Trying to understand Vuescan workflow: expert help needed

[Pedro Gomez]

Hi everybody

I'm trying to understand which is the actual workflow for exposure,
histogram, BP & BP determination when a preview/scan is executed My
understanding of what happens is described below: (I assume that
neither "exposure lock" nor "film base color" are checked, I assume as
well that "Input|Image" is not selected for the fbc part)

I assume that each time we click on preview, a scan with lower
resolution, exposure 1.0 and limited to the area defined in the crop
tab is carried out. Vuescan uses this to compute the histogram, RGB/IR
exposure (taking into account "exposure clipping").and BP/WP (taking
into account WP % and BP %), all of this taking into account only the
pixels comprised in the crop area. Is this assumption correct?

I assume as well that fbc determination is carried out at this point,
although this may not be very useful since fbc is sensitive to the
exposure and we haven't made a scan with the calculated RGB/IR
exposure yet (as a matter of fact, we can't lock fbc yet). Is this fbc
determination carried out at this stage at all?

I believe this first preview is necessary for determining the above
mentioned parameters, particularly RGB/IR exposure, and therefore this
preview happens every time we hit the scan button, unless "lock
exposure" is checked and "auto crop" is cleared. (why autocrop? Is
that a "sanity" check?) Is this assumption correct?

I understand as well that we can refine these calculations and obtain
more precise values for the histogram, BP and WP by making a crop and
clicking preview again, and this would be particularly critical for BP
and WP determination (I guess that more for the BP since by not
cropping appropiately we may introduce part of the "black frame"
around the image which may not represent the true black point of the
image). Is this correct?

I fail to understand, however, how fbc determination should yield
different results by cropping on the clearest representative area of
an image (leader or image itself) or by cropping the full scan area.
Shouldn't Vuescan fbc algorithm find the clearest area by itself and
produce the same results? I'm lost in this point

I would very much appreciate confirmation/comments on the above
mentioned points
Thanks
Pedro.

 

[Bart van der Wolf]

SNIP

I assume that each time we click on preview, a scan with
lower resolution, exposure 1.0 and limited to the area defined
in the crop tab is carried out.

Yes, although the Crop|Preview area allows preview scanning different areas.

Vuescan uses this to compute the histogram, RGB/IR exposure
(taking into account "exposure clipping").and BP/WP (taking
into account WP % and BP %), all of this taking into account
only the pixels comprised in the crop area. Is this assumption
correct?

Yes, although the Crop|Buffer% allows to ignore edge pixels from the crop in
the subsequent calculation, and Crop|Border% allows to include more pixels
which are not included in further calculations but only output.

SNIP

Is this fbc determination carried out at this stage at all?

Don't know, perhaps as an approximation.

I believe this first preview is necessary for determining the
above mentioned parameters, particularly RGB/IR exposure,
and therefore this preview happens every time we hit the scan
button, unless "lock exposure" is checked and "auto crop" is
cleared. (why autocrop? Is that a "sanity" check?) Is this
assumption correct?

Switching off Auto crop only saves time.

I understand as well that we can refine these calculations and obtain
more precise values for the histogram, BP and WP by making a crop
and clicking preview again, and this would be particularly critical for
BP and WP determination (I guess that more for the BP since by not
cropping appropiately we may introduce part of the "black frame"
around the image which may not represent the true black point of the
image). Is this correct?

Different crops will result in different effects of a BP/WP percentage. You
can improve the accuracy of the histogram data by scanning at the maximal
native resolution instead of a more rapid lower ppi setting or Auto
resolution.

I fail to understand, however, how fbc determination should yield
different results by cropping on the clearest representative area of
an image (leader or image itself) or by cropping the full scan area.
Shouldn't Vuescan fbc algorithm find the clearest area by itself and
produce the same results? I'm lost in this point

FBC also requires a certain percentage of the pixels to be the brightest in
the histogram. If the image itself doesn't have enough totally overexposed
areas to represent that percentage, the accuracy of determining the film
base color is compromised. In that case it can help to crop only those
reflection highlights.

Bart

 

[Erik Krause]

Hello, Pedro Gomez
you wrote...

I assume that each time we click on preview, a scan with lower
resolution, exposure 1.0 and limited to the area defined in the crop
tab is carried out. Vuescan uses this to compute the histogram, RGB/IR
exposure (taking into account "exposure clipping").and BP/WP (taking
into account WP % and BP %), all of this taking into account only the
pixels comprised in the crop area. Is this assumption correct?

I assume as well that fbc determination is carried out at this point,
although this may not be very useful since fbc is sensitive to the
exposure and we haven't made a scan with the calculated RGB/IR
exposure yet (as a matter of fact, we can't lock fbc yet). Is this fbc
determination carried out at this stage at all?

It is. FBC is determined any time you do a preview or scan just before
other color and brightness corrections are made and before a color
space is applied. It would not be usefull to do it afterwards because
it would fail if the FBC is already clipped away.

It must be done for the preview anyway. Otherwise there would be a
strong color cast if previewing color negatives. Since the FBC values
that you can lock are for the final scan, it is not allowed to lock
them for the first preview only, where they might differ.