Archived posting to the Leica Users Group, 2004/07/03

[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]

Subject: [Leica] Fungus now filters
From: jean.louchet at inria.fr (Jean Louchet)
Date: Sat Jul 3 06:16:12 2004


The reason for aliasing is sampling. There is some math behind that (the 
Nyquist-Shannon theorem) but to make things simpler, I would say that 
aliasing comes from the fact that in a CCD, the photodetectors  are only a 
part and not all the CCD's surface. These photodetectors are arranged in a 
periodic pattern, thus periodic objects (like patterns on fabrics etc.) 
will interfere with the diffrent period of the photodetectors and make the 
"moire" effect. Just like in music, two instruments not in tune that try 
to make the same note will produce beats.

There can't be real software remedies to aliasing. An digital image with a 
given sampling period has to be reconstructed into an analog, displayable 
image by interpolating between the samples, and in order to interpolate 
the only "reasonable" hypothesis is that the reconstructed image should 
not contain objects with higher spatial frequencies than what the sampling 
interval is able to give. So-called software remedies to aliasing just 
consist in eliminating some unwanted frequencies.

There can be hardware remedies to aliasing. 

The first one is to make CCD sensors with the highest possible proportion 
of their surface occupied by the photodetectors. This is in the hands of 
the microelectronics industry.

The second ne is to get a lens which eliminates the spatial frequencies 
higher than those compatible with the given CCD. In other terms, to use a 
lens that gives images smooth enough to eliminate them. Cheap lenses are a 
solution - this is probably why expensive digital cameras are often more 
prone to aliasing than cheaper ones. Or use a filter (I mean, not a 
software filter but a piece of glass) that smoothens the image.

Maybe some day electronic sensors will copy silver halide films and have a 
random distribution of photodetectors on the image surface, which would 
solve the problem... 

Jean

> Date: Fri,  2 Jul 2004 17:15:10 -0500
> From: George Lottermoser <george@imagist.com>
> Subject: Re: [Leica] Fungus now filters
> To: Leica Users Group <lug@leica-users.org>
> Message-ID:
        <r01050300-0921-5FD7797FCC4B11D8B2B0098FCC95711B@[66.239.168.102]>
> Content-Type: text/plain; Charset=ISO-8859-1

>Dan C7/2/04

>The important question is this:   Is the Kodak (and future digital R 
>back)
>better off with or without the anti-aliasing filter?

>Ideally - Choice.
>This moir? problem keeps coming up in the high end digital capture world, 
>especially in the
>fashion world where we add both fabric weave and printed patterns to the 
>array. Back to film
>anyone ;-)  ?
>
>Fond regards,
>
>G e o r g e   L o t t e r m o s e r,    imagist