Aberrations at Lens Elements as Unsharp Filter

[Mustafa Umut]

I want to learn do aberrations and compensation actions used as unsharp filter for final image ?

I read a Leica paper. There was symmetrical lens Summicron as a subject. I found that at the exit of the lens there were no aberrations without corrected. But when you look inside , some lens elements at the left side highly aberrated and there were lens elements with highly negative aberrations at the right side as an compansating action.

Unsharp masking is to blend a normal picture with its low contrast version to create edge effect and the sharpness. Low contrast version might be separated from the original , based on a color without being member of RGB CMYK separation colors.

Mustafa Umut Sarac
Istanbul

 

[Drakkith]

What is your question?

 

[Mustafa Umut]

I am asking is it true or not ?

Thanks,
Umut

 

[Drakkith]

Is what true?

 

[Mustafa Umut]

OK,I wrote my findings , do lens element aberrations and their compensation act like an unsharp masking filter. Is it true or a dream ?

 

[Drakkith]

OK,I wrote my findings , do lens element aberrations and their compensation act like an unsharp masking filter. Is it true or a dream ?

I suppose you could say that the two are similar, but I think it is more accurate to simply say that the lens elements act together to cancel out their individual aberrations and form the sharpest possible image. Unsharp masking then takes that image and attempts to sharpen it even further. I think the key is that unsharp masking acts results in an image that may or may not be an entirely accurate representation of that object, while reducing lens aberrations is always more accurate.

 

[sophiecentaur]

I think the Unsharp mask is a non-linear process (do they not include a 'threshold' level in the Unsharp Mask process?). That is not necessarily the same as the aberrations of a lens. Why should the two things be equivalent? It is not necessary that the two operations are the same - all that's necessary is that the subjective effect is to produce a 'better' picture yet minimise the effect on low level noise.
I am only giving an opinion as a user and not a writer of USM algorithms.

 

[Drakkith]

My Astronomical Image Processing book says the following about Convolution by Unsharp Mask:

Unsharp masking is a convolution technique that, at first sight, appears to be something else. It works because the linearity property of convolution allows you to break a kernel into additive component kernels.

Later...

Since convolution is linear, it is legitimate to multiply the values in the image generated by the unity kernel by a contrast factor, c; multiply the unsharp mask image by -(c-1); add the two images; and normalize the image by the sum of the elements in the mask.