Filmscanners mailing list archive (firstname.lastname@example.org)
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[filmscanners] Re: Sharpening after scanning (SS4000): questionforArt
What you are saying makes sense, in terms of the progressive unsharp
masking process, and indeed my own workflow sometimes includes this.
One of the reasons I came to this was because I found occasional
upsetting artifacts showing up once I had completed the manipulation and
compositing work when I then did the large USM at the end. Suddenly,
defects I should have corrected in masking, dust clean up, and other
artifacts showed up where they were not noticeable when the image was
still soft. This was particularly so with masking processes. By doing
some early-USM the edges were more defined and allowed for better
masking and cut and pasting, and even in cases of some types of clean up.
I also suspect doing a progressive USM (even if it were done at the end)
by in stages and steps, might allow for (ironically) "softer sharpening"
which might look more natural, sort of like a fractal-like process where
definition was generated by "massaging" the pixels into place.
Laurie Solomon wrote:
> There is a current wisdom among many including some industry gurus that
> because of the points you make regarding captures by scanners (and I might
> add digital cameras), it is beneficial to apply slight sharpening to an
> image prior to doing any editing of the image, additional sharpening at the
> end of the editing stage with focus on local sharpening, and final
> sharpening of the overall image prior to outputting. This does represent a
> sea change from the all-at-once prior to printing advice that use to be in
> fashion in the golden days of digital's youth.
> email@example.com wrote:
>>I may have mis-spoken or at minimum, been misunderstood.
>>You are correct that sharpening should occur prior to printing.
>>Saving the image sharpened is not necessary, and may, in fact, be
>>detrimental since sharpening adjustments vary depending upon final
>>output size and other factors. They may even depend upon the printer
>>type and driver software.
>>I cannot give you absolutes in terms settings in using unsharp
>>masking, because it depends upon many factors. Some include the type
>>of image or subject matter and contract, color intensity, etc, the
>>size the imagine is going to be reproduced to, and the scanning
>>resolution used, the type of source material (the film base used) and
>>indeed the type of scanner and if things like dICE is used or not.
>>By trial and error, I have a "sense" of the settings depending on
>>these factors, and how the image looks on the screen at differing
>>However, my principal point is this:
>>All CCD based scanners tend to introduce softening which can in part
>>be recaptured via unsharp masking. This softness is not a defect in
>>focus or optics or the CCD, but is intentionally introduced to reduce
>>the amount of noise and artifacting (Nyquist errors) that develop in
>>the analogue to digital transfer which occurs in the scanning process.
>>No image should be compared until optimum unsharp masking is
>>accomplished because some manufacturers will uses some USM to make
>>their scanners appear to have higher sharpness and resolution when
>>actually introducing this higher focal accuracy may add unnecessary
>>and even undesirable artifacts which cannot later be removed.
>>Keeping the image unsharpened for storage does indeed allow you to
>>adjust those measurements to the output method and size.
>>Stan Schwartz wrote:
>>>A while back, Art mentioned sharpening a scanned transparency image
>>>before saving it--to restore some of the loss of sharpness inherent
>>>in the SS4000 scan. I am curious to know what degree of sharpening
>>>you use, in Photoshop terms re: %,radius and threshold, for this
>>>I've usually reserved sharpening as the last step before printing,
>>>leaving my archived image unsharpened.
Unsubscribe by mail to firstname.lastname@example.org, with 'unsubscribe
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or