preview scan and final scan histogram

  • Thread starter Thread starter Jean-Michel LEON-FOUN-LIN
  • Start date Start date
J

Jean-Michel LEON-FOUN-LIN

Hi,

I'm using a Nikon LS 50 with the latest CoolScan software on Mac OS X. I scan mostly slides.

I try to get the best histogram in CoolScan before scanning, with no shadow or highlight clipping.
But the histogram of the final scan very often shifts to the left, indicating the presence of more
shadow values than seen during the preview. There is shadow clipping in the final scan when there
was none in the preview scan.

Maybe it is the basics of slide scanning but could someone explain the problem to me (if it is a
problem at all)?

I have noticed that with VueScan, if I choose the option scan from preview, the preview and final
are (obviously?) the same. So maybe it is a matter of resolution difference between preview and
final scan?

Thanks in advance.

Jean-Michel LEON-FOUN-LIN
 
Jean-Michel LEON-FOUN-LIN said:
Hi,

I'm using a Nikon LS 50 with the latest CoolScan software on Mac OS X. I scan mostly slides.

I try to get the best histogram in CoolScan before scanning, with no shadow or highlight clipping.
But the histogram of the final scan very often shifts to the left,
Click to expand...
indicating the presence of more
shadow values than seen during the preview. There is shadow clipping in the final scan when there
was none in the preview scan.

Maybe it is the basics of slide scanning but could someone explain the problem to me (if it is a
problem at all)?

I have noticed that with VueScan, if I choose the option scan from preview, the preview and final
are (obviously?) the same. So maybe it is a matter of resolution difference between preview and
final scan?

Thanks in advance.

Jean-Michel LEON-FOUN-LIN
Click to expand...


I guess that the preview is done with a fixed exposure time and the
scan with an adjusted exposure time (maybe longer). This longer
exposure time pulls more details out of the shadow.

Winfried
 
Jean-Michel LEON-FOUN-LIN said:
Hi,

I'm using a Nikon LS 50 with the latest CoolScan software on Mac OS X. I scan mostly slides.

I try to get the best histogram in CoolScan before scanning, with no shadow or highlight clipping.
But the histogram of the final scan very often shifts to the left,
Click to expand...
indicating the presence of more
shadow values than seen during the preview. There is shadow clipping in the final scan when there
was none in the preview scan.

Maybe it is the basics of slide scanning but could someone explain the problem to me (if it is a
problem at all)?

I have noticed that with VueScan, if I choose the option scan from preview, the preview and final
are (obviously?) the same. So maybe it is a matter of resolution difference between preview and
final scan?

Thanks in advance.

Jean-Michel LEON-FOUN-LIN
Click to expand...

The best instruction on scanning is found at:
http://www.scantips.com
 
Hi,

I'm using a Nikon LS 50 with the latest CoolScan software on Mac OS X. I scan mostly slides.

I try to get the best histogram in CoolScan before scanning, with no shadow or highlight clipping.
But the histogram of the final scan very often shifts to the left, indicating the presence of more
shadow values than seen during the preview. There is shadow clipping in the final scan when there
was none in the preview scan.
Click to expand...

As you suspected, some minor variations are bound to happen due to
resolution difference between preview and final scan. However, that's
negligible compared to additional processing that takes place, for
example, curves. If you've set clipping that's more likely to cause a
major shift in the final scan's histogram than anything else.

Another culprit, albeit not for slides, is autoexposure. That's
because in addition to autoexposure NikonScan "does you a favor" by
messing with the image.

Because of all of the above, many people do so-called "raw scanning".
If you're not familiar with the term, it basically means you turn off
"everything" and try to get as "unpolluted" an image as possible at
maximum optical resolution, also known as a "digital negative". The
only exception is ICE as that's hardware based and can't be done
afterwards. All the image editing is then done in post-processing
using dedicated software instead of NikonScan's "keyhole" and a very
limited subset of graphic tools. If you search the archives there were
lots of messages on the subject with more details.
I have noticed that with VueScan, if I choose the option scan from preview, the preview and final
are (obviously?) the same. So maybe it is a matter of resolution difference between preview and
final scan?
Click to expand...

VueScan is far too buggy and unreliable to be used for this kind of
testing (or, indeed, anything else). Results vary wildly between
different versions.

Although NikonScan has its problems, most of them can be circumvented
by turning off auto-exposure. This is not easy because NikonScan
desperately clings to it. You need to turn off all occurrences of
auto-exposure (positive, negative, preview) close NikonScan, turn off
the scanner (!) and they restart everything. Do note that if - after
that - autoexposure is used at any time, the whole process must be
repeated because once used autoexposure changes the baseline setting.

Don.
 
Hi Winfried,

thanks for your answer.

I've played a bit with the auto-exposure settings. To no avail. I
unchecked DEE which I had been using extensively for its "magical"
effect on shadows and highlights. And I had quit playing with the curves
out of laziness. And now I use the curves only again and the preview
histogram and scan histogram match.

I am happy that this works but I don't understand why DEE causes the
problem. When it is activated in Preview shadows are actually lightened
and the left of the histogram moves right accordingly. In the scan DEE
is also performed, the shadows lightened but the histogram doesn't
comply.

What am I missing?
 
Back
Top