Elitechrome 100 Slide Scanning with Coolscan V ED

[Oliver Kunze]

Hi,

I am a owner of a Nikon Coolscan V ED. The scan quality is a big step
forward compared to my former 2500 dpi HP photosmart S20, but I have a few
questions. I take my slides with a contax/zeiss 35mmSLR equipment using
elitechrome 100 film.

1) It seems to me that there is no improvement in detail resolution when I
compare 2500 dpi HPS20 scans with 4000 dpi nikon scams. Is it possible that
2500 dpi is already sufficient for the ISO100 elitechrome?

2) The 4000 dpi (with ICE and GEM function on) in fact creates a grainy,
streaky structure in contrary to my old 2500dpi HPs20 scans, for example in
blue sky areas of my slides. Is that the already the elitechrome film grain
scanned by the nikon? I scanned the slides in glasless frames. I heard the
elitechrome is not suited for scanning purposes, but other films are. Is
that right, what are they? I also found in scanned B/W negative ISO 50 films
grainy, but not streaky structures, but I guess that must be noise or any
other disturbance and not the B/W film grain?

3) I found out that 48 bit of color depth improves quality compared to 24
bit, but handling the resulting 114 MB tiff scans is not very practical
(zipping does not reduce size). Neither Nikon Scan nor photoshop (ver. 5
limited edition) can save these files as jpeg. Is jpg only a 24 bit file
color depth file standard? What to do?


Oliver

 

[Bart van der Wolf]

SNIP

Is it possible that 2500 dpi is already sufficient for
the ISO100 elitechrome?

No, at 4000ppi it should be much better.

2) The 4000 dpi (with ICE and GEM function on) in fact creates a grainy,
streaky structure in contrary to my old 2500dpi HPs20 scans, for example in
blue sky areas of my slides.

Are you looking at 100% zoom?

SNIP

Neither Nikon Scan nor photoshop (ver. 5 limited edition)
can save these files as jpeg.

Convert to 24-bit first, or use "Save for web" in Photoshop.

Is jpg only a 24 bit file color depth file standard?

Yes. Use Photoshop to convert to 8-bit/channel, but only after
colorbalancing/tonescaling and sharpening.

Bart

 

[ThomasH]

Hi,

I am a owner of a Nikon Coolscan V ED. The scan quality is a big step
forward compared to my former 2500 dpi HP photosmart S20, but I have a few
questions. I take my slides with a contax/zeiss 35mmSLR equipment using
elitechrome 100 film.

1) It seems to me that there is no improvement in detail resolution when I
compare 2500 dpi HPS20 scans with 4000 dpi nikon scams. Is it possible that
2500 dpi is already sufficient for the ISO100 elitechrome?

At what resolution are you scanning with the Nikon? The
resolution can be adjusted... I use LS4000, by the way.

2) The 4000 dpi (with ICE and GEM function on) in fact creates a grainy,
streaky structure in contrary to my old 2500dpi HPs20 scans, for example in
blue sky areas of my slides. Is that the already the elitechrome film grain
scanned by the nikon? I scanned the slides in glasless frames. I heard the

I have problems with Kodak material scanned on the LS4000,
but since I almost never use Kodak, I am not very impacted.
Especially E200 was really, really fatal.

Do not forget to *switch off* ROC!! Its really important. ROC
helps only if you have really old slides with substantial
color fading.

Which level of GEM are you using?
My experiences with GEM are very positive, as a matter of fact
I never saw any comparable tool to GEM. My experiences with the
settings are:

GEM 1 for Provia 100, Velvia 50
GEM 2 for Sensia 100,200 Agfa Precisa CT100 and 200
GEM 2 or 3 for Provia 400, Sensia 400
GEM 3 was enough for negatives such as Agfa Vista 800 or Fuji Press 1600

elitechrome is not suited for scanning purposes, but other films are. Is
that right, what are they? I also found in scanned B/W negative ISO 50 films
grainy, but not streaky structures, but I guess that must be noise or any
other disturbance and not the B/W film grain?

B/W material based on silver halide cannot be scanned using
IR ICE dust/scratch remover. Have you switched off ICE?

