Yup. if you want reliable profiling, then you should disable all corrections before scanning.Originally Posted by Patrick Dixon
Yup. if you want reliable profiling, then you should disable all corrections before scanning.
Wolf Faust makes targets for various films; I only use Velvia 100, Astia 100F and Provia 100F, so I only need two targets because certain targets suit more than one emulsion. I use Monaco EZColor to make the profiles.
I think a canned approach would work better for me. I don't scan enough to justify custom profiles. Is there a simple, idiot proof workflow for the 4990?
What do you mean by a "canned" approach?
Yes, all you have to do is find out what scanner settings work for each and every transparency, do the scan and then work on the file in Photoshop to correct any problems you find.
If you scan a transparency without a profile, what you usually end up with is a colour cast and incorrect luminosity levels. However, these colour casts are not consistent across the whole image; the difference between correcting colour casts in Photoshop and via a profile is that, in Photoshop, the correction is made over the whole range of colours whereas, with a profile, the cast is corrected differently for different shades and tones of each colour.
So, in summary, as long as you don't mind spending time doing a lot of work in Photoshop, scanning without a profile is easy. I have better things to do with my time than spend hours trying to correct colours; I use a profile.
Profilling the scanner is not difficult. You put the target in the scanner, go to the software that does the profiling, punch the button, and then identify the resulting profile in your scanning software. Or, you use the scanning software to make the profile if it has the feature.
Then, you scan transparencies and they turn out the same as they look on your light table.
That is the canned approach. Doing it without the profile is the uncanned method.
Rick "noting that Vuescan does have a built-in profile for many films, and it does get pretty close" Denney
I've been printing B&W and ignoring these issues, so let me see if my analogy works. (My background comes from the darkroom.)
When you print color in the darkroom, you have to color balance the film, paper/chemistry, and the enlarger to get good color. Usually this is all done at the same time by dialing filtration to the enlarger, but it could be thought of as separate color balances for the film, enlarger, and paper/chemistry. (The papers used to come with printed color balance information.)
In the digital darkroom, it seems similar that you have to color balance the scanner (input), and you have to balance the printer (output), and it's awfully nice to balance the monitor. If you shoot different films, then there's separate color balances for the films.
It seems you could pull out an old Kodak color check grid (colors and B&W), shoot that on your favorite films, develop, scan, and make a profile, but I guess you're re-inventing the wheel, and it's easier to simply purchase the premade targets for $75 and use existing software. (Damn this hobby is expensive.)
Am I in the ballpark?
Jay
In the past, I've been pretty happy with the results that my other scanner (Coolscan 5000) produced using the Nikon Scan software. Sure there were color casts to change to get the look I wanted, but my slides themselves usually have color casts or don't look exactly how I want them to.
Is the Epson 4990 scanner software just not as good as others I've dealt with?
Very much so. Just as with the colour darkroom, you should be able to determine what filtration will be required for a certain paper, exposed under a certain enlarger and developed in a certain developer, that would equate (approximately) to a printer profile.
So, now you get to insert, say, a Kodak transparency; this will require an adjustment to the filtration, as would a Fuji transparency require a different adjustment to the filtration. This secondary adjustment for a particular make of film is the equivalent of a scanner profile.
Now, some say that you can use the same filtration, regardless of the film used but, in reality, most of us find that it is better to go with an adjustment (profile) for each film type, as it would appear that there are characteristics that can add different colour casts dependent on the make of film.
However, profiling a scanner has a slight advantage over altering the filtration on an enlarger; profiling writes a colour translation table that is not necessarily linear for all shades of every colour, thus it is slightly more "refined" in the way that the profile alters the colour balance, rather than applying one colour filter for the whole image.
If you have been "pretty happy" with unprofiled results from another scanner, then you shouldn't notice too many differences in using the 4990. In that, the Epson software is really not that different.
However, profiling eliminates the "standard", non-linear, colour casts in a certain make of film; this then means that any residual colour discrepancies are now down to you and how you shot the film.
Don't forget that a lab can add a magenta or green cast, simply by not checking the pH of the solutions are within range; no amount of profiling can account for that and it is a sign that you need to either scream at the lab or to change to another one.
Assuming I can't afford profiles, what is the best way to use the Epson Scan software? Let it color manage using Adobe RGB as the output space?
Posting Permissions
Reply With Quote
Bookmarks