Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: filmscanners: VueScan 7.2.11 Available (new focus test)

On 12/4/01 2:30 AM, EdHamrick@aol.com <EdHamrick@aol.com>, wrote:

>In a message dated 12/4/2001 2:07:36 AM EST, julianv@mindspring.com writes:
>> Here are two sample scans (one from VueScan, one from Nikon Scan) that 
>>  show the benefit of a manually selected focus point:
>>  <http://julianv.home.mindspring.com/focus_test/focus_test.html>
>No, there are too many other differences for this to be a useful
>test.  It isn't clear if sharpening has been applied to the NikonScan
>result, and the range of intensities is different in the NikonScan
>and VueScan images which makes them look different.
>A more useful test would be to first use the Advanced Workflow
>suggestions to lock the exposure, film color and image color.
>Set "Device|Auto focus" to "Scan" and scan the whole slide.
>Then crop to a smaller area and scan again with auto focus
>on the smaller area.  Then change "Device|Auto focus" to
>manual and manually offset the computed focus value by some
>numbers while scanning the smaller area.

Hi Ed,

I followed the above suggestions as closely as I could.  The new results 
can be examined at:


I used the same slide from the earlier test, but this time, all scanning 
was done with VueScan.  The "crop 1" image was obtained by setting the VS 
crop rectangle to the entire slide, with "Device|Auto focus" set to 
"Scan".  I used Photoshop to trim the detail area out of the tif file, 
and then saved as jpeg.  The "crop 2" image was obtained by setting the 
VS crop rectangle to an area just slightly larger than the "version 1" 
detail area.  "Device|Auto focus" was still set to "Scan".  Photoshop was 
used to trim "version 2" to match the size of "version 1".

No sharpening was done to either image.  No other edits or Photoshop 
manipulations other than cropping and conversion to jpeg.  All VueScan 
filters were off.  Other info: Provia 100F in plastic mount, scanned at 
4000ppi on a Nikon LS-4000ED, using VueScan 7.2.11, Mac OS 9.2.1.

The difference between these images is not huge, but it is clearly 
visible to me (and even more apparent in the tif files).  Setting the 
focus point closer to the area of interest did make an improvement.  I'm 
actually a bit surprised, since the detail area is close to where you say 
that you are setting the autofocus point.  On these Nikons, a small 
change in focus location can make a difference.  Maybe this slide has an 
unusual bow shape (if you want to examine it, I would be happy to donate 
it to you).  It does not appear severely bowed.

At your suggestion, I did experiment with manual overrides to the auto 
focus setting on the "crop 2" scan.  The auto focus result was the best, 
at -0.431.  I tried four manual settings from -0.37 to -0.5.  For the 
"crop 1" scan, the auto focus result was -0.531.

Julian Vrieslander <mailto:julianv@mindspring.com>


Copyright © Lexa Software, 1996-2009.