Grain aliasing

[alan_rockwood]

I have looked all over the place for a good and credible explanation of the origin and cause of grain aliasing when

scanning film, but I can't honestly say I have seen a satisfactory explanation. Can anyone here add some fresh

insight into this issue?

 

Don't be afraid to make either a highly technical or a non-technical comment.

[dallalb]

I think it's the result of scanner's light scattering through the film's grain that produces "grain aliasing" when captured by the CCD...

[dallalb]

...If you search on Photo.net forums you will find plenty of informations about this argument.

Regards, Alberto.

[Ed_Ingold]

"Grain alienization" is a term bantered about but seldom explained. If grain is comparable in size and distribution to the pixel array, it will sometimes be exaggerated in the scanning process. A given pixel is either affected or not affected by grain, depending on this distribution. Consequently, the digitization process forces the frequency and distribution of the effects of grain to that of the pixel array.

 

Grain which is fine but widely distributed can also be exaggerated. Velvia, for example, exhibits a "pepper grain" effect in areas of medium density. These look like black specs resembling pepper which can be seen at 25x magnification or so. Because they are very dark, they dominate any pixel where they appear, and are magnified to the size of that pixel as a result.

 

In the attached photo, you will see evidence of grain aliasing in Velvia as irregular dark clumps in the lighter area (pepper grain is evenly distributed). You also see evidence of the little air bubbles found in velvia, which are represented fairly accurately in this scan. The scan was made using a Nikon LS-4000 at 4000 ppi.<div></div>

[leicaglow]

Edwards example is perfect, but I liken it to apparent "grain clumping". I find it tends to diminish if I turn off, or down, the dust removal feature, but also get it with Digital Ice.

[jim_hein]

The Definitive answer:

 

http://www.photoscientia.co.uk/Grain.htm

 

enjoy

Jim

[alan_rockwood]

Thanks for the link Jim. I have actually read that web page several times in the past, and it comes closest to a

real explanation of anything I have seen, but I would say the mechanism for the formation of grain aliasing is

still not entirely clear to me.

 

I suspect that grain aliasing is closely related to the phenomenon of aliasing that occurs in Fourier transforms,

obviously not a new insight on my part. However, the details of the mechanism of grain aliasing are still a bit

elusive to me. I have some familiarity with Fourier transforms, Nyquist sampling theorem, aliasing, convolution

theorem, etc., so I should be able to understand theory of grain aliasing if it can be clearly explained.

 

Edward, thanks for posting the example of grain aliasing.

[williamchuttonjr]

The comments about the discreet FT and Nyquist sampling are right on the mark. One thing to keep in mind is that in images we are talking about spatial frequencies (as opposed to harmonic frequencies such as sound waves).