Scanning Kodak Ektachrome Slides - Your Recommendations Please

[Kennedy McEwen]

Thanks for the informative post Kennedy!

Question:

How could you go about dust/scratch removal on film that retains silver
in the emulsion?

With great patience! ;-)
(and moderate skill with the clone and healing tools.)

Since silver is opaque to the wavelengths we're likely to have in our
scanners, IR and UV transmission difference won't work.

What about dealing with the surface?

Ah, now you are asking about how we *could* deal with it. ;-)

I wonder if UV light on a scratch or dust would appear differently when
just looking at the surface of the film?

No different to visible light, in essence.

Do the layers of film below the surface respond differently to UV light?

Not that I am aware of.

Could you shine two beams of single wave length collimated light at the
surface at right angles and look at the diffraction pattern to
distinguish dust (above the surface) and scratches (below the surface)
from the actual top of the emulsion?

In theory, you could do something like that - which is essentially how
ICE on reflective media in flatbeds works, although it isn't diffraction
that identifies defects, just different shadows. In film, of course,
you have two surfaces, and neither is usually pressed against an
optically flat surface for other reasons. So it needs to be a bit more
complex than the solution used on flatbeds.

 

[Kennedy McEwen]

I'm using the same set of "dark" KCs for my tests. And by now I must
have scanned them 100s of times - at least! I always wondered if this
repeated exposure to the bright LEDs had any effect on the slides.

My instinct told me it should and above paragraphs suggests that is
does. Presumably, exposing slides to scanner's LED light is no
different to slides being projected. Or is it?

Scanning exposes the slide to *much* less light than conventional
projection.

Think about it this way, a projector has to pass enough light through
the slide (of approximately 1.5sq.in) to illuminate a screen that is
perhaps 15sq.ft in area with enough brightness for you to see all of the
detail. That is about 225x as much light as necessary to view the slide
directly without any magnification. The scanner has to illuminate the
slide with enough light to cover the detector length, about the same
size as the short edge of the slide. So, ignoring the sensitivity
differences between the scanner sensor and your eye, and any reflection
efficiencies from the screen, the projector exposes the entire slide to
about 225 as much light as the scanner does during the scan time of a
single line. And I bet you project you slides for more than a couple of
milliseconds each...

Scanning exposure is negligible compared to projection.

BTW. Wilhelm's work makes it quite clear that it is the light of the
projection lamp, not the heat, that is the damage mechanism with
Kodachrome.

 

[Kennedy McEwen]

I've
been of the opinion for dome time that Kodak want Kodachrome to die as
it's far too much trouble. If it weren't for the fact that a fair
number of high profile supporters are around I suspect they would've
killed it years ago. Maybe the current formulation is deliberate ;-)

K-14 was first introduced in March 1974, so it has taken them rather a
long time to kill it off. ;-)

 

[Peter D]

Let's rephrase that: because digital can, at least a lot of the time,
be copied losslessly, as long as the file hasn't been corrupted, saved
to a medium which has become corrupted, and is copied with reasonable
frequency to allow for disk substrate failure, disk failure and
radiation (cosmic). ;-)

Yes, digital is susceptible to damage/corruption/loss. But so is the actual
slide. Given the ease with which digital copies can be created (and
distributed) it seems to me that it is better than doing nothing.

My reasons for wanting to create digital copies are in case the originals
are ever lost or damages or degrade. If they don't, great. If they do, I
will have digital copies. If they are lost or degraded, then I am no worse
of than if the originals were and I have no digital copies. I dont' see the
downside other than the work.

I would recomend to anyone setting out on a project like this to make lots
of copies and then pass them on to family/others. Keep track of who has what
and if you lose yours you can go get another copy.

 

[Hecate]

K-14 was first introduced in March 1974, so it has taken them rather a
long time to kill it off. ;-)

LOL!

--

Hecate - The Real One
(e-mail address removed)
Fashion: Buying things you don't need, with money
you don't have, to impress people you don't like...

 

[Hecate]

Yes, digital is susceptible to damage/corruption/loss. But so is the actual
slide. Given the ease with which digital copies can be created (and
distributed) it seems to me that it is better than doing nothing.

My reasons for wanting to create digital copies are in case the originals
are ever lost or damages or degrade. If they don't, great. If they do, I
will have digital copies. If they are lost or degraded, then I am no worse
of than if the originals were and I have no digital copies. I dont' see the
downside other than the work.

I would recomend to anyone setting out on a project like this to make lots
of copies and then pass them on to family/others. Keep track of who has what
and if you lose yours you can go get another copy.

