Profiling 35mm film scanner with IT8 targets

[DenverDad]

I am trying to decide whether or not to invest in calibration targets
to profile a dedicated film scanner, as part of a move to a
color-managed workflow. So, I was wondering if anyone would care to
comment on their experiences profiling a film scanner this way. I
realize of course that the subject comes up fairly regularly on this
(and other) forums, and I have read as much as I could find. But I
want to convince myself it's going to really make a difference before I
purchase the targets (probably the "everything" pack of IT8 targets
from Wolf Faust). So I guess you could say I'm taking sort of a poll
on the subject.

It occurs to me that there may be a lot of variables involved;
including the specific scanner being used, the effectiveness of
whatever built-in profile the manufacturer offers; how old the scanner
is, how much use it has seen, or how stable it is over time; variation
from unit to unit; the software which is being used to actually do the
profiling (for example, Vuescan) and probably how "particular" the
given user is when it comes to discerning color differences. And then
too, any errors or misunderstandings of how to use color spaces and
profiles could probably blow it all out of he water anyway (possibly
explaining reports of any unsatisfactory results?)

My specific scanner is a Canoscan FS4000US, and I will likely use
Vuescan to profile it. So any advice from that particular combination
would be especially useful!

Thanks,

Jeff

 

[Djon]

It'll be interesting to see responses. One thing you said is especially
interesting:

" and probably how "particular" the
given user is when it comes to discerning color differences."

This isn't just a matter of "particular"...I've noticed that few have
the basic photographic skillset with which to meaningfuly discuss
color visually, so you'll notice that it's rarely done..it can't be
done without visually quantifying according to the standard RGB/MCY
terminology that custom photolab technicians have long utilized..

eg "this print needs 5Y 10M" would get very close agreement from a
dozen skilled printers, evaluating the same cyan/blue print WITHOUT the
use of any technology.Very few digital printers seem aware that this
skillset is common among skilled traditional color photographers and
custom lab printers...they're illiterate visually.

 

[Jim]

I am trying to decide whether or not to invest in calibration targets
to profile a dedicated film scanner, as part of a move to a
color-managed workflow. So, I was wondering if anyone would care to
comment on their experiences profiling a film scanner this way. I
realize of course that the subject comes up fairly regularly on this
(and other) forums, and I have read as much as I could find. But I
want to convince myself it's going to really make a difference before I
purchase the targets (probably the "everything" pack of IT8 targets
from Wolf Faust). So I guess you could say I'm taking sort of a poll
on the subject.

It occurs to me that there may be a lot of variables involved;
including the specific scanner being used, the effectiveness of
whatever built-in profile the manufacturer offers; how old the scanner
is, how much use it has seen, or how stable it is over time; variation
from unit to unit; the software which is being used to actually do the
profiling (for example, Vuescan) and probably how "particular" the
given user is when it comes to discerning color differences. And then
too, any errors or misunderstandings of how to use color spaces and
profiles could probably blow it all out of he water anyway (possibly
explaining reports of any unsatisfactory results?)

My specific scanner is a Canoscan FS4000US, and I will likely use
Vuescan to profile it. So any advice from that particular combination
would be especially useful!

Thanks,

Jeff

Profiling tools create a profile for a particular scanner using a particular
IT8 target. The target shows a series of patches, etc., whose LAB values
are furnished by Kodak. A particular batch of targets have a particular
description file containing those values. The purpose of the profiling tool
is to adjust the results of the scanner to agree with the LAB values of the
target.

The Kodak IT8 target applies to E6 film. You get targets from other types
of film from Wolf Faust; I have never used his products, and I certainly do
not know how well they work.

To accomplish this task, you turn off all color management in the scanner;
hence whatever profile the manufacturer provides is not used.

I used Monaco to create a profile for my Nikon LS-40, and the results were
very good indeed. I would expect the same from the Canon.
Jim

 

[bmoag]

I use this scanner, although less and less as dSLR use predominates, in a
color managed workflow. It is an old but useful device.

Unless you have specific professional needs then using calibrated targets
with this scanner for high end amateur work is essentially pointless. Even
with a calibrated transparency target you would have to use the same
emulsion and recalibrate prior to every session. Even then there are so many
variables that the process would not likely be worth the effort.

My recommendation would be to obtain the best scan of a transparency or
negative that you can using whatever software you prefer. My usual method is
to make minimal adjustments in the scanning software. Suffice it to say that
I find the Canon software superior for color negative materials and Vuescan
superior for transparencies.

Color management begins once the image is scanned.