3) I found out that 48 bit of color depth improves quality compared to 24
bit, but handling the resulting 114 MB tiff scans is not very practical
(zipping does not reduce size). Neither Nikon Scan nor photoshop (ver. 5
limited edition) can save these files as jpeg. Is jpg only a 24 bit file
color depth file standard? What to do?

Oliver

Why not practical? Its digital photography! We have untold gigs of
space and DVD's to archive the data! Keep 48bit scans! I do without
exception. I use Vuescan to process raw 48bit files, but for a really
fine job you probably need something like Picture Window or than
of course Photoshop! Anyway, 48bit is essential if you use slide
material with high dynamic range. Wolfgang Faust who makes IT 8
calibration targets wrote that Fujichrome has dynamic range of 4!

Thomas

 

[Alan Browne]

I have problems with Kodak material scanned on the LS4000,
but since I almost never use Kodak, I am not very impacted.
Especially E200 was really, really fatal.

Do not forget to *switch off* ROC!! Its really important. ROC
helps only if you have really old slides with substantial
color fading.

Which level of GEM are you using?
My experiences with GEM are very positive, as a matter of fact
I never saw any comparable tool to GEM. My experiences with the
settings are:

GEM 1 for Provia 100, Velvia 50
GEM 2 for Sensia 100,200 Agfa Precisa CT100 and 200
GEM 2 or 3 for Provia 400, Sensia 400
GEM 3 was enough for negatives such as Agfa Vista 800 or Fuji Press 1600

Any for E100GX and E100S? I have trouble scanning them, grain
aliasing perhaps?

Cheers,
Alan

 

[Don]

2) The 4000 dpi (with ICE and GEM function on) in fact creates a grainy,
streaky structure in contrary to my old 2500dpi HPs20 scans, for example in
blue sky areas of my slides. Is that the already the elitechrome film grain
scanned by the nikon?

Yes, it is film grain, and (ironically) it manifests how much better
Nikon is. All other scanners use a different light source which
"fuzzifies" the image so you don't see the grain. Nikon just shows you
what's on the film.

This is similar to someone needing glasses looking at a wall and
admiring its smooth surface. Then they put on glasses and discover all
the bumps, smudges and other imperfections...

One advice I read - but don't use myself - is to slightly de-focus
(click on manual focus and enter a value by hand). That should smooth
out the image somewhat and reduce the graininess.

Don.

 

[Oliver Kunze]

Which level of GEM are you using?
My experiences with GEM are very positive, as a matter of fact
I never saw any comparable tool to GEM. My experiences with the
settings are:

GEM 1 for Provia 100, Velvia 50
GEM 2 for Sensia 100,200 Agfa Precisa CT100 and 200
GEM 2 or 3 for Provia 400, Sensia 400
GEM 3 was enough for negatives such as Agfa Vista 800 or Fuji Press 1600

I used GEM level 3, ROC off. I further found out that ICE is responsible for
the "streaks" between the grain in my scans. Turning ICE off takes them
away, but for the price of dust marks in the scan, which is not a good
option.

Why not practical? Its digital photography! We have untold gigs of
space and DVD's to archive the data! Keep 48bit scans! I do without
exception. I use Vuescan to process raw 48bit files, but for a really
fine job you probably need something like Picture Window or than
of course Photoshop! Anyway, 48bit is essential if you use slide
material with high dynamic range. Wolfgang Faust who makes IT 8
calibration targets wrote that Fujichrome has dynamic range of 4!

Since what version number is Photoshop fully capable (filters etc.) of
handling 48bit? I understand elements 2.0 which came with my Coolscan is
not. Is PS 6, which is available at a fair price?

Thomas

Oliver

 

[Mac McDougald]

Since what version number is Photoshop fully capable (filters etc.) of
handling 48bit? I understand elements 2.0 which came with my Coolscan is
not. Is PS 6, which is available at a fair price?

Oliver

Many 16 bit editing functions have been added to Photoshop PS.
Elements won't even open a file at 16 bit.

Mac

 

[Sander Vesik]

Since what version number is Photoshop fully capable (filters etc.) of
handling 48bit? I understand elements 2.0 which came with my Coolscan is
not. Is PS 6, which is available at a fair price?

