Re: "Pepper grain" (was Minolta Multi Pro)

[Arthur Entlich <artistic-1@xxxxxxx>]

There is a difference between what causes aliasing, and how it may be 
represented in a final image after processing, which is the point I am 
trying to make here.  You will note I stated "emphasize aliasing" NOT 
create aliasing; hence the difference in meaning.

I am aware how Nyquist's equation enters into this discussion, but that 
is not how I interpreted Byard's question. I worked with sampling error 
and filtration in audio sampling nearly 15 years ago.

Methods of upsampling or, if you prefer "interpolation", can determine 
what aspects of an image are emphasized.  Some methods downplay smaller 
details of low contrast, and might tend to minimize aliasing artifacts. 
  Other methods may emphasize or define them further.

Lastly, it is still unclear if pepper grain is about aliasing.  It may 
simply (or not so simply) be due to certain lighting conditions or 
characteristics of CCD sensitivity tending to emphasize an aspect of the 
structure of some films which under other applications is not visible.

Perhaps if you weren't so intent on trying to make me "wrong" you might 
have understood the intent of my posting.

Art

rafe.bustin@verizon.net wrote:

> On 5 Apr 2002 at 1:44, Arthur Entlich wrote:
> 
> 
>>
>>byard pidgeon wrote:
>>
>>
>>>WOuld also be nice to know if the prints were made with condenser or
>>>diffusion.
>>>
>>>Also, I wonder if the pepper grain and bubbles show as much if one enlarges
>>>the neg/slide to final size during the scan, rather than in an image editor?
>>>
>>
>>Are you referring to upsampling via an image program, versus using the 
>>native scanning optical scan resolution?  I would expect upsampling 
>>would tend to emphasize the aliasing, but that might depend upon the 
>>method of upsampling used.
>>
> 
> 
> Gee, for a moment, I thought you "got it."
> 
> Aliasing has to do with the bandwidth of the input, relative 
> to the sampling rate.  Nothing more and nothing less.
> 
> Any input with a frequency content beyond (sampling rate / 2) 
> leads to aliasing.  That's all there is to it.
> 
> "Upsampling" is post processing of the sampled data.  It's usually 
> just a euphemism for interpolation -- taking a best guess at what 
> the data "might have been" with a higher sampling rate.
> 
> 
> rafe b.
> 


-
Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.