I now use the Monaco optix system because it allows for accurate monitor
calibration and also for the generation of custom printer/paper profiles
that work for me somewhat better than the canned profiles Epson provides for
its papers and printers. I presume you are using a quality photoprinter and
a color managed photo program and understand how to implement color
management properly in the printer driver. I presume you understand that the
differences between what you see on a monitor and what you will see in the
final print using the best calibration techniques will differ somewhat
because of the physical differences between the media.

If your monitor is properly calibrated and your color eye sight reasonably
reliable, which is not the case more often than people (men) realize, this
is all you require to produce stunningly high quality color prints. If your
vision is color impaired and you want an absolute color workflow independent
of subjective judgement I suggest you change to an all digital system as
then the camera can be calibrated as well.

 

[pawelb]

I am trying to decide whether or not to invest in calibration targets
to profile a dedicated film scanner, as part of a move to a
color-managed workflow. So, I was wondering if anyone would care to
comment on their experiences profiling a film scanner this way. I
realize of course that the subject comes up fairly regularly on this
(and other) forums, and I have read as much as I could find. But I
want to convince myself it's going to really make a difference before I
purchase the targets (probably the "everything" pack of IT8 targets
from Wolf Faust). So I guess you could say I'm taking sort of a poll
on the subject.

I described my experience with scanner profiling some time ago on
photo.net - I hope you don't mind if I quote it here. My experience
refers to colour slide scanning only:

--- begin ---

Reading your conversation I thought I'll add my two cents. I also use
Minolta DSE 5400 (first version) with Vuescan and I am in no way a
professional scanner operator however I spent some time profiling my DSE
5400 a couple of weeks ago. To evaluate color fidelity I was comparing
my slides put on a lightbox with the same slides scanned and displayed
on a monitor calibrated and profiled with hardware calibrator.

First, when using the scanner with built-in profile of Vuescan and color
balance set to neutral I noticed my scanned slides had consistently
red/magenta cast - very annoying as these were from my winter holidays
with lots of snow that looked pink.

After reading for a while on the subject I bought an "Agfa IT8 Color
Reference" target set and began to profile the scanner using Vuescan.
The results using Vuescan produced ICC profile and color balance set to
neutral were better, however the cast was not gone completely and when
comparing the slides on a lightbox with the scans I still could see a
difference.

Together with my IT8 sets I got for free a long discontinued profiling
software - Agfa Colortune 3.02 (now Agfa is selling only its
professional version costing several thousand euros). As the software
was over 5 years old and I was using the latest and newest version of
Vuescan I didn't even install Colortune. However, seeing still not
perfect results with Vuescan I decided to give Colortune a try.

I scanned an IT8 target with Vuescan (with color balance set to none or
neutral and ICC setting set to default "built-in" profile) and then
generated scanner ICC profile using Colortune. I tried to set this
profile as a scanner ICC profile in Vuescan, but it didn't work so I
rescanned my slides using exactly the same Vuescan settings I used when
scanning the target and then attached my Colortune generated ICC profile
to the TIFF files produced by Vuescan. And... it made all the
difference! My slides on a lightbox looked now _exactly_ the same as the
scans on a monitor! On some slides only minor brightness and contrast
adjustments were neccessary - the color was spot on.

What I learned afterwards is that Vuescan produces very small,
simplificated ICC profiles of inferior quality - so called matrix
profiles (mine was only about 500 bytes long!). Agfa's Colortune instead
produced a "complete" ICC profile with size over 600 kB containing
detailed color lookup tables (CLUTs). These tables are used by photo
processing software to interpret and display the colors outputted by
Vuescan and they describe the scanner behaviour much better than matrix
profiles produced by Vuescan. Of course, the big Colortune profile
doesn't have to be attached to the TIFFs indefinitely - after using the
profile as an input profile I'm free to convert my TIFFs to whatever
color space I wish and forget about the scanner profile.

So, if you want maximum color fidelity for your scans you should give
some serious scanner profiling software a try. Agfa's Colortune is not
an option I'm affraid as it is discontinued but try a free XLProfiler or
cheap Profile Mechanic - Scanner from Digital Light and Color
(www.dl-c.com).

--- end ---

I would add some thoughts to the text above.

First, there is scanner profiling and film profiling possible. Scanner
profiling is when you want your scans to resemble an _image_ _on_ _a_
_lightbox_ in every possible way (it is like scanning media type "image"
in Vuescan). Film profiling is necessary when you want your scans to
look like an _original_ _scene_ photographed (taking into account how
particular film handles different lighting situations and so on - this
is like media type "slide film" in Vuescan).

If your aim is to make scans that would look as close as possible to
what you see on a (good) lightbox then - contrary to what bmoag wrote in
this thread - you don't have to profile your film nor scanner every time
you scan a new roll - you can profile your scanner once a month (like
your monitor) and it will suffice.

