Maskless color negative scanning was Scanner woes

  • Thread starter Thread starter Ernst Dinkla
  • Start date Start date
E

Ernst Dinkla

Everybody here seems to agree that the mask in color negative
film is a good thing. I think we would be better off without a
mask if it is for scanning. Agfa and Kodak made maskless
negative films for aerial photography and copy work that had
to be scanned. Making color prints was not the first goal of
that material. There must be a good reason why they made it
maskless. Better S/N and grain is mentioned for the Aviphot
maskless films.

http://www.agfa.com/docs/sp/aerial/aviphot_x100_2004-06-15_en.pdf

Ernst

--

--
Ernst Dinkla


www.pigment-print.com
( unvollendet )
 
Ernst said:
Everybody here seems to agree that the mask in color negative
film is a good thing. I think we would be better off without a
mask if it is for scanning. Agfa and Kodak made maskless
negative films for aerial photography and copy work that had
to be scanned. Making color prints was not the first goal of
that material. There must be a good reason why they made it
maskless. Better S/N and grain is mentioned for the Aviphot
maskless films.

http://www.agfa.com/docs/sp/aerial/aviphot_x100_2004-06-15_en.pdf

Ernst

Ernst,
Wouldn't it be better to scan the film as a positive transparency and
convert in Photoshop? You won't get perfect color, but adjustable from
there. Problem is that by far the most color neg film has the orange
mask. Earlier scanners seemed to have film types built into their
software, lately that doesn't seem to be the case.

Tom
 
tomm42 said:
Ernst,
Wouldn't it be better to scan the film as a positive transparency and
convert in Photoshop? You won't get perfect color, but adjustable from
there. Problem is that by far the most color neg film has the orange
mask.

I think the problem Ernst wants to resolve by using maskless film is the
fact that the mask reduces the density range of negative film, not that
it influences the inverted colors.
 
Ernst Dinkla said:
Everybody here seems to agree that the mask in color negative
film is a good thing. I think we would be better off without a
mask if it is for scanning.

The benefit would manifest itself in reduced scan times if compared
with masked CN film. For pictorial, portrait or product photography
however, color accuracy should be the deciding factor IMHO. Films like
Portra NC scan quite well, and they are not terribly dense (which also
means they can become grainy with boosted contrast/gamma). I usually
overexpose them a little (1/3rd stop) for scanning, it helps shadow
definition and may reduce graininess a bit.
Agfa and Kodak made maskless negative films for aerial photography
and copy work that had to be scanned. Making color prints was not
the first goal of that material. There must be a good reason why
they made it maskless.

Indeed, color accuracy isn't the main goal with Aerial photography,
but feature discrimination usually is, so they went for resolution and
contrast (to cut through the haze) and a simpler/cheaper design. The
faster scan speed apparently played a roll as well, as they make
specific mention of it (but they also hasten to add that the colors
are clean and saturated).

In the PDF link you provided, it is striking that they omitted the
Spectral Dye Density Curves commonly provided to illustrate the amount
of secondary absorptions (implying the effect a mask could have on
color accuracy/saturation).

Wat also is clear, is that the goal was for a neutral film base and,
while they came close, there is still some scanner gain/channel
exposure adjustment needed (as far as we can take the graphs for
accurate), but it might scan twice as fast as a masked film. Also
clear is that the non-parallel RGB density curves will require a white
balancing (+ complex gamma adjustment) after invertion, as usual,
because we're not printing but scanning.
Better S/N and grain is mentioned for the Aviphot maskless films.

http://www.agfa.com/docs/sp/aerial/aviphot_x100_2004-06-15_en.pdf

As usual, by relaxing a few constraints, it is possible to improve
some other aspects.

In most types of aerial photography, resolution (for high altitude
recognisance or topography) or speed (both film sensitivity and
processing speed) are the more important aspects. For low altitude
oblique photography, films with more contrast allow to cut through the
haze (heavy UV-filtering alone is not enough), unless there is an
exceptionally clear atmosphere.

Bart
 
Bart said:
The benefit would manifest itself in reduced scan times if compared with
masked CN film. For pictorial, portrait or product photography however,
color accuracy should be the deciding factor IMHO. Films like Portra NC
scan quite well, and they are not terribly dense (which also means they
can become grainy with boosted contrast/gamma). I usually overexpose
them a little (1/3rd stop) for scanning, it helps shadow definition and
may reduce graininess a bit.

