Pages: 1 ... 8 9 [10]   Go Down

Author Topic: deconvolution sharpening plug in  (Read 43857 times)

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland
Re: deconvolution sharpening plug in
« Reply #180 on: February 29, 2016, 02:20:45 pm »

The solid lines all refer to the same final pixels, the 4:1 downsized ones, so their results are as measured. The dashed original line is there for reference.

Yes, it's the dashed original, in particular, that I don't understand (the MTF curve seems way too high).

Quote
Did you resize the image 4:1 using the various methods?

MTF Mapper never gets confused, if anything it's operator error :)  But in this case it looks like you are using the original edge at its native resolution so that's where the discrepancy comes from.  And you are probably unknowingly adding a little sharpening somewhere in your workflow, because the MTF50 value in cy/px looks high.  Have you tried running Imatest on the cropped tiff I provide there?

I was using the original image as supplied in your link ... which corresponds to the dashed line, correct?  I didn't add any sharpening - if I had my MTF result might have been a bit nearer yours, but as it is the Imatest MTF result was much lower, as you can see.

Here is the MTF with bicubic downsized by 4x (which actually leaves the image not really big enough to run Imatest on).  This has added quite a bit of sharpness - or possibly the Imatest result is not reliable because the sample has become too small.  At any rate the result is not too far off your bicubic result. 

So it's really your dashed curve that I don't get.



Cheers,

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana

bjanes

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3354
Re: deconvolution sharpening plug in
« Reply #181 on: February 29, 2016, 03:43:42 pm »


Here is an image that was upsized by 2.95x, with all of the above steps.  BTW ... these are very small flower-heads and the flowers have a grainy look ... the white dots are not caused by sharpening.




In my opinion the white dots are specular reflections caused by harsh lighting of the flower. They are more prominent when using an undiffused flash and can be reduced by diffusing the flash or better yet by using a soft box or other means to produce soft lighting. I have often noted these artifacts when photographing orchids.

Bill
Logged

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland
Re: deconvolution sharpening plug in
« Reply #182 on: February 29, 2016, 05:15:31 pm »

In my opinion the white dots are specular reflections caused by harsh lighting of the flower. They are more prominent when using an undiffused flash and can be reduced by diffusing the flash or better yet by using a soft box or other means to produce soft lighting. I have often noted these artifacts when photographing orchids.

Bill

I think you're right (although I didn't use a flash).  I photographed the flower again in very diffuse, but direct, light and I still got the speckles.  Then I photographed it with indirect light only and there were no speckles, but I could see that the surface is quite bumpy and shiny, so that with any light shining directly on the flower you will get these reflections.  It's difficult to see because the flower heads are very small (about 1/2 cm in diameter).

Cheers,

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana

Jack Hogan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 794
    • Hikes -more than strolls- with my dog
Re: deconvolution sharpening plug in
« Reply #183 on: March 01, 2016, 03:10:33 am »

Yes, it's the dashed original, in particular, that I don't understand (the MTF curve seems way too high).

Right that one is there for reference and its frequency axis is in lp/px: it has been scaled  to reflect the different pixel size, assuming one would view final images at the same size, see the bottom of the post.  Take a look at the first image in the article in lp/mm: does it make sense to you?  If so, so should the second one.  Once you understand it there are some interesting insights to be had.  Such as that when viewing an image at standard size, what really matters as far as perceived sharpness is concerned is MTF 90.  Think about that when you next consider a lens for purchase :)

Jack
Logged

Robert Ardill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 658
    • Images of Ireland
Re: deconvolution sharpening plug in
« Reply #184 on: March 01, 2016, 04:43:39 am »

Right that one is there for reference and its frequency axis is in lp/px: it has been scaled  to reflect the different pixel size, assuming one would view final images at the same size, see the bottom of the post. 

OK, I see. You're multiplying the MTF value for the original image by 4.  I don't understand why. The scale is surely the same.  Is it to show somehow the degradation in image quality?  Would you then divide the original MTF by 4 when showing an image upscaled by 4?  I would be interested in understanding this, although I guess this is way off topic.

What is interesting is the MTF of the nearest neighbour downsample - which does horrible things near and beyond Nyquist.

Robert
Logged
Those who cannot remember the past are condemned to repeat it. - George Santayana
Pages: 1 ... 8 9 [10]   Go Up