[filmscanners] Re: Dynamic range

["Todd Flashner" <tflash@earthlink.net>]

on 8/27/02 9:26 PM, David J. Littleboy wrote:

>
> "Roy Harrington" <roy@harrington.com> writes:
>
>>>>>>>>>>>>
> The notion that "Dynamic Range is not based on the actual values the data
> represents" is so ludicrous that I'm just going to have to bow out of this
> fruitless endeavor.  I had hopes that you might be able to "get" it but ...
> <<<<<<<<<<<<
>
> I think we need to retreat to basics:
>
> Density range: the range of densities the scanner can recognize.
>
> Dynamic range: How finely the scanner chops up ("resolves") the density
> range.
>
> Which means that dynamic range has very little to do with the physical
> quantities the values represent. (More accurately, the meaning of the
> individual values are determined by the meanings of the endpoints and how
> finely the actual range is chopped up, but it is the density range that
> determines the meaning of the endpoints.)
>

Please help me understand this.

I thought what determined the DyR of a scanner was its ability to accurately
read into dense film. IOW, to discern low voltages (readings through high
density film) from noise. IOW, how far from clear film into density it can
distinguish -- not how finely the image gets chopped.

I understand that if you don't have enough "chops" (bits) your scanner will
run out of encoding room, so to speak, and won't be able to continue
assigning values to the sequentially lower voltages. Thus insufficient bit
depth can LIMIT DyR. But this is just incidental to how scanners work, not
what DyR is. Thus, simply adding more bits to a noisy scanner may do nothing
to increase DyR. Therefore, isn't a scanner's DyR in fact a function of
being able to discern shadow detail above noise? IOW, the "distance" it can
record from one endpoint (clear film base) to the other (film dMax)?

If you designed a scanner that could only handle film with a DyR of 2.2, but
it could do so by spreading that density across 16-bits, would it have a
higher DyR than a scanner that uses 14-bits to accurately record film with a
DyR of 3.9? I think not, not by any conventional use I've seen of the term
DyR.

If I'm correct, here is an analogy. We have two rulers, one a 12" marked in
1/16" increments, the other 24" marked in 1/4" increments. If we understand
DyR to be the ability to measure further from one endpoint to another, the
longer ruler has the capacity to record the greater DyR (24"), not the one
with the greater resolution (12").

Todd Flashner


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