You probably want Photoshop CS. However, if youare not totaly set on photoshop
you could simply go here

rsb.info.nih.gov/ij
ij-plugins.sourceforge.net/plugins/imageio

and download imagej and image i/o plugin that will handle 16 bit (and 32bit & fp)
depth in tiff. You can then do lossy export as jpg from it.

 

[RSD99]

The URL you posted has apparently been 'mugged' by some form of redirection command on
your server. This one might work better for "the rest of the world."

http://sourceforge.net/projects/freeimage/

 

[Bart van der Wolf]

The URL you posted has apparently been 'mugged' by some form of redirection command on
your server.

SNIP

Both links work fine on my computer.

Bart

 

[ThomasH]

Any for E100GX and E100S? I have trouble scanning them, grain
aliasing perhaps?

Perhaps! I had so many problems with Ektachrome E200 that I dropped
this material log ago in favor of Agfa RSX200 or Sensia 200 or
Agfa Precisa CT200. I have also encountered problems in scanning
of Elitechrome 100, but I concluded back than that possibly Qualex
did a lousy job with development.

(I used to have Kodak mailers for E6 film, but all this changed
after someone recommended me Calypso in Santa Clara! Jeez, the
best place for E6 development was just 5 min away, and I was
oblivious to its existence!)

Which scanner are you using? Nikon uses LEd light, I suspected
this light cannot penetrate Kodak's celluloid (??) Just hunches,
but Kodak is slide is really difficult to scan.

Thomas

 

[ThomasH]

Since what version number is Photoshop fully capable (filters etc.) of
handling 48bit? I understand elements 2.0 which came with my Coolscan is
not. Is PS 6, which is available at a fair price?

Oliver

How about the http://www.dl-c.com/, Picture Window Pro!
Its all around 48bit processor!

Norman Koren uses this tool and he has also posted very
detailed users guide on http://normankoren.com/

I like his site a lot, I just wish he would be a bit
less "Canon focussed."

Thomas

 

[Alan Browne]

Perhaps! I had so many problems with Ektachrome E200 that I dropped
this material log ago in favor of Agfa RSX200 or Sensia 200 or
Agfa Precisa CT200. I have also encountered problems in scanning
of Elitechrome 100, but I concluded back than that possibly Qualex
did a lousy job with development.

(I used to have Kodak mailers for E6 film, but all this changed
after someone recommended me Calypso in Santa Clara! Jeez, the
best place for E6 development was just 5 min away, and I was
oblivious to its existence!)

Which scanner are you using? Nikon uses LEd light, I suspected
this light cannot penetrate Kodak's celluloid (??) Just hunches,
but Kodak is slide is really difficult to scan.

Minolta 5400 (Cold cathode flourescent).
I just scanned some E100G and it went very well, but S and GX are
not easy to scan.
Elitechrome 200 (consumer v. of Ektachrome 200) scans fine,
'though it does have grain ... but not in an unpleasing way.

Cheers,
Alan

 

[RSD99]

FWIW:
Corel PhotoPaint has some "48-bit" capabilities. I haven't "kept up" on the program
recently, but version 8 had better "48-bit support" than the version of PhotoShop that was
being sold at the time (IIRC ... version 5.x).

 

[ThomasH]

Minolta 5400 (Cold cathode flourescent).
I just scanned some E100G and it went very well, but S and GX are
not easy to scan.
Elitechrome 200 (consumer v. of Ektachrome 200) scans fine,
'though it does have grain ... but not in an unpleasing way.

I though so far that these problems with Ektachrome were intrinsic
to Nikon and their LED monochromatic light sources... Obviously
there is something troublesome about Kodak's celluloid.

Thomas

 

[Bart van der Wolf]

SNIP

I though so far that these problems with Ektachrome were intrinsic
to Nikon and their LED monochromatic light sources... Obviously
there is something troublesome about Kodak's celluloid.

I haven't seen any examples (maybe I missed them) but I wouldn't jump to
conclusions too fast. Many issues result from people cutting corners, e.g.
scanning at less than full resolution. Yes, they may save some scan time,
but they will regret it when they compare lower res scans with properly
downsampled ones. And a 4000 or 5400 ppi scan allows lots of downsampling
for many uses (not everybody makes posters of each scan ;-) ).
Some make the mistake to judge results at less than 100% zoom on screen,
which will create all sorts of aliasing artifacts.

