Thanks Dave.
Well, I hope the mystery of violet casting will be resolved soon.
After numerous calibration and profiling efforts I recalled that after
acquiring the new monitor
I immediately started to scan the batch of slide featuring my Italian
vacation last year.
Moreover, I recalled also, that the slide film I used there was 21 roll of
Sensia II 100
outdated by one and half - two years at the time of purchasing. I bought it
intentionally knowing that it is outdated by the Fuji importer I bought from
stored it in the fridge all the time, so I hoped it will be fine and was
happy paying only about half price.
Bearing that in mind, I picked up my older slide storage album with similar
film but acquired new at the time and scanned several slides. No any unusual
cast !
I repeated that again and again for 5-7 slides from that bunch - nothing.
That almost insured me that I was really foolish relying on the outdated
film, although upon my arrival from the trip I enjoyed the slides projecting
them on the wall and didn't notice anything like color casts. Now I guess
this weird thing might come up in outdated emulsions even after being
exposed and processed, speeding up their aging process drastically.
Besides of that, I used LittleCMS software to create monitor's profile after
the monitor technical info I requested arrived from the manufacturer.
They provided me with phosphors x,y coordinates for RGB as well gamma
setting for those values.
The gamma they advised seemed me quite exaggerated being 3.06 but they
insisted then if I'll use their coordinates values for profiling I'll must
use this gamma.
I did it.
Afterwards, I defined this profile in NikonScan (as Custom monitor profile)
and had to change the RGB space to Scanner RGB instead of Adobe98. Similarly
I did in Photoshop defining this profile
instead of Adobe98. Overall, the results were not bad at all, however the
brightness looked exaggerated a bit (seemed like overexposure by 0.3 - half
stop).
(Would I leave Adobe98 in as RGB space in NikonScan or Photoshop it would
results in even more greater brightness (1-2 stops for my visual
estimation) ).
I suspect the excessive gamma might be the reason for this visual
overexposure, so I'll try to lower it creating new profile to see what will
happen.
Regards, Alex
-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Dave King
Sent: Tuesday, April 23, 2002 7:00 AM
To: alexz@zoran.co.il
Subject: [filmscanners] Re: Scanner calibration
I would think it's that simple Alex. All I did is select AdobeRGB as the
working space when installing Nikonscan and then point to my monitor
profile, and scan previews match Photoshop visually and numerically nearly
exactly.
Dave
----- Original Message -----
From: "Alex Zabrovsky" <alexz@zoran.co.il>
To: <kingphoto@mindspring.com>
Sent: Monday, April 22, 2002 12:29 PM
Subject: [filmscanners] RE: Scanner calibration
So Dave, do you mean this seems to be a matter of monitor profiling only ?
That means I have to create precise monitor's profile and specify one for
NikonScan CMS, right ?
Regards,
Alex Z
-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Dave King
Sent: Monday, April 22, 2002 4:19 PM
To: alexz@zoran.co.il
Subject: [filmscanners] Re: Scanner calibration
Without going to the other computer and actually looking (so I may be
forgetting something), Nikonscan only has two places for CM, monitor
compensation and output tag. AFAIK the output tag chosen cannot affect the
preview anyway, but the profile used for monitor compensation certainly
will. Considering this happened with the introduction of a new monitor, I
wonder about the quality of your monitor profile.
I haven't seen any misbehavior of Nikonscan's CM, the preview matches
photoshop exactly using PhotoCal/Monitor Spyder and AdobeRGB 98.
Dave
----- Original Message -----
From: "Alex Zabrovsky" <alexz@zoran.co.il>
To: <kingphoto@mindspring.com>
Sent: Monday, April 22, 2002 5:33 AM
Subject: [filmscanners] RE: Scanner calibration
The NikonScan is configured fro Adobe 98 RGB (Preferences/Color Management
tab) as well the Photoshop.
Well, yesterday evening, for my great disappointment, I revealed that the
strong bluish or to be more precise violet cast I reported previously has
gone just because once struggling with it I had to tweak analog gain on RGB
channels trying to compensate for it, then closing the application the Nikon
software remembered that analog gain setting and automatically loaded it
once the NikonScan has being re-launched, whilst I though it was cleared
off.
That was the reason of the cast disappearing, and not my pervious guesses
about removing monitor profile.
Once I cleared the analog gain settings to 0, the violet cast came back
:-(( driving me crazy again. Playing with that more, I figured out that
switching Nikon CMS off diminishes the cast greatly making it hardly
distinguishable, which means Nikon Color Management caused it with no
relation to monitor profiling at all !
