batch color and area

[braver]

Greetings -- I'm using SilverFast Ai 6 to archive my family scans. I'm using Nikon CoolScan 4000 with an SF-200 slide feeder. I'd like to enable maximum flexibility and uniformity for batch scans, for subsequent processing with SilverFast HDR and then Photoshop 7 or CS.

Perusing Ian Lyons' tutorials from computer-darkroom.com, notably the HDR and Batch ones, I see that there're two general ways to perform the archival color management: set Scanner->Internal to None, and not embed any ICC profiles, or set Scanner->Internal to Calibration (or ICM and then choose the Scanner's ICM profile) and embed a Photoshop profile, e.g. Adobe RGB (1998).

Hence my questions --

-- Is there a difference between
(a) choosing Calibration in Scanner->Internal, so that the SF_T (Nikon LS4000) is greyed out in the ICM subsection below, and
(b) choosing ICM for Scanner->Internal, and then explicitly specifying the same SF_T (Nikon LS4000), now white/selectable, in the ICM subsection?

-- What happens if I do choose Calibration, but don't use a Photoshop Internal profile, and/or do not embed it? Ian Lyons warns not to mix the settings from the two methods, but it looks like you can simply apply the Internal change in HDR afterwards, or select the profile in Photoshop, after all? Generally speaking, how does using Scanner->Internal set to Calibration relate to choosing Internal set to Adobe RGB (1998), and embedding the latter -- or not?

My underlying desire is to make an archival record of the scanner's IT8 calibration at the time of scanning -- I recalibrate it before major batch scans -- but leave myself free for batch-processing it later. Since the processing will be done, finally, in Photoshop, I don't mind embedding things so long as they simplify workflow, and record anything I need to remember about the scan in the results, but do not lead to information loss in any way. So, why _either both_ Calibration/Embedding or None/None, in both, as Lyons says?

Another question concerns cropping the scan area for batch slide scans. Different frames differ in the slide area they expose. I'd like to err on the side of the larger scan area size -- but I'd also like to avoid tedious manual cropping too much black areas off. How do I set the crop rectangle to safely and reliably capture the whole batch with minimal extra black? Is it possible for SF to do this automatically for each slide? On the other hand, if I end up with a lot of extra black, can HDR or Photoshop be set to remove it in an easy way, for large batches?

When tuning the position/size of the scan area rectangle, I found out its annoying sticky behavior -- e.g., I was capturing an unusually positioned slide, leaving the rectangle in an odd place, -- then starting SF next time would find the rectangle there. If you forget to prescan, you can misalign the whole new batch! Is there a way for the rectangle to return to a "normal" setting? Moreover, if I were adjusting it and lost track, I'd love to be able to reset it to the "normal" centered 24x36 mm in 50x50 mm area expected of a slide -- again, is there a simple way to do it?

Danke,
Alexy

[LSI_Noack]

Dear Alexy

Options > CMS > Scanner to Internal: Calibration
is an outdated option. More recent releases of SilverFast either use "None" or "ICM" (Win) resp. "ColorSync" (Mac) as settings.

The "Calibration" differs from "ICM"/"Colorsync" in that not only the data from the scanner input are standardized, but already matched to a specific color workspace.
This now is done internally in a separate option (with setting the "Internal" ICC/ICM profile in the middle section of the CMS tab).

So, recent SF version use their scanner profile indepently from the color workspace your are profiling. But still, the gamma value used is put into creating the scanner profile while calibrating.

When choosing "Scanner to Internal: Calibration" you must not use another internal color workspace than that the calibration was done with.
Of course, you can have more than one of such profiles.

I suggest you drop the awkward option and version altogether and go for the ICM/ICC method.


None / none is a different thing. While I suggest you might want to ask why Mr Lyons wrote what he did instead of me, the reason perhaps is:
If you are using an internal color workspace for profiling your scanned images, you should be embedding that piece of information into the file so that application which will work with the resulting data need to know how this data is 'formatted' to apply its routines correctly.