The color could be sufficiently corrected by custom profiling.
I do not see a reason for masking if the film itself is no
longer used for analogue prints. Increasing the AG (exposure
time actually) for Blue at 3.5x and Green at 2.5x and on top
of that a correction profile doesn't sound as a better
solution than good software + a custom profile. Slides are
scanned too and they also suffer of color inaccuracy but that
is solved with profiling as well. The inconsistency of the
mask influenced by the development process is a known issue,
so at one hand there is an analogue color accuracy created by
the mask but it also creates inaccuracies.

I also use 100/160-320/400 for Portra NC. Portra has been made
more suitable for scanning (including new surface coats for
better ICE results since 2005), it still is a color negative
film for analogue prints in the first place. Even the Agfa
Aviphot can still be used for analogue prints if needed. I
included it as an example of a maskless film but do not see it
as a ideal maskless film for scanning.
Indeed, color accuracy isn't the main goal with Aerial photography, but
feature discrimination usually is, so they went for resolution and
contrast (to cut through the haze) and a simpler/cheaper design. The
faster scan speed apparently played a roll as well, as they make
specific mention of it (but they also hasten to add that the colors are
clean and saturated).

In the PDF link you provided, it is striking that they omitted the
Spectral Dye Density Curves commonly provided to illustrate the amount
of secondary absorptions (implying the effect a mask could have on color
accuracy/saturation).

Wat also is clear, is that the goal was for a neutral film base and,
while they came close, there is still some scanner gain/channel exposure
adjustment needed (as far as we can take the graphs for accurate), but
it might scan twice as fast as a masked film. Also clear is that the
non-parallel RGB density curves will require a white balancing (+
complex gamma adjustment) after invertion, as usual, because we're not
printing but scanning.


As usual, by relaxing a few constraints, it is possible to improve some
other aspects.

In most types of aerial photography, resolution (for high altitude
recognisance or topography) or speed (both film sensitivity and
processing speed) are the more important aspects. For low altitude
oblique photography, films with more contrast allow to cut through the
haze (heavy UV-filtering alone is not enough), unless there is an
exceptionally clear atmosphere.

Bart

I'm not convinced that a general purpose color negative film
needs a mask if it is only used for scanning. The Aviphot x100
is just one of several aerial Agfa films, some with higher
contrast, some with a mask, some with higher speeds. They are
made more suitable for Aerial photography but their features
for that task are not in a 1:1 relation with being maskless.
The color accuracy of a masked film is more in relation to
analogue printing but maskless will not be more of a problem
in scanning than slides are. The same secondary absorptions
happen with slide photography and can be dealt with in
profiling and software.

There's some analogy in the separation workflow of printing.
Thirty years ago it was a long and slow process to separate
slides for FC offset printing. Secondary absorptions in the
film emulsion and the printing inks had to be taken care off.
Film masks were needed and special analogue methods + color
masking films were used. Today that's all done with software
for even the simplest desktop inkjet. Custom ICC profiling on
top of that to increase the accuracy. I do not see why that
can't be done with color negatives.

Of course one could use slide film. But reversal film has in a
way a similar dual task, it is still made to suit projection
display. For that it has a higher contrast so it can't have
the tone compression and latitude a color negative film has.
An ideal scanner film could be negative or positive but should
have a dynamic range and Dmax that just suits the best desktop
scanners like the Nikon 9000 and still have the nice
compression/latitude of the color negative film. Color
inaccuracies can be cured in another way. That film will be
worthless for any other purpose than scanning.

Ernst


--

--
Ernst Dinkla


www.pigment-print.com
( unvollendet )
 
SNIP
The color could be sufficiently corrected by custom profiling. I do
not see a reason for masking if the film itself is no longer used
for analogue prints. Increasing the AG (exposure time actually) for
Blue at 3.5x and Green at 2.5x and on top of that a correction
profile doesn't sound as a better solution than good software + a
custom profile.

It is indeed possible to correct some of the issues with a good
profile, but I'm more in favor of prevention, rather than cure,
especially when it is so easy to adjust for the more obvious
contaminations by altering the gain.

Profiles are often very dependent on exposure level. If the exposure
for profile creation is different from the actual scan, colors will
shift all over the place. Removing the film base mask does exactly
what we want, create an optimal exposure level and as a bonus corrects
secondary absorptions *in the analog domain*!

I view profiling as supplemental, but not without its own issues.
Slides are scanned too and they also suffer of color inaccuracy but
that is solved with profiling as well. The inconsistency of the mask
influenced by the development process is a known issue, so at one
hand there is an analogue color accuracy created by the mask but it
also creates inaccuracies.

