On Sun, 09 Jun 2002 00:12:26 +0300, Kari Kivisalo
<pro### [at] luxlabcom> wrote:

>Glen Berry wrote:
>>
>> That sounds interesting, at least for simulating electronic scans and
>> digital camera images. Have you ever given any thought to simulating
>> film grain?
>
>Hmm. Isn't the noise from the film grain? What I ment was that the grain
>visible at 6000dpi looked very much like blurred Gaussian noise :)

If we are talking about a color scan of a negative or transparency,
there will often be at least three types of noise in the final scan. 

First, there is "film grain" (or "grain clusters" or  "dye clouds", as
the case may be, depending on the type of film.)

Second, are the electronic artifacts contributed by the scanner
itself. This can come from the image sensor, the analog amplifiers
that feed the digital converter, or the converter itself. Technically,
all these would contribute a little "noise" to the scan, although one
may dominate the others, for noise contribution. 

Third, there is sometimes a film grain aliasing effect in a scan. This
is where the resolution and pixel pattern of the film scanner creates
a sort of "moire" effect with the grain of the film, and exaggerates
the apparent size of the film grain. This would be remotely similar to
the exaggerated artifacts one sees when scanning half-toned materials
on a flat-bed scanner. This effect is very dependent on the DPI
setting of the scan, and on the size and nature of the particular film
grain being scanned. It seems to be most often noticed on high speed
films, but could appear with any type of film.

So when you look at the scan of a negative or slide, there are several
potential sources of noise. Actual film grain looks a bit different
than what you have done with these images, although these images do
"hint" at film grain.  

Some photographers will actually choose very grainy films, and process
them in ways that maximize the effective perceived graininess of the
film, for an artistic effect. Take a look at some images like that, if
you want to see some film grain up close and personal.  :)



Later,
Glen

7no### [at] ezwvcom     (Remove the numeral "7")

Post a reply to this message

Glen Berry wrote:
>
> If we are talking about a color scan of a negative or transparency,
> there will often be at least three types of noise in the final scan.

Yes, but the combination still looks like Gaussian noise. Some  of
the samples I found actually demonstrated noise aliasing which seemed
to be a mystery to photographers.


_____________
Kari Kivisalo

Post a reply to this message