Digital BW, The Print

Hi Todd,

> Austin, there was one point in our last conversation that you
> promised you'd
> clear up for me but did not, perhaps you could do so now? You said two
> things that seemed to contradict each other:

I'll do my best, as I have a somewhat clear head right now...could last only
about 15 minutes though ;-)

> A) any DR can be defined by two bits of data. First off, why two bits and
> not one.
> I'm assuming you were speaking to what you consider "range" when
> you said that; If so, why not one bit of data: dmin-dmax?

A very good question.  One bit only gives you ONE threshold.  If 0, the
value is either greater than or equal to the "threshold", and if 1, the
value is less than the "threshold".  You need two bits to have the three
states you need.  Where 00 is value is greater than the HIGH threshold
(dMax), where 11 is the value is less than the LOW threshold (dMin), and 01
or 10 is any value in between...and that gives you a dMin and a dMax
representation.

Just for those that are wondering...this discussion has to do with the
representing dMax and dMin (density ratio values).  It started out as a
typical question, what do N bits from the scanner's A/D have to do with
density ratio values...and what this example shows, is you can represent ANY
density ratio RANGE with three states.  That is separate from any discussion
of dynamic range.

> B) DR is defined
> by the number of discreet steps of data. That seems in direct
> contradiction
> to A.
>
> Could you explain the relationship of A to B?

A is a representation of "a" set of density ratio value, and B is dynamic
range, as B has steps.  This is at the root of the entire discussion Martin
and I are having.  Showing the difference between density ratio values and
dynamic range, and that they are not the same.

We need to stop using DR, because it could be Density Range (or Ratio) OR
Dynamic Range ;-)

Did that answer your question?

Austin