I agree that processing variability causes additional issues, but
profiling won't help either in that case. A benefit of slides is that
it is easier to produce an IT8 or similar target, and the dye-set is
the same for a large family of films (e.g. Ektachrome). Thus a profile
can be made more reliably than with CN film. Photographing a target on
film oneself to create a profiling target introduces a whole lot of
variables to the equation, making it a dubious solution.
I also use 100/160-320/400 for Portra NC. Portra has been made more
suitable for scanning (including new surface coats for better ICE
results since 2005), it still is a color negative film for analogue
prints in the first place.

I still regret the demise of Supra 100, a very good and easy to scan
film. The best film I've used over the years.
Even the Agfa Aviphot can still be used for analogue prints if
needed. I included it as an example of a maskless film but do not
see it as a ideal maskless film for scanning.

And I saw that Kodak has discontinued all of its Lab films, with some
interesting candidates.

SNIP
I'm not convinced that a general purpose color negative film needs a
mask if it is only used for scanning.
SNIP
The color accuracy of a masked film is more in relation to analogue
printing but maskless will not be more of a problem in scanning than
slides are.

I guess we differ in the sense that I more strongly believe in
"prevention is better than cure", but we agree that Profiling is a
useful step in the modern workflow.
The same secondary absorptions happen with slide photography and can
be dealt with in profiling and software.

Yes, since color dye impurities are the cause for secondary
absorptions, all films suffer from them:
<http://www.kodak.com/global/en/professional/support/techPubs/e190/e190.pdf?id=0.1.14.34.7.16.4&lc=en>
page 13 top left Spectral-Sensitivity Curves show huge amounts of
Magenta formation (=Green transparency) for originally Blue light,
*mixed* with intended Blue transparency from Blue light. To me that
seems very hard to unravel for a profile, but relatively
straightforward for a mask. At 500nm all three layers react with dye
formation. Thank goodness the 160NC on page 8 behaves *much* better.

Slide film is also reasonably behaved:
<http://www.kodak.com/global/en/professional/support/techPubs/e4024/f009_0526ac.gif>
but not entirely without issues:
<http://www.kodak.com/global/en/professional/support/techPubs/e4024/f009_0527ac.gif>

The unanswered question however is, how much better would the
resulting image be if it were possible to apply a mask? Despite my
scanners being profiled, It proofed to be almost impossible for me to
get a good looking scan from last years Scanner Shoot-out slide. I
unfortunately missed the deadline for my entry due to a medical
emergency in the family, but the results from others were as
troublesome as mine (which was both a relief and a disappointment to
me).
There's some analogy in the separation workflow of printing. Thirty
years ago it was a long and slow process to separate slides for FC
offset printing. Secondary absorptions in the film emulsion and the
printing inks had to be taken care off. Film masks were needed and
special analogue methods + color masking films were used. Today
that's all done with software for even the simplest desktop inkjet.
Custom ICC profiling on top of that to increase the accuracy. I do
not see why that can't be done with color negatives.

The question is; How good could it ultimately be from color negatives
as a source? There is little joy in using CN for process printing, but
slides are/were very common as input.

SNIP
An ideal scanner film could be negative or positive but should have
a dynamic range and Dmax that just suits the best desktop scanners
like the Nikon 9000 and still have the nice compression/latitude of
the color negative film.

Yes, something I also suggested a couple of years ago in this group,
as a last resort for film manufacturers in defence against a rapid
decline in film sales; give us a scanner (forget print) film closer to
a gamma 1.0, that would have probably combined the best of several
worlds.

Bart
 
I still regret the demise of Supra 100, a very good and easy to scan
film. The best film I've used over the years.


It was a fine film, as was Royal Gold 100.

But Reala's excellent, also. Unfortunately
it's not available in large format (4x5") sheets -
but still available in 35mm or MF.


rafe b
www.terrapinphoto.com
 
nothing spectacular, grain bigger than most Fuji negs. Relatively
contrassty good for aerial which is inherently low contrast scene.
It is a film design for a special purpose. Lack of masking dyes will
result with lesser color accuracy, which is not a big deal in
technical phtography. Just try it with faces, you may be unpleasantly
surprised.


On Thu, 15 Jun 2006 12:00:50 +0200, Ernst Dinkla

#
#
#
#
#Everybody here seems to agree that the mask in color negative
#film is a good thing. I think we would be better off without a
#mask if it is for scanning. Agfa and Kodak made maskless
#negative films for aerial photography and copy work that had
#to be scanned. Making color prints was not the first goal of
#that material. There must be a good reason why they made it
#maskless. Better S/N and grain is mentioned for the Aviphot
#maskless films.
#
#http://www.agfa.com/docs/sp/aerial/aviphot_x100_2004-06-15_en.pdf
#
#Ernst
 
which scanner can register all densities from masked negs of the
current crop?



