[filmscanners] RE: [filmscanners_Digest] filmscanners Digest for Tue 3 Sep,2002

["Austin Franklin" <darkroom@ix.netcom.com>]

Jonathan,

> I have been reading the ongoing discussion regarding dynamic range and it
> appears to be approaching the discussion about how many angels can sit on
> the head of a pin.  There is a practical side of dynamic range however and
> that is this, " How much information can I get out of the shadow areas of
> Velvia film.  This is the real world test.   When I look at fuji's film
> databook, it appears that the Dmax of the green sensitive layer is closer
> to 4.0 than 3.5.  So in order to get shadow detail out of this area  you
> need a scanner with dynamic range.

That's actually one of the issues...you need a scanner with DENSOTY range,
not dynamic range....you can capture any density range you want with any
dynamic range...they are not the same...but, if you want the DETAIL in the
shadow detail, yes, you do need DYNAMIC range.  That's one of the points.
Dynamic range is about tonal resolution, and not about ability to capture a
density range.

> I don't really care how closely you
> divide up the numbers somewhere along the line I need a scanner
> to get that
> detail.

Exactly, and that is what dynamic range is.  Those who equate it with
density range are not understanding that they are different, and
why...again, one of the big issues, as there are a lot of people who,
through misinformation, have been led to believe the two are the same.

> It seems to me that drum scanners did the job reasonably well using 8 bits
> per channel.

What drum scanners scan at 8 bits/channel?  Yes, you can get a resultant
data output of 8 bits, but the actual scanner scans using much higher bit
depth, then applies the setpoints and the tonal curves to the high bit data
to give you the 8 bit data file...but the point is, you have to scan using
higher bit depth, or you lose the resolution you are "seeking".

> If you have a scan with a lot of noise in the shadows, who
> cares whether the noise is divided into big chunks (8 bit files)  or tiny
> chunks (16 bit files), it is still noise.

Correct, but 8 bits is insufficient to provide both the density range you
want to digitize (because of how scanners are designed) and the dynamic
range you need to manipulate the image.  Noise is typically much more than 8
bits.

Austin

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