If none / none is setting, you don't have an internal profile. Whether you then select to embed this info, is not really relevant, because if no such data is embedded, the graphics app has no other choice as to work if no color workspace information was given.


In conclusion, update and either you want to use a profiled scan: ICM/ICM (or: Colorsync/Colorsync) with the scanner profile, embedding might proof to be very useful.


Unfortunately, SilverFast doesn't offer an automatic image frame detection.


Deleting all scan marquee frames will let you "return to the standard setting". However, you might find it useful to learn that frame settings can be saved and retrieved on the general tab.


Finest regards
Sonny Noack
- tech support, LSI AG -

[braver]

Sonny -- thanks for the detailed reply! Just to make sure, I'm now downloading the updates to my SF 6 Ai/HDR, without the Calibration option. However, I did already perform many scans using Ian Lyons' "User Group 2" method (from his HDR tutorial):

(1) Selecting Scanner->Internal set to Calibration, thereby ICM is preselected as SF_T (Nikon LS4000) in the ICM section
(2) Selecting Adobe RGB (1998) as the Internal, and embedding it.

Does it mean that Calibration will map to Adobe RGB (1998) in the resulting image, properly reflecting the IT8 calibration, so I can assume I'm opening a correct Adobe RGB and forget about the scanner in HDR/Photoshop?

That is, Will I have the same result from the new version, 6.4.1, when
Scanner->Internal set to ICM, ICM set to SF_T, and Internal set to Adobe RGB and embedded to On, as in the above "old" setup with Calibration? My "old" version was 6.0.1r19.

Cheers,
Alexy

[braver]

OK, I upgraded to Ai/HDR, 6.4.1, and now have just one choice of Scanner->Internal, None/ICM, effectively meaning Color Management Off/On.

My questions becomes this: I have three choices of setting color management for archival scans:

(1) Everything Off -- Scanner->Internal=ICM/Input=None.

(2) ICM On, Scanner profile for Input, None for Internal.

(3) ICM On, Scanner profile for Input, Ektaspace for Internal, Embed On.

(I've read that Ektaspace by J. Holmes, www.josephholmes.com , is better than Adobe RGB for archival purposes.)

My question is reproducibility/equivalence: aren't these three in fact *equivalent* in the sense that I should be able to losslessly transform a scan done in one of those regimes to any of the other two states, *given* I keep my Scanner profile which was used at the time of scanning, and remember what Internal space was used (or just embed it)?

For example, say I forgot to turn ICM On, i.e. scan in (1) mode, what's the final difference if I know that and provide ICM/Input profile of my Scanner in HDR and set the Internal as I wish? Will there be any difference between the scenarios where the scanner profile is added in HDR, not Ai, and/or the Internal color space such as Ektaspace is also chosen in HDR, not Ai?

But this raises further questions: If I know exactly what Input and what Internal space were used, why should I care about choosing Ektaspace versus Adobe RGB at the sime of scanning, then? Is the transform from the calibrated Input to Internal in ICM lossy, and I have to save in Ektaspace if I want to go to Adobe RGB later, since the other way is impossible? Why not choose None for Internal then, losing nothing at all, hopefully, and pick an archival Internal space later?

Is the Internal space then only for memory, i.e. you explicilitly specify the target space which doesn't change, such as Ektaspace, as opposed to remembering a scanner's profile which is individual to the scanner and also may change in time?

I see there're some wide gamut color spaces there, is any of them preferable to Ektaspace for archival purposes in any way?

While I'm on it: how often should I recalibrate my scanner, LS4000ED, if at all? If I got an IT8 target 3 years ago, should I replace it before I recalibrate? And should I keep an old recalibration performed when the IT8 target was new, as opposed to recalibrating the scanner which became older with the target which also became older?

Cheers,
Alexy