Going further, I switched the Nikon CMS On again and set RGB to be Scanner
RGB instead of my regular Adobe 98 RGB setting in NikonScan, in order to
make the Nikon software not to use the
scanner's ICC profiles embedded into the software - that cured the problem
again removing the cast, but making certain impact on brightness (the image
is a bit darker then).
So I inferred that the reason for this violet cast is Nikon's ICC profile
embedded into their software. (According to the manual, implying Scanner RGB
space in Color Management/RGB tab will dismiss usage of Nikon's ICC profile
in the driver, without need to switch off the whole CMS)
What really confused me is that fact that a few weeks ago, scanning several
rolls of slides and negs with ordinary NikonScan configuration (Nikon CMS is
ON) I haven't noticed any unusual color casting over the image. The only
changed since then is the monitor...
Amy it still be the reason somehow ?
Regards,
Alex Z
-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Laurie Solomon
Sent: Sunday, April 21, 2002 7:54 PM
To: alexz@zoran.co.il
Subject: [filmscanners] RE: Scanner calibration
>The scans delivered to Photoshop seemed overexposed
Completely speculative, what is the Photoshop working color space set for?
Could it be that this working space is is the problem in that Photoshop is
translating the scan input into its working space which is being displayed
on the monitor while the scanner output that is not going through Photoshop
is being displayed directly in either sRGB or your monitor's default windows
or default custom profile?
-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of Alex Zabrovsky
Sent: Sunday, April 21, 2002 6:40 AM
To: laurie@advancenet.net
Subject: [filmscanners] RE: Scanner calibration
Well, must admit I cannot complain on the image quality delivered by my IV
ED using his default driver NikonScan 3.1.2. Nikon's CMS seems to work good
indeed.
However, I thought having the scanner calibrated precisely using the Q-60
TI8 slide and appropriate software might create even more precise results,
which I would like to compare with driver's native colors interpretation.
Frankly, this isn't something pressing me hardly, since the scanner already
deliver good results without his profile customization.
However I have another issue bothering me a lot.
A few days ago I changed my display purchasing new Iiyama VisionMaster 454.
It comes with his own setup software (INF file and ICC profile) which I
downloaded from their site and installed. into my system (Win98SE). Besides
of that, I tweaked the monitor's profile using Adobe Gamma utility
and installed it as working monitor's profile.
That was the starting point of my troubles. Once I did all that, the scanner
started to deliver
heavy bluish cast on slides which is easily distinguishable on both preview
and scan in Photoshop
and to correct it back I had either to play with analog gain feature on RGB
or to try to remove that in Photoshop using Curves in RGB channels. That was
really frustrating, considering the fact that with the previous monitor I
didn't have such problems from the beginning.
The friend of mine has recommended to remove this manually created monitor's
profile from the system (it is actually started automatically altering the
video card's LUT replacing the
default Window's values with those generated during my own calibration).
Since it is loaded automatically upon Windows startup, to remove it I had to
remove the link to this file in System StartUp menu.
After I did it, the cast has disappeared indeed for my happiness and the
scanned results started to look much more neutral resembling the situation
with old monitor.
This was relaxing.
However, I noticed new effect I didn't pay attention of earlier with new
monitor.
The scans delivered to Photoshop seemed overexposed (excessive brightness).
Trying to lower the brightness (by Analog Gain, Master channel or by Curves
tool in Photoshop) helps, but then, opening the pictures in something like
ACDS viewer shows much darker image then it appears in Photoshop. That
confused me.
I was told the ACDs and most other available viewer doesn't support Color
Management which means the results aren't corrected by the monitor's profile
the Photoshop uses.
This is frustrating.
The only way left to go is to remove the Iiyama monitor's profile from
Display Properties so that the default Window's display profile (if any)
would only one for all applications, until I'll acquire monitor calibration
hardware/software tool to create precise monitor's profile.
Any thoughts ?
Regards,
Alex Z
-----Original Message-----
From: filmscanners_owner@halftone.co.uk
[mailto:filmscanners_owner@halftone.co.uk]On Behalf Of michael shaffer
Sent: Saturday, April 20, 2002 2:53 PM
To: alexz@zoran.co.il
Subject: [filmscanners] RE: Scanner calibration
Alex writes ...
> Recently I was given the IT8 35mm slide to try out my
> IV ED calibration, but I have no calibration software
> to handle IT8 pattern.
> ...
Use littleCMS with care
... the results seem to be respectable, but I doubt if it will produce
anything better than the CLUT profiles which cane with your scanner. What
scanning software are you going to use it with? What settings?
cheerios ... shAf :o)
Avalon Peninsula, Newfoundland
www.micro-investigations.com (in progress)
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
