Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Re: Density vs Dynamic range>AUSTIN (2a)

  • To: lexa@lexa.ru
  • Subject: [filmscanners] Re: Density vs Dynamic range>AUSTIN (2a)
  • From: "Isidoro Orabona" <isiora02@hotmail.com>
  • Date: Tue, 18 Jun 2002 19:31:36 +0200
  • Unsubscribe: mailto:listserver@halftone.co.uk

Hi Austin,

>Austin wrote:
>the sampling is still done as DC.  The digitized signal (all the samples)
>does not know anything about the A/C nature of the signal.

I agree. More precisely the digitized signal does not know anything, period.
It is a sequence of numbers. Only software or a circuit knows what these
numbers means (maybe ;-) ).

> > I claim that all useful signals are AC, or that can be modeled as AC
> > signals.
>For audio, yes, but not for imaging.

It seems strange, but this isn't completely true. I can model (and it's
useful to do it) an image as a signal. Your discussion is about dynamic
range and then about noise. Well, let me to ask you:

Do you agree that moire is noise ?
Do you agree that grain aliasing is noise ?

If yes, then ask ourself how discover and analyse it. We must do it, because
we are interested to the whole scanned image and not to a single pixel.

Both phenomena are related to spatial arrangement of CCD pixels.
The first is made worse by the Bayer mask.
Now, changing our mind, we make a model that reassemble the matters as
sampling a AC signal. The AC signal is represented by the changing of
luminance with distance. Let fs to be the sampler frequency (2700 or 4000
dpi) and let fmax to be the maximum spatial frequency on the medium
(6000-8000 dpi for high quality slide film). When scanning, all the
frequencies f > fs produce aliasing, in the same manner as sampling a
electrical Alternative-Current signal.
So now I can apply to this signal all the signal processing know-how and
analysing tools as Fourier Transform, Cosine Transform etc.
Incidentally, as sampling AC signal, we can reduce or eliminate the aliasing
noise applying a low pass filter to signal before sampling. This is easy
with electrical signal, but is difficult with optics. The "glass" on the
surface of a digital camera CCD has this function too. It don't work with
scanner CCD because this is a single stripe of pixel (linear or trilinear

>Though A/C (frequency) can certainly be ascribed to imaging, that has
>nothing to do with the sampling, digitization, and reproduction of the
>image.  Also true for audio, as I said above.  I do agree audio is an A/C
>signal, no doubt.

I don't agree. See above.


Entra a far parte del pi grande servizio di posta elettronica del mondo con
MSN Hotmail. http://www.hotmail.com

Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 


Copyright © Lexa Software, 1996-2009.