Skip to main content

⚠️ Please note that this topic or post has been archived. The information contained here may no longer be accurate or up-to-date. ⚠️

Sharpening with rotation or keystone corrections

NN21962
Rotation and keystone corrections have the effect of displacing pixels on the image space - in fact, computing new pixels.

I suppose sharpening happens after geometrical correction in the processing of the image.

Question: what is the best sharpening strategy in case of a geometrical correction? My impression is that corrected images are not as sharp as straight ones: should the radius or the intensity of capture sharpening be increased? By how much?

Thanks for sharing your ideas and experiences.

Reto
0

Comments

5 comments

Date Votes
  • SFA
    Good question Reto.

    It is one that happened to occur to me yesterday as I was processing an image with a lot of fine detail shot upwards against an indirect as a light source. I was playing with "correcting" the perspective for the upwards distortion and realised that it resulted in quite a significant adjustment that might have a major effect on some adjustments in some parts of the image.

    So far I have not investigated further but hope to do so soon.

    A quick view led me to believe the challenge had been handled well (at least for my RAW file from a Canon G11) but the underlying question remains.

    It might be a very good subject for one the Image Professor's blog posts.

    I suppose selective sharpening is an option using local adjustments so that the modified and infilled areas could be treated differently as required but the question about the best approach to the settings for any adjustments is more important perhaps?

    Grant Perkins
    0
  • NN21962
    Thanks for your answer, Grant. I also thought the issue would make a good blog entry. On the other hand, the C1 blog is more oriented towards the capacities of the program than towards solving problems created by those capacities... Compensating for loss of sharpness when changing the geometry of the picture is a general processing issue. Hey, Cambridge in Color would eventually be a better address than the present forum.

    Actually, the pro solution would be to use tilt-shift lenses and primes with little distortion...

    Reto
    0
  • SFA
    [quote="NN21962" wrote:
    Thanks for your answer, Grant. I also thought the issue would make a good blog entry. On the other hand, the C1 blog is more oriented towards the capacities of the program than towards solving problems created by those capacities... Compensating for loss of sharpness when changing the geometry of the picture is a general processing issue. Hey, Cambridge in Color would eventually be a better address than the present forum.

    Actually, the pro solution would be to use tilt-shift lenses and primes with little distortion...

    Reto



    I agree Reto.

    However with the continuing developments of reasonably priced smaller lenses that require software "fixing" in some nearly mainstream commercial uses one of the smarter things of software development may indeed be how well a program deals with distortion correction and the resulting "problems". Realistically, at the Pro and serious non-pro levels, these two things surely go hand in hand? So optimising the technique and/or the steps of the workflow could be an important USP to be broadcast rather than ignored.

    There is probably a Doctoral Thesis (or several) to be written on the subject. Perhaps they already exist but not in the context of RAW file conversion software at the user level.

    If you come across anything that discusses the options I, for one, would be really interested to read it. I have not checked Cambridge in Colour for that sort of subject recently ... 😉

    Grant Perkins
    0
  • Johnny Walter

    I cannot find any Doctoral Thesis written on the subject, tried to look already. I would love to use at least one Doctoral Thesis for my research paper to retrieve information.

    https://essays.agency writer, Johnny B. Walter

    -1
  • SFA

    SPAM

    0

Post is closed for comments.