Digital BW, The Print

Hi Roy,

I am going to just snip indiscriminately and not say I did so ;-)

> From my point of
> view, your way made less intuitive sense than mine, but I perserved and
> tried to understand what the end result meant.  I.e.  could I
> conceptualize
> the quality or property so defined.  Mine is certainly pretty
> simple (in fact
> I think you believe its too simple).

Your statement of "the ratio of maximum contrast to minimum contrast" works
fine for me, where maximum contrast is (dMax - dMin), and I believe because
contrast requires two references, minimum contrast works all by it self with
no qualification, and that minimum can lie anywhere within the density
range.

> Anyway, given all that, I looked at a lot of the imaging
> literature and the
> descriptions of dynamic range all seem to agree with my view.

A LOT does, but not all.  That's why I am saying that it is an "old",
obviously deeply rooted, misconception...

> So here we are.  You've pretty much blown off all the "experts" in imaging
> saying they are not engineers and don't understand

Whoa there, leap of something...  I have not blown off ANY "IMAGING"
experts, but pointed out that the people referenced are PhotoShop EXPERTS,
which does NOT make them imaging experts.

> -- only audio experts
> know what this is all about.

Well, obviously some imaging experts (real ones who do hardware design) seem
to know all about it, and...it is also a very common concept in mechanical
engineering...where they view it exactly the same as I have described it (my
undergrad was ME, BTW).

> Minor question: If dynamic range is a measure of how many gray tones
> there are, how come its in dB and not just an integer.   Why doesn't
> it just say there are 237 grays in this image??

Good question.  As I've said, number is not a good term.  It (meaning
dynamic range) is a precept for "expanse" (which isn't a very good word
either, since dynamic range doesn't define the bounds, just what can go on
within bounds).  I really do have to get a better way to word this...but you
are right, calling it a hard number is not the intent, at least in the
analog world.  It IS a hard number in the digital world, simply because the
minimum discernable happens to be a hard number.

> Not that it matters much, but BSEE MIT, MSEE Stanford, My field has been
> mostly computer science.  Personally designed and personally
> written a ton of code.

[SIDEBAR] I used to give a lecture years ago called "Is Computer Science?",
later also one titled "Is Software Engineering".  We can get into that
discussion in some other lifetime ;-) They weren't as "bad" as you may take
their titles to be.  Their point was to separate out programming from
engineering, and get management types to understand they are not
(necessarily) the same.  This was at a time where programmers were called
"programmers"...and programming wasn't really part of any engineering
discipline, actually, any schools that taught programming did so as a part
of some other curriculum.  If they did have a degree program in
 "programming" it was called "computer science" and typically part of their
mathematics department, and not associated at all with engineering.  Also,
many astronomy and music graduates were now hired as programmers, and some
of them made damn good programmers in fact.

> You can probably claim more dynamic range usage in your specific area,
> but my diversity has served me well.

My background is also quite diverse as well...as well as in senior
engineering management for one a quite large computer company, and my share
of many startups, some quite successful, and some not.  I've been on both
sides of many coins.

> So you're asking how I think the English definition of "dynamic"
> is relevant to
> the technical definition??  I thought we were talking about a technical
> definition that was explicit and precise technically.  No?

Well, dynamic is a concept/understanding, and I believe it is entirely
applicable, and technically appropriate, to discuss a system in terms of it
being "dynamic".  Everyone else does ;-)

> Actually, the word I've equated "dynamic" to is "useful".  In
> ranges such as
> exposure, Zones, density.  The useful part of those ranges is where the
> image information is.  The part of the range is "dynamic", the
> parts outside
> are dead, useless, have no image.  Hence "dynamic" rather than "dead".
> Its just an English language analogy, but then that's all you asked for.

Well, useful seems to be more related to the endpoints/bounds than dynamic
is.  Dynamic, in and of it self, doesn't really care about the bounds,
except in calculating it, but after it's calculated, you can throw out the
bounds.  Useful, on the other hand, to me at least, seems to require the
bounds to be meaningful.

Two systems can have the same dynamic range, but have entirely different
"useful" ranges.  For example, positive and negative film.  The "useful"
range of positive film is much larger than the "useful" range of negative
film, if you are talking about density as the property...BUT...there are
more tones compressed within the smaller "useful" range of the negative film
then there are in the larger "useful" range of the positive film.

> I guess we're pretty much at an impasse.  Too bad.

For now ;-)  Well, I have more to come on this, as soon as I'm finished with
my discussions with some "others" on this subject.  So far, so good (for me
that is ;-).

> Sorry about any grumpy tone to this. I'm frustrated as I'm sure you are.

Why, thank you, I appreciate that very much.  I've tried to not be
frustrated...but I have been, no doubt...I'm sorry if I've come across as
arrogant and/or grumpy.  Please take any arrogance as confidence, not as
arrogance, and grumpy as, well, simply grumpy...but probably due more so to
lack of sleep with a sick dog, and two small children who want "up" at 11,
1, 3... ;-)

Regards,

Austin