Also, if you are not after film profiling then you can use whatever
transparency target you have available and your results will be fine
(you don't have to match emulsion types because you don't try to recover
an original scene but the image on a lightbox). Actually, my Agfa
targets are Agfa RSX if I remember correctly - a slide film I hardly
ever used - I shoot mostly Fuji - and the results are good.

So, my experience with scanner profiling is very positive but the key
question is: why do you want to profile your scanner? In my case the
reasoning behind profiling my scanner was I wanted my scans to resemble
original slides as much as possible - hence the calibrated monitor and a
lightbox as a tools to evaluate results. Now when I say that now I am
content with the results I get using Colortune profiles it doesn't mean
the images I get are perfect but rather that the colours on the scans
are identical to those on film (this is done without _any_ color
correction in Vuescan) - even if it means that all exposure and colour
cast errors from the slides are visible on the scans.

If this is kind of result you are after - definitely go for it and
profile your scanner. In my opinion however you should look for a better
profiling software then Vuescan.

 

[pawelb]

I use this scanner, although less and less as dSLR use predominates, in a
color managed workflow. It is an old but useful device.

Unless you have specific professional needs then using calibrated targets
with this scanner for high end amateur work is essentially pointless. Even
with a calibrated transparency target you would have to use the same
emulsion and recalibrate prior to every session. Even then there are so many
variables that the process would not likely be worth the effort.

I think you are referring here to film, not scanner profiling. If your
aim is not to make scans that look exactly like an original scene
photographed but rather scans that look like an image on a lightbox then
no film profiling is necessary and scanner profiling will suffice.

Scanner profiling is not a difficult operation and once you have a
target set, a profiling software and your workflow set it can be done
once a month or so and results should be good - providing your aim is to
get scans looking like slides on a lightbox.

 

[DenverDad]

First, there is scanner profiling and film profiling possible. Scanner
profiling is when you want your scans to resemble an _image_ _on_ _a_
_lightbox_ in every possible way (it is like scanning media type "image"
in Vuescan). Film profiling is necessary when you want your scans to
look like an _original_ _scene_ photographed (taking into account how
particular film handles different lighting situations and so on - this
is like media type "slide film" in Vuescan).

If your aim is to make scans that would look as close as possible to
what you see on a (good) lightbox then -

Yes that is basically the goal - matching what's on the film, not
necessarily "correcting" the way the film rendered the scene.

Also, if you are not after film profiling then you can use whatever
transparency target you have available and your results will be fine
(you don't have to match emulsion types because you don't try to recover
an original scene but the image on a lightbox). Actually, my Agfa
targets are Agfa RSX if I remember correctly - a slide film I hardly
ever used - I shoot mostly Fuji - and the results are good.

I hesitate commenting on this because I know it may derail my own topic
- but that's the way I always thought of it too. Namely, that ANY
calibration target should be usable if the goal is to profile the
scanner (as opposed to profiling and correcting for specific film
emulsions). I have never understood all the explanations for why you
HAVE to have a target for a specific film if you are only profiling the
scanner.

 

[Bart van der Wolf]

SNIP

I have never understood all the explanations for why you
HAVE to have a target for a specific film if you are only
profiling the scanner.

For scanner profiling it is 'only' necessary to profile the
interaction between the Scanner's lightsource, the film's dye-set (!),
and the CCD or other sensor's color filters (if any). So you only need
a target per film dye-set (e.g. all Ektachrome variations use the same
dye-set).
Wolfgang Faust's targets are high quality, use a large(r) part of the
available saturation, and are decently priced. I've had good
experiences with his products and prompt service.

Particular/individual film response to exposure/processing conditions
could be approximated with a special profile, but it's only practical
if you have control over the lightsource(s) used for shooting and use
identical (same run) processing.
It is typically accomplished by shooting a reflective target on
(single batch, same storage/usage conditions) film (or Digicam
sensor), and by using that session's profile only for the images shot
at the same session. Even then, color balancing will probably be
needed to make the image look more appealing.

Bart

 

[Stewart]

I used to have a 4000US (traded it in on a 9000) and I profiled it using
Wolf Faust targets and Vuescan. There was an improvement - noticeable
but not huge, which suggests that it was not doing too badly on the
default settings. For me the biggest single improvement I made to my
results was getting dedicated profiles for my Epson 1290 printer.

Stewart

 

[DenverDad]

Frankly, mine does a pretty good job with the built in profile too, at
least most of the time. So I'm not expecting any kind of huge
improvement either. Rather, just wondering if I can get a little
closer on the color in the intial scan and maybe with greater
consistency.

