Making Scanner Profiles from IT8 Targets, Coloraid vs. Colour Science

[ThomasH]

Recently I have discovered an alternative source of the IT8 targets
to the well known and respected http://www.targets.coloraid.de/.

I was in the quest to find profiles for Agfa slide material and
to resolve the problem which I had with scanning negative material.
Especially all my ISO 800 and 1600 shots are made on negatives and
frequently produced very ugly color balance and lots of noise in
shadows. Wolf's delay in making Agfa targets is immense, the A3
target was postponed for several years already.

Very recently I stumbled across a Swiss company Colour Science.

http://www.colour-science.com/

They have a different approach in making calibration targets. If I
may a bit speculate, my impression is that they do not try to achieve
the absolute perfection in color projection. They rather seem to
follow a pragmatic practitioner way. They expose regular material
and provide a large number of calibration strips for both negative(!)
and positive material. They provide film strips, as opposed to
mounted slides produced by Wolf's Coloraid. On these strips they
expose the IT8 pattern with some +/- steps, so that you can make a
target for over/under exposed images as well and try to salvage them
using such profile. An example film strip is depicted here:

http://www.colour-science.com/quality test tools/IT87negatives/IT87_neg_productlist.htm

Furthermore, while Wolf provides a charge data for each of his
targets, Colour Science offers only one charge, common to all
targets. But while Wolf provides one target for a family of films,
Colour Science makes separate targets for each of the materials
in a film family, pro and consumer alike. E.g. there is a target
for Sensia 100, 200 and 400. Naturally, since the consumer material
may have deviation in color balance, so can every target for consumer
class film. This is probably something for what Wolf will not stand
for, if I may draw this conclusion based on his struggle with
achieving a stable Agfa RSX-II profile.

Being a hobby photographer, I do not seek the absolute colorimetric
truth, I wanted be helped in obtaining reliable, close to correct
results. And I did. I love these profiles in the practice! I scanned
5 different calibration strips and Vuescan made *.icc profiles from
these scans without a problem. The result is very good: For the
first time I can scan negatives, like the Superia 800XT or Agfa
Vista 800, without making some wild and radical adjustments of the
black point and without countless tweaks of color channel balance.
The settings in Vuescan are now similar across all materials.

A real mprovement in Agfa slide scanning comes from the CT-100 Agfa
target. Following Wolf's recommendation I used so far the Velvia
profile for my CT-precisa and RSX-II slides. It is astonishing how
different are the histograms of scans made with the Agfa profile
versus with Coloraid's Velvia profile. It appears to me that Agfa
has a significantly lower dynamic range and that its black dyes are
much less dense compared to (say) Velvia or Provia.

However, once adjusted to a similar brightness level, corrected
black point etc., the result is optically difficult to distinguish,
and certainly I cannot make a "quality" judgment by looking on the
results and trying to guess which was made which way. With the
Velvia profile one must use higher black point setting and than
sometimes Agfa's shadows appear dark bluish.

If someone of you made also experiences (or maybe even) measures
of these target, I would like to share this experience. My
recommendation is: Get some of these targets and give them a try.

Thomas

 

[Erik Krause]

Recently I have discovered an alternative source of the IT8 targets
to the well known and respected http://www.targets.coloraid.de/.

I was in the quest to find profiles for Agfa slide material and
to resolve the problem which I had with scanning negative material.
Especially all my ISO 800 and 1600 shots are made on negatives and
frequently produced very ugly color balance and lots of noise in
shadows. Wolf's delay in making Agfa targets is immense, the A3
target was postponed for several years already.

Very recently I stumbled across a Swiss company Colour Science.

http://www.colour-science.com/

They have a different approach in making calibration targets. If I
may a bit speculate, my impression is that they do not try to achieve
the absolute perfection in color projection. They rather seem to
follow a pragmatic practitioner way. They expose regular material
and provide a large number of calibration strips for both negative(!)
and positive material. They provide film strips, as opposed to
mounted slides produced by Wolf's Coloraid. On these strips they
expose the IT8 pattern with some +/- steps, so that you can make a
target for over/under exposed images as well and try to salvage them
using such profile.