You are absolutely right. It was just that Don's answer gave the
impression that you made a digital file and it was alright for ever
whereas, as you rightly say, you need rather more than one copy (I
know Don didn't mean that either, but impressionable people may be
watching <g>).

--

Hecate - The Real One
(e-mail address removed)
Fashion: Buying things you don't need, with money
you don't have, to impress people you don't like...

 

[Don]

Scanning exposes the slide to *much* less light than conventional
projection.

OK, so that means: Keep on testing, Don! ;o)

Think about it this way, a projector has to pass enough light through
the slide (of approximately 1.5sq.in) to illuminate a screen that is
perhaps 15sq.ft in area with enough brightness for you to see all of the
detail. That is about 225x as much light as necessary to view the slide
directly without any magnification. The scanner has to illuminate the
slide with enough light to cover the detector length, about the same
size as the short edge of the slide. So, ignoring the sensitivity
differences between the scanner sensor and your eye, and any reflection
efficiencies from the screen, the projector exposes the entire slide to
about 225 as much light as the scanner does during the scan time of a
single line. And I bet you project you slides for more than a couple of
milliseconds each...

Scanning exposure is negligible compared to projection.

That's great news! Thanks as always for the explanation, Kennedy!

BTW. Wilhelm's work makes it quite clear that it is the light of the
projection lamp, not the heat, that is the damage mechanism with
Kodachrome.

I'm sure the heat does all sorts of other damage such as causing
excessive curling etc.

Don.

 

[Don]

(I
know Don didn't mean that either, but impressionable people may be
watching <g>).

LOL!

Don.

 

[Kapecki]

Some of the notes in this thread reflect a basic misunderstanding of
image stability, reinforced, alas, by other sources who should know
better.

Both contemporary Kodachrome and Ektachrome are very stable products,
but both emphasize dark stability over light stability. This is for the
simple reason that they spend over 99% of their time in the dark. There
are no "perfectly stable" dyes from anyone, and the balance of dark
(more properly "thermal"), light, humidity and pollutant stability has
to be chosen depending on the application.

The amount of light exposure, either in normal projection or scanning,
has a minimal effect on dye loss--period. Alas, there have been those
who have emphasized (or sometimes even limited testing to) light
stability, ignoring the other factors.

If your uses call for a transparency display material, then by all
means you should consider light stability first. But over 90% of
consumer prints are stored in the dark. Even most displayed pictures
are put away after 10 to 20 years and virtually all after 40 years.

Of course, many dye transfer products, not encumbered by having to form
dyes in situ, have exceptional stability, and if you're ready to spend
the time or money involved to produce them, they may be your answer.

 

[Kennedy McEwen]

Some of the notes in this thread reflect a basic misunderstanding of
image stability, reinforced, alas, by other sources who should know
better.

Both contemporary Kodachrome and Ektachrome are very stable products,
but both emphasize dark stability over light stability.

"Very stable" is a subjective term - objective measurements show that
both current Kodachrome and Ektachrome have significant deficiencies
compared to competitive products. Hence your use of such a subjective
term is highly questionable. The question may be "is Kodachrome or
Ektachrome good enough?" and, contrary to your claim, the independent
assessment indicates that the answer is a resounding "No!".

Current Ektachrome has nothing like the dark stability of Kodachrome,
which has "the worst of all current slide films" for light stability and
which is *very* significant under normal projection. Current Ektachrome
is indeed much improved over earlier Ektachrome in terms of dark
stability, but it is still grossly inferior to Fuji equivalent films.

The popularity of automated colour restoration features, such as ROC, on
film scanners is evidence in itself that popular film stability is
woefully inadequate even for amateur use.

This is for the
simple reason that they spend over 99% of their time in the dark. There
are no "perfectly stable" dyes from anyone, and the balance of dark
(more properly "thermal"), light, humidity and pollutant stability has
to be chosen depending on the application.

The amount of light exposure, either in normal projection or scanning,
has a minimal effect on dye loss--period.

Again, the published measurement data from *independent* authorities
disputes this claim. Please provide referenced which support your
claims - the ANSI group responsible for such assessments, of which
Henry Wilhelm is a founder member, disputes your claim.

Furthermore, whilst Wilhelm has measured and reported the fading
characteristics of these films according to published methods, he
actually points out that the established ANSI measurement method is, if
anything, likely to overestimate the stability of the emulsion (ie.
underestimate just how bad the problem actually is) when compared to
"typical" use!

Alas, there have been those
who have emphasized (or sometimes even limited testing to) light
stability, ignoring the other factors.