So how do you like the Nikon? That's a switch I've considered making
myself. Are you scanning 35mm with it or MF (or both)? Judging by
everything I've read, the 9000 should be getting you quicker scans, and
greater dynamic range (as seen in the shadows). Any other improvements
or differences you can report relative to the FS4000? And, in
particular have you had any issues with the focus depth of field?

Jeff

 

[Stewart]

Jeff

I bought it because I got into MF, and given that, it was a good buy.
Leaving that aside, it is slightly better than the FS4000 but not
breathtakingly so. Compared to the FS4000 the 9000 shows a discernable
improvement in noise (ie a single-pass scan on the 9000 is like an
8-pass scan on the FS4000, and no worries about registration as the 9000
does multisampling on a single pass). DoF is worse than the 4000 - less
than 1mm in my estimation and you therefore need to be very careful. I
think Nikon is derelict in not selling it with a glass holder. Or they
should have given it a narrower aperture and not achieved their scan
speed by compromising DoF.

Purely on my subjective response to the quality of the scans and prints,
I'd say stick with the FS4000 if you are only into 35mm. The extra money
does not buy you a proportional increase in quality.

I should qualify that by saying that I used Vuescan with both. VS gives
significantly better results than Canoscan, and only slightly better
results than Nikonscan. So if you are not using VS then it might prove
to be a cheaper option!

Happy new year to all from southeast Australia (where it is currently 38
degrees Celsius).

Stewart

 

[Roger S.]

Hi Jeff,
I use the Wolf Faust Provia target with the FS4000US and recommend it.
I only use Provia and Sensia, so I don't need multiple targets, and for
the occaisional times I have used Elitechrome 100, the profile was
reasonably close.

You can get good results using the VS default profile setting and
"white balance" but "neutral" comes out very reddish for me. Custom
profiles fix this to a good extent.

Roger

 

[Roger S.]

Bmoag wrote:
"Unless you have specific professional needs then using calibrated
targets
with this scanner for high end amateur work is essentially pointless.
Even
with a calibrated transparency target you would have to use the same
emulsion and recalibrate prior to every session. Even then there are so
many
variables that the process would not likely be worth the effort. "

As an enthusiastic amateur, I respectfully think that you are
overstating the difficulty. I've noticed almost no drift with my Canon
scanner since first profiling it last March, meaning that the August
profile is nearly identical to the March one. Using a calibrated
target gives you slide scans that are much closer to the original than
you would get with no profile and this is true using a Fuji Provia
target with Provia and Sensia emulsions processed by Fuji, A&I, my
local lab, and labs in Japan. White balance may work fairly well in
Vuescan, but it will change the color balance of the original in some
cases.

There are ways to "color manage" your negative output as well to
largely let you see what's on the film and Erik Krauss has some posts
on different techniques in the archives.

My monitor is a CRT calibrated with the $60 Eye-One and it's good
enough to give me consistent output to the Fuji Frontier. My goal
isn't to take subjective judgement out of the loop but to get a more
consistent starting point to minimize the amount of color correction
needed.

 

[DenverDad]

Hi Jeff,
I use the Wolf Faust Provia target with the FS4000US and recommend it.
I only use Provia and Sensia, so I don't need multiple targets, and for
the occaisional times I have used Elitechrome 100, the profile was
reasonably close.

You can get good results using the VS default profile setting and
"white balance" but "neutral" comes out very reddish for me. Custom
profiles fix this to a good extent.

Roger

Well, I can report that finally I just cracked down and bought the
things (Wolf Faust target set special, to be exact). They're "in the
mail" I'm told, and I can't wait to try them out ! I think I will see
how the Vuescan profiling goes first. I understand that the "matrix"
profile may not be as good for the color rendition as a full blown
(normal) profile. On the other hand, the idea of the profile being
exposure independent sounds like a big plus considering my scanning
workflow. We'll see how it goes.

Roger, my experience has been similar - the default profile with white
balance often does a remarkably good job, while "neutral" is more
spotty; sometimes close to perfect, but then othertimes it approaches
funkadelic! Of course there are times when the white balance setting
is just the wrong transformation too.

As a side note, Roger, did you ever resolve the more recent IR-clean
issues you were seeing with VueScan? I'm sorry I haven't had time to
make up some good raw scans and send them to Ed, as you requested. My
experience of late has been that as long as I don't have any
particularly large goobers, then I don't really see the effect. In any
event I'm hoping that by now this one has been resolved.

Jeff

 

[Roger S.]

Hi Jeff,
No luck yet. I have been away and haven't had the chance to do any
more testing. I went back to an older version of VS that doesn't have
these problems. If you have a raw scan lying around, send it on to Ed.
I'll eventually get around to it. Otherwise 8.3.01 is working
bug-free for me at the moment (now that my transparencies are all in
strips).