#tomm42 wrote:
#> Ernst Dinkla wrote:
#>> Everybody here seems to agree that the mask in color negative
#>> film is a good thing. I think we would be better off without a
#>> mask if it is for scanning. Agfa and Kodak made maskless
#>> negative films for aerial photography and copy work that had
#>> to be scanned. Making color prints was not the first goal of
#>> that material. There must be a good reason why they made it
#>> maskless. Better S/N and grain is mentioned for the Aviphot
#>> maskless films.
#>>
#>> http://www.agfa.com/docs/sp/aerial/aviphot_x100_2004-06-15_en.pdf
#>>
#>> Ernst
#>>
#>> --
#>>
#>> --
#>> Ernst Dinkla
#>
#> Ernst,
#> Wouldn't it be better to scan the film as a positive transparency
and
#> convert in Photoshop? You won't get perfect color, but adjustable
from
#> there. Problem is that by far the most color neg film has the
orange
#> mask.
#
#I think the problem Ernst wants to resolve by using maskless film is
the
#fact that the mask reduces the density range of negative film, not
that
#it influences the inverted colors.
 
nailer said:
which scanner can register all densities from masked negs of the
current crop?

Many scanners can do that, provided they have adequate control over
the removal of the film's base+fog density by adjusting the individual
channel exposures. Most CN films have a Dmax minus Dmin of say 2.0, so
that shouldn't be too difficult.

I'd say most of the Epson and Nikon scanners that allow to scan
transparent materials can cope with that. I assume Canon scanners also
can do that, although some earlier models had a slightly limited way
of accurately adjusting exposure levels.

Your question is probably easier to answer if you mention a specific
model that you're contemplating, and then users of such a model may
share their experience.

In general, dedicated filmscanners should have no issues with coping
with the limited density range(!) of average negatives, but there are
quality differences in how they cope.

Bart
 
I'm not convinced that a general purpose color negative film
needs a mask if it is only used for scanning. ....
Of course one could use slide film. But reversal film has in a
way a similar dual task, it is still made to suit projection
display.

Just curious, but if I understand correctly, you would like to
continue using negative film and then scan it. Is that right?

Why not go completely digital and skip that intermediate step?

Especially now that digicams match and even supercede analog. And no
matter how good the film is for scanning, one still ends up with a 2nd
generation image. A film scan is essentially a "picture of a picture".

Yes, digital photography has its own problems (the Bayer pattern for
one) but on balance and given proper equipment it should outperform
scanned film and make the whole process much simpler (i.e. no grain,
no focusing problems with warped film, no scanner bias, etc.).

Don.
 
I stumbled upon Kodak patent 6,274,299 for a "maskless" negative film
that is designed for scanning. (Search for it here:
http://patft.uspto.gov/netahtml/PTO/srchnum.htm )

Skip to the "Background" and "Preferred Embodiments" sections. It goes
into some detail describing the design considerations. Eliminating the
mask and reducing the use of inter-layer effects allows improved
sharpness, lower grain, and better color reproduction. Lower "gamma
ratios" are involved, (its not clear to me what those are).

Cliff
 
CRames said:
I stumbled upon Kodak patent 6,274,299 for a "maskless" negative film
that is designed for scanning. (Search for it here:
http://patft.uspto.gov/netahtml/PTO/srchnum.htm )

Skip to the "Background" and "Preferred Embodiments" sections. It goes
into some detail describing the design considerations. Eliminating the
mask and reducing the use of inter-layer effects allows improved
sharpness, lower grain, and better color reproduction. Lower "gamma
ratios" are involved, (its not clear to me what those are).

Cliff

Helen Bach had some experience with Kodak maskless film, she
wrote a reply in a similar thread:

Which Kodak have done: HD Color Scan Film 7299. Unfortunately
it is
only available in 16 mm at the moment.

EK did produce a nother scan-only film a few years ago:
Primetime 5620
640T. It was unprintable, unless you wanted a weird, muddy
mess. It
was intended for direct transfer to video. The colour could be
taken
care of during transfer. For the time it was a high speed film
with
comparatively low graininess. Unfortunately the radical nature
of the
film caused problems for colourists (the guys who transfer film to
video) and it never really stood a chance. I was involved in
the field
tests in a small way, and made a short with it - entirely at
f/1 and
16⅔ fps (that's 50/3: three fields per film frame at 25
fps PAL,
slightly jerky). EK made sure that the transfer was a good one
and I
was happy with the result.

Even by modern standards it was not grainy.

Best,
Helen

End of quote.

Ernst

--

--
Ernst Dinkla


www.pigment-print.com
( unvollendet )
 
Back
Top