The dynamic range from these targets is very limited, since they seem
to be shot, not laser exposed. Wolf Faust targets are laser exposed and
cover the whole dynamic range of the respective slide film without need
of under- or overexposed samples.

Providing different targets for different exposures means you never get
the whole dynamic range from one profile. You get at least clipping
but more likely colors outside the target dynamic range will be far
off.

Furthermore negative color depends on processing. Just compare the film
base color values from different rolls of the same film brand. They
differ a lot. According to my favourite lab the film base color depends
not only on the chemicals, their age and temperature but also on the
age of the film and the conditions it was stored (humid, dry, hot,
cold).

It is a far better idea to shoot a grey card together with some white
and black object to have the ability to do a simple white- neutral- and
black adjustment. you can do this for any lighting condition and get
the ability to adjust not only neutral daylight, but different
incandescent and flourescent light, too - with the typical colors of
your lab included.

I experimented with the coloraid targets shot on negative film and got
exactly that effect. It depends on the profiling software whether the
colors outside are simply clipped or whether you get some weird
psychedelic results...

The best resuls I got so far was from a slide target shot on color
negative film. But even this technique does not provide the whole
dynamic range. However, it should be sufficient for all standard
photography work

 

[ThomasH]

The dynamic range from these targets is very limited, since they seem
to be shot, not laser exposed. Wolf Faust targets are laser exposed and
cover the whole dynamic range of the respective slide film without need
of under- or overexposed samples.

The dynamic range of the targets is... identical to the dynamic range
of any images on the same material. This argument escapes the logic.

May I ask, how do you know that Coloraid targets are laser exposed,
and why should this be considered an advantage over an exposure of a
reflective target? I would say: exposure of a target is precisely
reproducing the condition in which the film is being used! Great,
I like targets made this way, they should be able to provide
excellent profile. I do not care about the laser!

Providing different targets for different exposures means you never get
the whole dynamic range from one profile. You get at least clipping
but more likely colors outside the target dynamic range will be far
off.

You might be here correct in all aspects, but these +/- nstop targets
will be in use for extreme rare cases only, in which you would like
to salvage an over/uder exposed image. In such case the damage is
done already, with or without the target. I have not applied any
of these +/- nstep profiles yet.

I have a few such images in my archive, foremost cases in which
I use shift lens and forget to close apperture. I will make soon
a test if such targets provide really some level of help. Until
than I have no real idea of just how effective these +/- nstep
targets really are.

Furthermore negative color depends on processing. Just compare the film
base color values from different rolls of the same film brand. They
differ a lot. According to my favourite lab the film base color depends
not only on the chemicals, their age and temperature but also on the
age of the film and the conditions it was stored (humid, dry, hot,
cold).

It sure does, I know this as well. By writing this however you almost
make the impression that you disagree with use of scanner targets, but
below you write by yourself that you tried to expose Wolfs reflective
target using the specific negative material, and you consider this
'sufficient for all standard work.'

This is precisely what ColousSceince does, except that the exposure
occurs under controlled condition, which I am not sure how to create
and how to reliably reproduce by myself. You mileage may vary.

Exposing a neutral gray target on the shoot scene is an excellent
and well known method, but the majority of my images are legacy
and I do not have such exposure for the vast majority of them.

Thus facing a reality as it is, lets assume for the sake of argument
that many of us do not have any gray neutral spot memorized on each
negative roll available to setup the color balance.

It is a far better idea to shoot a grey card together with some white
and black object to have the ability to do a simple white- neutral- and
black adjustment. you can do this for any lighting condition and get
the ability to adjust not only neutral daylight, but different
incandescent and flourescent light, too - with the typical colors of
your lab included.

I experimented with the coloraid targets shot on negative film and got
exactly that effect. It depends on the profiling software whether the
colors outside are simply clipped or whether you get some weird
psychedelic results...

The best resuls I got so far was from a slide target shot on color
negative film. But even this technique does not provide the whole
dynamic range. However, it should be sufficient for all standard
photography work

....well this is what I did using these targets, ...I think.
Or, if you mean something else, I am not sure did you meant
to confirm the method or rather to suggest an alternate course
of action.