On the contrary - all factors should be considered, and many are by
Wilhelm, although he has emphasised light fading because it is much more
significant than you suggest. The whole point of an image is for it to
be viewed - and that cannot be achieved without light exposure. A
typical projector, such as a Kodak Carousel, exposes the slide to a
light intensity about 10x that of direct sunlight at the equator - and
some high intensity projectors used in large lecture theatres can exceed
100x! Needless to say, it doesn't take long *accumulated* exposure
through the life of the slide to cause fading - irrespective of whether
it spends most of its life stored in the dark at -18degC or not.

If your uses call for a transparency display material, then by all
means you should consider light stability first. But over 90% of
consumer prints are stored in the dark. Even most displayed pictures
are put away after 10 to 20 years and virtually all after 40 years.

The loss in Kodachrome K-14 density and consequential colour shift is
measurable and considered unacceptable for museum and colour critical
applications after only 20minutes of accumulated exposure in a typical
projector! It is considered unacceptable for the average commercial and
amateur use (considered to be at most 25%, 20% and 35% loss of red,
green and blue density respectively) in only one hour of accumulated
exposure. (Reference Table 6.1 of Wilhelm's book). And *most* of that
dye fading occurs within the first 10 to 15 minutes of accumulated
exposure!

One hour total projection isn't a lot, particularly for a well enjoyed
image, over even the limited period of use that you suggest, of 10 to 20
years! That is only one 1-minute display (the maximum recommended by
Kodak) every 4 months! The average slide might only see a few minutes
of projection exposure in its life, but your 'winners' can certainly
expect to get a lot more - and rapidly turn into faded losers!

Furthermore, this light induced fading also has a form of reciprocity
failure in that many short bursts of exposure induce the damage faster
than the same total exposure made on a continual basis. So those
Kodachrome favourites can really suffer.

 

[Kapecki]

Obviously one is free to choose a product based on whatever attributes
that one deems most important. For a product that will spend most of
its time in the dark, I'll put thermal stability first, and my own
collection of transparencies dating back to 1964, using materials from
multiple manufacturers, convinces me that I have made the right choice.


Further, I view "independent data" as that appearing in
peer-reviewed technical journals or from laboratories whose primary
funding comes from grants and institutions, not payments from
interested manufacturers. There is nothing inherently wrong with using
data from these other sources (including manufacturers) as long as you
know and acknowledge who is paying the bills and don't try to pretend
otherwise.

That said, let's be clear about a few things. It is absolutely
incorrect to say, referencing my earlier note, that "the ANSI group
responsible for such assessments...disputes your claim." As I'm
sure you must know from reading the ANSI standards, that group (along
with the ISO) writes test and/or reporting protocols for many things,
among them image stability testing. They do not evaluate products. They
do not do tests. They do not assess claims.

While in my opinion and that of others, these standards are in need of
updating (new media bring with them new or at least different
permanence problems, plus new research studies have provided further
insights), such revisions need to be based on real data, published in
the journals where others can evaluate the data's validity, not just
based on someone's opinion.

Having worked with numerous publications over the years that specified
or required the very products you denigrate, I would hardly say they
are "...considered unacceptable for the average commercial and
amateur use."

Further, despite your citation of ANSI standards, the endpoints you
quote are not the ANSI Illustrative Endpoints, nor are they based on
any psychophysical data published in a peer-reviewed journal. The only
such data that I know of (though it is for prints and not
transparencies) is by Oldfield, et. al., [Journal of Imaging Science
and Technology, 48, 495 (2004)].

You can, of course, choose any endpoint you want, and for some
archivists and curators a "Just Noticeable Difference" or JND (as
determined by a double stimulus response test) is the standard. But
there is no data to suggest that at 1 JND most users deem an image
"unacceptable." In fact, 1 JND is approximately within the
variability of processing or viewing illumination.

"Typical use" can cover a broad range of applications and, as I
indicated above, should play a major role is choosing which attributes
of a system one values over others. Again, however, data, not
speculation, is critical to making an informed decision. The largest,
best, and longest term studies I know of (though primarily focused on
negatives and prints) began in 1987 with the work of Anderson and his
colleagues [J. Imaging Technology, 13, 49 (1987); 17, 127 (1991)] and
were continued by Bugner and his colleagues into the present time.

As for "reciprocity failure" in accelerated testing, it can work
both ways, and it is not limited to light-induced changes only. The
only sure lifetime predictions are those based on natural aging, but
any serious accelerated test predictions must take into account the
most extended reciprocity data the tester can gather in the time
available and are always subject to revision.

 

[Kennedy McEwen]

Obviously one is free to choose a product based on whatever attributes
that one deems most important. For a product that will spend most of
its time in the dark, I'll put thermal stability first, and my own
collection of transparencies dating back to 1964, using materials from
multiple manufacturers, convinces me that I have made the right choice.