Some links to some comparison examples would be appreciated, because an
image says more than a thousand words...

Bart

 

[Alan Browne]

I though so far that these problems with Ektachrome were intrinsic
to Nikon and their LED monochromatic light sources... Obviously
there is something troublesome about Kodak's celluloid.

Troublesome only in the digital sampling sense. Projected look fine.

Some time ago there were posts regarding scanners picking up air
bubbles (or something) in Fuji film base. Had never been a
problem before scanning. Fuji (reportedly) did something to
their process (or supplier of base) and you don't hear of the
problem anymore... (was prob. much ado about nothing in the first
place).

Nikon's 5000/9000, I believe, have s/w desigend specifically for
the Kodak grain (grain-aliasing?) problem.

Cheers,
Alan

 

[Alan Browne]

SNIP



I haven't seen any examples (maybe I missed them) but I wouldn't jump to
conclusions too fast. Many issues result from people cutting corners, e.g.
scanning at less than full resolution. Yes, they may save some scan time,
but they will regret it when they compare lower res scans with properly
downsampled ones. And a 4000 or 5400 ppi scan allows lots of downsampling
for many uses (not everybody makes posters of each scan ;-) ).
Some make the mistake to judge results at less than 100% zoom on screen,
which will create all sorts of aliasing artifacts.

I believe that issues such as grain-aliasing occur at the higher
res scans, not the lower.

 

[Kennedy McEwen]

I believe that issues such as grain-aliasing occur at the higher res
scans, not the lower.

A common, and completely wrong, misconception!

Grain aliasing, in fact aliasing of any subject matter, occurs because
the resolution of the sensing system exceeds the resolution which can be
supported by the sampling density, as defined by the Nyquist criteria.
For a fixed size sensor (in this case the individual sensing elements in
the CCD) and a given optical system the resolution is fixed. Thus
aliasing becomes more problematic at lower sampling densities because
this fixed resolution of the sensor is more likely to exceed the
resolution capable of being supported.

The reason for the common misconception that you have reiterated is that
lower resolution scanners (specifically *NOT* lower resolution scans
from high resolution capable devices) have lower resolution sensing
elements and optics. Consequently aliasing is less noticeable with such
scanners at low resolutions than it is with high resolution systems at
the same sampling density. In addition, certain features in the object
being scanned, in this case grain, cannot be resolved at all by the low
resolution scanners and therefore cannot alias. However, since they
clearly are resolved by the high resolution scanners, they alias even
more at low resolution scans than at high resolution ones made on such
systems.

The worst possible configuration is a low resolution scan made directly
from a high resolution scanner without the proper downsampling that Bart
alluded to.

 

[Alan Browne]

A common, and completely wrong, misconception!

Grain aliasing, in fact aliasing of any subject matter, occurs because
the resolution of the sensing system exceeds the resolution which can be
supported by the sampling density, as defined by the Nyquist criteria.
For a fixed size sensor (in this case the individual sensing elements in
the CCD) and a given optical system the resolution is fixed. Thus
aliasing becomes more problematic at lower sampling densities because
this fixed resolution of the sensor is more likely to exceed the
resolution capable of being supported.

The reason for the common misconception that you have reiterated is that
lower resolution scanners (specifically *NOT* lower resolution scans
from high resolution capable devices) have lower resolution sensing
elements and optics. Consequently aliasing is less noticeable with such
scanners at low resolutions than it is with high resolution systems at
the same sampling density. In addition, certain features in the object
being scanned, in this case grain, cannot be resolved at all by the low
resolution scanners and therefore cannot alias. However, since they
clearly are resolved by the high resolution scanners, they alias even
more at low resolution scans than at high resolution ones made on such
systems.

The worst possible configuration is a low resolution scan made directly
from a high resolution scanner without the proper downsampling that Bart
alluded to.

I get what you're saying, and it makes sense.

OTOH, when I scan at 5400 on E100S I get strong grain appearance
that does not show projected. The same film scanned at 1/4 the
res of the scanner does not exhibit the grain as strongly (and of
course nor does it show the detail). So pehaps grain aliasing is
the wrong term, but something is happening (or not happening)
resulting in the chunky looking grain. Other films (incl. GX,
Sensia, Velvia) do not show these artifacts...

Cheers,
Alan