Thomas

 

[Erik Krause]

May I ask, how do you know that Coloraid targets are laser exposed,

....Wolf Faust told me...

and why should this be considered an advantage over an exposure of a
reflective target?

because you can profile the whole dynamic range in one go. BTW.: there
is a specification for this target and Wolf Faust is proud to be much
closer to that specification than Kodak.

I would say: exposure of a target is precisely
reproducing the condition in which the film is being used! Great,
I like targets made this way, they should be able to provide
excellent profile. I do not care about the laser!

You're kidding. Any natural scene has potentially more dynamic range
than a reflective target.

You might be here correct in all aspects, but these +/- nstop targets
will be in use for extreme rare cases only, in which you would like
to salvage an over/uder exposed image.

If the target would cover the whole dynamic range there would be no
need to use a different one.

I have a few such images in my archive, foremost cases in which
I use shift lens and forget to close apperture. I will make soon
a test if such targets provide really some level of help. Until
than I have no real idea of just how effective these +/- nstep
targets really are.

Try it. It might work somehow for slides but I see little chance for
negatives. My experience using profiles from low dynamic range targets
are not good at all. It depends on the profiling software whether you
get only clipping or very psychedelic colors. To avoid this you will
have to set hard black and white points to the dynamic range of the
target and use exactly the same settings for scanning.

It sure does, I know this as well. By writing this however you almost
make the impression that you disagree with use of scanner targets, but
below you write by yourself that you tried to expose Wolfs reflective
target using the specific negative material, and you consider this
'sufficient for all standard work.'

Yes. But the target is on the same roll of film as the images. Hence
the film base color and color casts due to processing and film age is
the same for all images. This will not be the case if the filme is
processed in a different lab.

I would think about using the color science targets if they would offer
that I can send them a film, they would expose the target and send it
back unprocessed.

This is precisely what ColousSceince does, except that the exposure
occurs under controlled condition, which I am not sure how to create
and how to reliably reproduce by myself. You mileage may vary.

Take your flash against a white wall or white sheet of paper. This
should give at least the same result for different shots. Flash should
be pretty standard daylight - see the specs...

Exposing a neutral gray target on the shoot scene is an excellent
and well known method, but the majority of my images are legacy
and I do not have such exposure for the vast majority of them.

Thus facing a reality as it is, lets assume for the sake of argument
that many of us do not have any gray neutral spot memorized on each
negative roll available to setup the color balance.

There will be future shots...

However, images from color negatives are guesswork, even if you get
them processed by a specialized lab. Only you can know how the colors
where while shooting. You will have to take the time and adjust them
manually if you are not satisfied with some automatics. Did you ever do
color chemical darkroom work?

 

[ThomasH]

...Wolf Faust told me...


because you can profile the whole dynamic range in one go. BTW.: there
is a specification for this target and Wolf Faust is proud to be much
closer to that specification than Kodak.


You're kidding. Any natural scene has potentially more dynamic range
than a reflective target.

I agree of course, its about the active versus passive light source,
yea! A natural scene can even make you blind from all the excessive
photons... And yet the film has its own dynamic range... Something tells
me it is so!

If the target would cover the whole dynamic range there would be no
need to use a different one.

You confuse here dynamic range, or some miracles in "increasing of
dynamic range of a film," with the problem at hand, what is a drastic
changed color balance of an overexposed image, especially of a slide.

Yes. But the target is on the same roll of film as the images. Hence
the film base color and color casts due to processing and film age is
the same for all images. This will not be the case if the filme is
processed in a different lab.

Great, I agree here completely. Now just tell me how to "put" an
IT8 pattern of (say) 700-800 already exposed rolls and I am happy
to use your method.

Otherwise, Earth calling Erik, I will stay with my doable compromise.
It provides results better than a guesswork from a neutral color
profile. Its over with drastic black point settings for negatives,
it is a tangible improvement.

I will barely have new rolls of film, film is over, I am just using
out what I have in the fridge. Other than that, its all about the
pixels and new problems: How to keep the sensor clean in DSLR.

Scanning was at the beginning a labor for love for me, later on
it only became a nuisance and significant time effort to get the
results right. I am glad it is going to its natural end.