Further, I view "independent data" as that appearing in
peer-reviewed technical journals or from laboratories whose primary
funding comes from grants and institutions, not payments from
interested manufacturers.

As you, of all people, know Wilhelm's work has not only been peer
reviewed but he has been the subject of failed litigation by Kodak whom
it found more wanting than almost any other photographic manufacturer.
Accepting payment from interested manufacturers, many of whom receive
less than satisfactory reports as a consequence of that work, does not
preclude an author from being independent.

There is nothing inherently wrong with using
data from these other sources (including manufacturers) as long as you
know and acknowledge who is paying the bills

As for who is paying the bills:
"Kapecki is an imaging consultant who was a senior researcher for more
than 30 years with the Eastman Kodak Company".

Yes, we do need to keep an eye on who is paying the bills, especially
when they dispute facts without divulging their own interests!

and don't try to pretend
otherwise.

The terms pot, kettle and black come to mind - but black has probably
faded from yours!

 

[Ole-Hjalmar Kristensen]

I do not doubt your claim about Ekatchrome being inferior to Fuji
products, but I have 40 year old Ektachrome slides which are virtually
indistinguishable (by eye) from slides taken yesterday. Good enough
for me.

KM> "Very stable" is a subjective term - objective measurements show that
KM> both current Kodachrome and Ektachrome have significant deficiencies
KM> compared to competitive products. Hence your use of such a subjective
KM> term is highly questionable. The question may be "is Kodachrome or
KM> Ektachrome good enough?" and, contrary to your claim, the independent
KM> assessment indicates that the answer is a resounding "No!".


KM> Current Ektachrome has nothing like the dark stability of Kodachrome,
KM> which has "the worst of all current slide films" for light stability
KM> and which is *very* significant under normal projection. Current
KM> Ektachrome is indeed much improved over earlier Ektachrome in terms of
KM> dark stability, but it is still grossly inferior to Fuji equivalent
KM> films.

 

[Kennedy McEwen]

I do not doubt your claim about Ekatchrome being inferior to Fuji
products, but I have 40 year old Ektachrome slides which are virtually
indistinguishable (by eye) from slides taken yesterday. Good enough
for me.

Of course they are - and all those years developing and improving the
Ektachrome process has all been in vain.

Cut the BS. 40 years ago Ektachrome wasn't anything like as good as it
is today. For one thing, Ektachrome of that era was renowned for its
excessively blue shadows which, thankfully, Kodak fixed around the
mid-80s . So even if 40 year old Ektachrome slides did survive unfaded,
which *is* possible if they were stored at low temperature in a dry
environment, they certainly would not be "indistinguishable from slides
taken yesterday" - simply because the emulsion was nowhere near as good
as it is today. If it was, Kodak would never have gone through however
many iterations of the E-x process in that time that they have.

 

[Ole-Hjalmar Kristensen]

The main point was that it has survived unfaded.

And of course old Ektachrome has excessively blue shadows, but that
was not the point. I also have lots of slides which *were* taken last
year with nasty colors, which also give me the same problems when
scanning. I usually get better colors when scanning negative film,
btw.

KM> Of course they are - and all those years developing and improving the
KM> Ektachrome process has all been in vain.


KM> Cut the BS. 40 years ago Ektachrome wasn't anything like as good as
KM> it is today. For one thing, Ektachrome of that era was renowned for
KM> its excessively blue shadows which, thankfully, Kodak fixed around the
KM> mid-80s . So even if 40 year old Ektachrome slides did survive
KM> unfaded, which *is* possible if they were stored at low temperature in
KM> a dry environment, they certainly would not be "indistinguishable from
KM> slides taken yesterday" - simply because the emulsion was nowhere near
KM> as good as it is today. If it was, Kodak would never have gone
KM> through however many iterations of the E-x process in that time that
KM> they have.

KM> --
KM> Kennedy
KM> Yes, Socrates himself is particularly missed;
KM> A lovely little thinker, but a bugger when he's pissed.
KM> Python Philosophers (replace 'nospam' with 'kennedym' when replying)

 

[Kennedy McEwen]

The main point was that it has survived unfaded.

How are you so sure? You have said your assessment is by eye, so you
have no firm data to back your statement. You acknowledge the
limitations of the emulsion of the time, so how do you know it has not
degraded? The Ektachrome emulsion of that time is known to fade, even
in dark storage, so what makes you so confident that your slides
haven't? What special precautions have you taken, because unless you
stored them under archive conditions by accident (since the ideal
conditions were not even identified at the time your storage started)
then they certainly will have faded - even if you *think* they look OK.

