On 6/20/2012 2:34 PM, Joel Wilcox wrote:
> On Tue, Jun 19, 2012, at 10:24 PM, Moose wrote:
>> Vuescan uses standard ICC color profiles. I suppose there might be some
>> B&W profiles out there, although I'm not sure
>> how that might work. IT8 profiling works by comparing scan values for the
>> color patches to absolute numbers I suppose it
>> may first brightness correct the whole thing using the B&W steps? I can't
>> at the moment imagine how to use that for B&W
>>
>> And you might try Ed again on B&W profiling. The renaissance in B&W film
>> work may have put more pressure on him in the
>> meantime. If anyone could figure out a way to create B&W profiles using
>> just the B&W steps on the IT8 target, I'd think
>> he would be the one. I don't know if they could be standard ICC format,
>> though.
>
> I think he has always had a halting gait when it comes to profiles.
> Initially, the canned profiles were supposed to be sufficient, but they
> didn't cover all the films, then he suggested using the Advantix ones
> for color as starting points and the TMAX for BW (if I recall his
> responses of the old Epson scanner list correctly). Finally, he
> provided a way to do custom profiling, but I had to coax it out of him
> as to what canned profile to set it to when building a custom profile.
> I suggested "generic" and he agreed -- and I'm not meaning to complain
> -- but it seems odd that there is even any other choice if one means to
> create a custom profile.
All that really matters for practical, local use is to document which VS film
profile you used in making an ICC profile,
and set the same one when applying the ICC profile.
But if you want generalizable ICC profiles for color film that include any the
film base color in their correction,
select Media: Image in the Input tab. If you then use an ICC profile for the
scanner on the Color Tab, the ICC profile
is strictly for the film, and usable anywhere.
I think there's a way to do that for color neg film, but haven't cared enough
to figure it out.
> Vuescan is a so many things to so many people. It is hard to imagine how he
> has done what he has done, but there seem to be a few doors without a landing
> at times.
True enough!
>
>> If it's close to being a gamma problem, one could make custom PS color
>> settings with different gammas and find one you like for a particular film.
>> If you use VueScan's RAW output option, you can then apply any gamma you
>> want to the image in PS.
>
> If you don't mind, I will paraphrase this conversation and give it another
> try with Ed.
Sure.
I think it really is a difficult problem. The same issues as in my recent
treatise post about color scanning apply. Is
one looking for a particular tonal distribution, regardless of film used? Or is
one interested in letting the 'native'
tonalities of the film/development come through?
But if something like the first is desired, there is nothing like ICC color
profiling and targets for B&W. One possible
solution might be to shoot the IT8 target on the B&W film, scan with generic VS
profile, linear curve, color balance
"None", and compare to a desaturated color shot of the target that has been
fully profiled. Using Brightness, Contrast,
Levels, Curves, etc., it should be possible to create a PS Action that tonally
adjusts scans of that particular film to
match the B&W part of the IT8 target pretty closely.
I assume from your post that the second alternative, being satisfied with the
native tonalities, isn't what you want.
>>> I don't see why there would be any point in scanning in 16bit rather
>>> than 8bit if doing BW, ...
>>
>> You are conflating bit depth with color. There is exactly the same reason
>> to work in 16 bit in B&W as in color. ...
>
> OK, I have no problem with that. I can select RGB or Grayscale and work
> in 16bit with either. Ken suggested scanning in RGB and then converting
> the scan to grayscale in PS. That seems to me to add an unnecessary
> step. What do you think?
I don't see any point in RGB scanning, just makes bigger output files with
three channels. As Ken says, some film
backings have a color cast. So scan one in RGB, look at the channels, determine
which one works best, and select that
channel in VS for scanning.
> I'm familiar with the ratios. It is necessary to know this in order to
> make one of the input settings line up with another setting in
> preferences (I think -- have to check that. If you have 16bit in one
> place and 24 bit in the other, the scan takes forever or doesn't work at
> all.
Input Tab
- Media: B/W Negative
- Bits per pixel: 16 bit Gray
Output Tab
- TIFF file Type: 16 bit Gray
[A bit of caution for anyone using chromogenic B&W film, not silver based. To
take advantage of IR based dust removal,
you need to set Bits per pixel to '64 bit RGBI, which does indeed seem to take
forever. That's because RGB takes one
scanner pass - and IR a full second pass. There is actually no wasted scanner
activity. You can leave Output at 16 bit
gray to avoid a three times larger than necessary output file.]
Scan That Moose
--
What if the Hokey Pokey *IS* what it's all about?
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|