Thomas

 

[Wolf Faust]

Very recently I stumbled across a Swiss company Colour Science.
http://www.colour-science.com/

Seems like they simply made shots of one of the coloraid.de C1 targets
on negative films and than shipp the reference file of that orignal C1
target (see the coloraid.de target reference files for charge R011220)
after changing the header info. I guess you are better of making this
yourself by simply using a C1 target with your actual films and your
lab... And even than, do not expect a reliable solution as there are
still lot's of unfixed problems...

Interistingly they also claim to produce IT 8.7/2 targets. But after
looking at the reference files, I can only say that they definitly do
not meet the IT 8.7/2 standard in several ways.

shadows. Wolf's delay in making Agfa targets is immense, the A3
target was postponed for several years already.

Yes, everytime I did start working on the Agfa films, something
happened causing delays. But as you can see from my website, the 4x5"
Agfa film targets have finaly become available last week and the 35mm
targets should follow soon. The density range is much bigger than the
original Agfa targets I have seen so far ;-)

Often the problem causing the delays were related to the lab not
developing the Agfa films correctly. My experience so far is, that
especialy with the handling of shadows Agfa RSXII films can differ
greatly between labs. Also, dmax can "jump" a lot. Sometimes really
frightening. But maybe this is just because the material I do have
from Agfa is a bit old (Agfa stopped the production of large format
RSX II films and I bought up their stock). With the limited films, I
guess I can only produce 20-30 4x5" targets... an expensive waste of
time and money considering the amount of work needed to adjust the
production devices.

The dynamic range from these targets is very limited, since they seem
to be shot, not laser exposed. Wolf Faust targets are laser exposed and
cover the whole dynamic range of the respective slide film without need
of under- or overexposed samples.

Yes, I did use lasers, but not anymore. Currently I use filtered light
sources that produces identical results (actually, I get a slightly
better dmax).

Furthermore negative color depends on processing. Just compare the film
base color values from different rolls of the same film brand. They
differ a lot. According to my favourite lab the film base color depends
not only on the chemicals, their age and temperature but also on the
age of the film and the conditions it was stored (humid, dry, hot,
cold).

I agree. The IT 8.7/1 method used for positive slide films can not be
easily adopted for negative films. Masking, lab and material problems
can be rather big and thus a negative film target will not give you a
reliable result. So it is better to do the things Colour Science did
yourself by making a shot of a target. And still this will not be a
perfect reliable solution....

However, a negative film target that makes full use of the density
range and color gamut should allow removing the scanner errors and a
correction based on that target should level the field for further
corrections and make them much easier. Producing such a wide gamut
target requires production methods similar to the IT 8.7/1 targets and
thus this is nothing you can simply do by making a normal shot of a
reflective target.

One should also not forget, slide film targets can be measured and
that measurement is the intended final color value (unless you take a
color appearance model into account... but this is a different
story...). You can measure a negative target, but how do you convert
that measurement into a final color value? The "Colour Science" target
simply ships the color values of the object (ie. C1 target) shot. For
a real negative calibration solution things will not be that easy.

Software is the other problem... one needs software that is specificly
designed for negatives. And in order to work in an optimal way, this
software already starts with the scanner driver. I haven't found much
scientific info on how exactly lab, masking and film aging affect the
colors. I intend to produce some negative targets with different labs
so that other developers have something to work with. Let's see what
we can find out. But I guess, in the end there will be some sort of
intelligent "guessing" needed based on the image content scanned in
order to get maybe 99% reliable results. Currently I only know some
high end devices able to get good negative film scanning results
without much user adjustments.

 

[Wilfred]

Seems like they simply made shots of one of the coloraid.de C1 targets
on negative films and than shipp the reference file of that orignal C1
target

Apparently they purchased targets from you to begin with - that's
interesting to know ;-)

 

[David Chien]

eh, long thread, but for those interested, Faust's color target worked
great for me on my Epson 1200S scanner, was super-cheap, did the trick
just perfectly, and works fine with the free tools at
http://www.coloraid.de/

 

[ThomasH]

Seems like they simply made shots of one of the coloraid.de C1 targets
on negative films and than shipp the reference file of that orignal C1
target (see the coloraid.de target reference files for charge R011220)
after changing the header info. I guess you are better of making this
yourself by simply using a C1 target with your actual films and your
lab... And even than, do not expect a reliable solution as there are
still lot's of unfixed problems...