Or do they just look like you recall they should? It is surprising how
many people believe that colours are so much more brilliant these days
than they were in the 50s and 60s - because that is the diet of images
from that period that they are fed on.

 

[Ole-Hjalmar Kristensen]

KM> How are you so sure? You have said your assessment is by eye, so you
KM> have no firm data to back your statement. You acknowledge the

I *said* it was by eye, so I don't really need any firm data to back
it up, do I? Yes, it has probably degraded, but not so much as to make
much of a difference to my eye, and there is no problem at all
scanning it, and the histograms indicate there is plenty of
information in all three colors. I have negatives from the same period
which now are essentially monochrome, though, stored under identical
conditions in the dark. The color rendition is certainly not
completely accurate, but then it never was.

KM> limitations of the emulsion of the time, so how do you know it has not
KM> degraded? The Ektachrome emulsion of that time is known to fade, even
KM> in dark storage, so what makes you so confident that your slides
KM> haven't? What special precautions have you taken, because unless you
KM> stored them under archive conditions by accident (since the ideal
KM> conditions were not even identified at the time your storage started)
KM> then they certainly will have faded - even if you *think* they look
KM> OK.

I mainly take pictures to look at them, so if they *look* good, they
*are* good. As I said before, good enough for me

KM> Or do they just look like you recall they should? It is surprising
KM> how many people believe that colours are so much more brilliant these
KM> days than they were in the 50s and 60s - because that is the diet of
KM> images from that period that they are fed on.

If I place these slides on the light table, side by side with other
slides taken within the last few years, lots of the newer slides have
more brilliant colors. I can also find quite a few examples of the
opposite. As you said yourself in a previous post, the Ektachrome
emulsion has been improved, so it is difficult to know which
differences can be attributed to aging, and which are just the
consequence of using another emulsion.


KM> --
KM> Kennedy
KM> Yes, Socrates himself is particularly missed;
KM> A lovely little thinker, but a bugger when he's pissed.
KM> Python Philosophers (replace 'nospam' with 'kennedym' when replying)

 

[Peter D]

How are you so sure? You have said your assessment is by eye, so you have
no firm data to back your statement.

Of course he does. He has his own impressions form his own eyes looking at
the slides. He may not have any empirical data -- other than that of a
witness --- but he certianly has sufficient data to make his opinion worth
consideraiton. Unless you know he is delusional, a liar, or incapable of
properly assessing the evidence of his own eyes, his statement is stronger
than yours.

 

[Hecate]

Of course he does. He has his own impressions form his own eyes looking at
the slides. He may not have any empirical data -- other than that of a
witness --- but he certianly has sufficient data to make his opinion worth
consideraiton. Unless you know he is delusional, a liar, or incapable of
properly assessing the evidence of his own eyes, his statement is stronger
than yours.

Actually, no it's not. Real data measurements will always beat "eye"
measurements because "eye" measurements are both subjective and, the
farther back you go, rely on more and more tenuous memory. Can you
remember what a slide you took 50 years ago looked like then?

--

Hecate - The Real One
(e-mail address removed)
Fashion: Buying things you don't need, with money
you don't have, to impress people you don't like...

 

[Kennedy McEwen]

KM> How are you so sure? You have said your assessment is by eye, so you
KM> have no firm data to back your statement. You acknowledge the

I *said* it was by eye, so I don't really need any firm data to back
it up, do I?

That is the point I was making - you are relying on your eyes and memory
as the basis of an objective assessment. An image can *look* quite
acceptable even after substantial level of fading and it is only when
you do a side by side comparison of the original that the loss becomes
obvious. You only need to look at the Usenet records of the Epson/Canon
inkjet orange fade issues for examples of that - literally thousands of
owners making claims that these inks were perfectly stable, only to
rescind them a few days or weeks later after making proper comparative
tests.

As you said yourself in a previous post, the Ektachrome
emulsion has been improved, so it is difficult to know which
differences can be attributed to aging, and which are just the
consequence of using another emulsion.

Precisely, and since you can't tell, how can you be sure that your
images are "virtually indistinguishable" from recent slides? The fact
is you can't. Neither can you determine, by comparing an eye viewed
image with your memory, both modified by cultural expectations of the
images from that period, how much fading has actually occurred. And if
you cannot do that, how can you possibly determine how close to an
objectionable level of fading your examples actually are - or even if
they have actually crossed that level already? How much longer can you
be confident that your original images on Ektachrome slides will
actually last in an acceptable form? Viewing "by eye" against a the
reference of a filtered memory certainly will not yield any of those
answers.