dot)com

rafe b wrote:

On Tue, 10 Jan 2006 08:03:05 -0500, G- Blank
wrote:

In article ,
rafe b rafebATspeakeasy.net wrote:

On Sat, 07 Jan 2006 02:10:24 GMT, "Scott W"
wrote:

It would also be interesting if one of the people who claims that optical
prints will have more detail then form scanning could do the same thing you
have, make an optical print and then scan the print.

So "BOTH" prints are scanned from the relative output?

What is "relative output"?

Really, the only thing "novel" about my post was that I
presented scans of prints, rather than (or in addition to)
a raw film scan. All the images (but one) are scans of
prints; either that smalll Letter-sized print from the Epson
R1800, or a 24x30" print from an Epson 7000.

No takers. I could have guessed.

First maybe it was missed (Like in my case) When
reading the LF group I tend to look for people asking questions
about how to do such and such versus people looking to prove a point.

We talked a few months ago about a "print exchange,"
but it never happened. I figured this was another way
to go about it, albeit on a very small scale.

All it involves is getting the print onto a scanner, which
can be a bit of a challenge, if it's a big print.

However: in both analog and digital, one can *enlarge*
as if making a big print, and yet print a small area of
that enlargement. I suspect that's what John C. will be
doing with my negative when he gets it.

Enlarge? Digital images don't "enlarge." Sorry.
Never read your PS manual, I guess. Also, any time
you scan you lose some detail, other than perhaps
from a drum scan. Although I've seen Imacon scans
that were pretty good.

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.

G- Blank wrote:

In article ,
rafe b rafebATspeakeasy.net wrote:

On Sat, 07 Jan 2006 02:10:24 GMT, "Scott W"
wrote:

It would also be interesting if one of the people who claims that optical
prints will have more detail then form scanning could do the same thing you
have, make an optical print and then scan the print.

So "BOTH" prints are scanned from the relative output?

No takers. I could have guessed.

First maybe it was missed (Like in my case) When
reading the LF group I tend to look for people asking questions
about how to do such and such versus people looking to prove a point.


I miss most of rafe b's posts on purpose, since
he's fixated on trying to prove digital points
in a LF nsg yet often doesn't know what he's
talking about...

In article ,
Tom Phillips wrote:

The detail in a digital image is maximized at the
scanned resolution and is all there is.

True however what you are not giving to consideration; is that a given
scan resolution can cover print sizes from y-x to x-y.

So a variety of size prints can be made from a high resolution
scan without any lose especially when one considers the relatively
low resolution required to make a print.

--
"To announce that there must be no criticism of the President,
or that we are to stand by the President, right or wrong,
is not only unpatriotic and servile, but is morally treasonable
to the American public."--Theodore Roosevelt, May 7, 1918

greg_____photo(dot)com