Hi,

Interesting... So the values are identical! Actually I asked them
why they provide only one charge, but they are now off to PMA and
they did not replied yet.

Could you give us a hint or recommendation as to how to expose the
reflective target?

Interistingly they also claim to produce IT 8.7/2 targets. But after
looking at the reference files, I can only say that they definitly do
not meet the IT 8.7/2 standard in several ways.

Regardless all this, with the Velvia profile (or without any profile)
I never resolved this "blue dark shadow issue" with RSX or with Precisa.
Now I did, albeit while violating all sorts of standards

Lets take a look:

http://www.pbase.com/phototalk_thh/agfa_ct_it8_profile

As I said, while not seeking a Colorimetric exactness, I am happy with
a visual improvement, and my old, tired eyes are the still my preferred
lab-instrument at hand.


That's why in our private communications I expressed a hope that you
would maybe release such "less perfect" Agfa targets at "our own risk"!
Wolf, you are like the Leica company, you either release with precision
0.000001, not 0.0000015, god forbid, or you not release at all.


Anyway, while knowing that their charge has identical values to the
R011220, I reckon you are correct about the targets and of course,
I would love to shoot the targets by myself. The little pesky detail
is:

How to make a shoot setup? Any specific light source?
Or simply use daylight?

Yes, everytime I did start working on the Agfa films, something
happened causing delays. But as you can see from my website, the 4x5"
Agfa film targets have finaly become available last week and the 35mm
targets should follow soon. The density range is much bigger than the
original Agfa targets I have seen so far ;-)

Often the problem causing the delays were related to the lab not
developing the Agfa films correctly. My experience so far is, that
especialy with the handling of shadows Agfa RSXII films can differ
greatly between labs. Also, dmax can "jump" a lot. Sometimes really
frightening. But maybe this is just because the material I do have
from Agfa is a bit old (Agfa stopped the production of large format
RSX II films and I bought up their stock). With the limited films, I
guess I can only produce 20-30 4x5" targets... an expensive waste of
time and money considering the amount of work needed to adjust the
production devices.


Yes, I did use lasers, but not anymore. Currently I use filtered light
sources that produces identical results (actually, I get a slightly
better dmax).


I agree. The IT 8.7/1 method used for positive slide films can not be
easily adopted for negative films. Masking, lab and material problems

I also agree, and yet.... it helps and it brings good results
See below:

can be rather big and thus a negative film target will not give you a
reliable result. So it is better to do the things Colour Science did
yourself by making a shot of a target. And still this will not be a
perfect reliable solution....

However, a negative film target that makes full use of the density
range and color gamut should allow removing the scanner errors and a
correction based on that target should level the field for further
corrections and make them much easier. Producing such a wide gamut
target requires production methods similar to the IT 8.7/1 targets and
thus this is nothing you can simply do by making a normal shot of a
reflective target.

Indeed, what I made are scanner profiles, and they really help.

I hate negatives, they are almost not scannable, they cause an immense
amount of labor and the results are often miserable. Its just so that
over the years a number of them has accumulated, and seemingly using
a *scanner* profile simplifies the scan. With or without the profile,
the final color balance is achieved by manual tests and adjustments.
Needles to say, this does not provide any correct color mapping, its
subjective!

One should also not forget, slide film targets can be measured and
that measurement is the intended final color value (unless you take a
color appearance model into account... but this is a different
story...). You can measure a negative target, but how do you convert
that measurement into a final color value? The "Colour Science" target
simply ships the color values of the object (ie. C1 target) shot. For
a real negative calibration solution things will not be that easy.

Software is the other problem... one needs software that is specificly
designed for negatives. And in order to work in an optimal way, this
software already starts with the scanner driver. I haven't found much
scientific info on how exactly lab, masking and film aging affect the
colors. I intend to produce some negative targets with different labs
so that other developers have something to work with. Let's see what
we can find out. But I guess, in the end there will be some sort of
intelligent "guessing" needed based on the image content scanned in
order to get maybe 99% reliable results. Currently I only know some
high end devices able to get good negative film scanning results
without much user adjustments.

As always, its a pleasure to read the explanations!

Thanks for replying, I would be happy to hear more about a good
method to expose the reflective target.

Thomas