Images, Purpose, Effort

[Peter Rauch]

Jim Croft comments and inquires:

>From: jrc at anbg.gov.au (Jim Croft)
>Subject: .tiff to .gif conversion
>To: unite at mailbase.ac.uk (unite listerver)
>Date: Fri, 2 Jul 1993 21:57:22 +1000 (EST)

>At the Australian National Botanic Gardens, we are extending our
>public information services to include images of Australian native
>plants, their habitats, and other aspects of Australian botany.

>Our gopher (155.187.10.12) delivers a meagre sample of TIFF compatible
>photographs, and we would like to expand it to a working tool for
>research and investigation.

 ...questions about specific file formats and conversions among them
    deleted; PAR...

>Or are there any other recommendations you would like to offer on the
>storage of colour images (mostly from 33mm slides) and their
>delivery over the internet?  What format is going to be most widley
>acceptable?  What compromises should we be prepared to accept on
>resolution as it impinges on data file size?

>cheers
>jim
>Jim Croft           [Herbarium CBG]               internet: jrc at anbg.gov.au
>Australian National Botanic Gardens                  voice:  +61-6-2509 490
>GPO Box 1777, Canberra, ACT 2601, AUSTRALIA            fax:  +61-6-2509 599
>______Biodiversity Directorate, Australian Nature Conservation Agency______


Jim,
Access to useful digital images of biological themes and interest is a
topic dear to my heart. But obvious solutions are not near at hand, for
a lot of reasons. (Or, lots of "solutions" are near at hand if you punt
on most of the issues and just toss anything and everything out on the
net.)

Your question asks about formats, but I'd like to highlight some other
issues that bear on the use of images, as described in your first two
paragraphs.

Among the more conspicuous, fundamental issues for me, around which I
continue to observe confusion, compromise, and costly (but perhaps
unnecessary) choices, include:

- scholarly use of the images
- tools to acquire, manipulate, and share images
- rights to use the images

Taking these issues in order, I'll identify a few points that emerge
from their consideration, and which would/should/could influence any
decision process about how to provide images.

Scholarly Use of the Images

I want the technology that will permit scholars to extract the abundant
and substantive information about the subject matter contained in
images.

The main issue here seems to be that little discussion/study of how
images (and I'm not talking satellite data, or medical imagery, or
other high profile infomation) can be used for scholarly analysis of
the subject matter content in the images. We all "use" photograph and
slides for many purposes, but little has been done to examine and
describe either the precise way in which (ecological, biological,
environmental) imagery information is acquired, represented and used
(again, excluding consideration of the specialized imagery from
satellites and other remote sensing technology, for which lots of work
is available).

E.g., photographic slides are often used in "slide shows" for classroom
instruction, seminar/workshop/conference presentations, illustrations
in publications, etc. We all recognize poor presentations/images, and
enthuse over the great ones. We appreciate the understanding conveyed
when an image and presentation is of high quality.

Well, how do these notion of quality, understanding, utility and
scholarly content (i.e., the intellectual, aesthetic, factual, gestaltic
content), and others characteristics of images and their use (what
are they?) get conserved when the medium of representation is
digital/network/computer/software/etc., rather than the more
traditional film/print, or real object?

The problem I often encounter is that discussions and decisions usually
revolve around the "Wow, look what I can do! Digitize a photo!" and when
things get more serious, around the realization that the first images
captured are often of poor quality and very lossy in content, the
focus turns to the idea that "Well, these are great for browsing and
finding subjects of possible interest, after which we go to the real
source materials [i.e., photos, or the objects/places for which they serve
as surrogates]."

I would like to see more work done on how one captures more of the scholarly
value to be had in imagery. For this, we must start asking the right
questions (and consequentially, spending a lot more time, talent and money
on the topic).

This sort of leads into the next topic.

Tools to Acquire, Manipulate, and Share Images

I want scholars to be provided with tools that permit them to work at
their full knowledge capacity and work pace, to use/study images. In
particular, I want them to have instant access, instant response times,
and a broad spectrum of information seeking and manipulating tools at
hand to perform and share their studies.

An obvious part of the reasons why one observes the common scenario
described in the previous section is because the technology of
acquiring digital images (esp. from photographic source material) is
only recently affordable at "high" levels of resolution and
quality. "High" quality and resolution image capture is still relatively
expensive, and experience with it is still very modest. Sharing and
otherwise using the captured images is also fettered by the high costs
of the current commonly available technology. So, it is somewhat to be
expected that things are where they are. But, times they are achanging,
and we should start to ask some of the more interesting questions about
use of images.

Subject content questions invariably lead to issues about the level of
resolution and depth/range/color attributes required to achieve a
particular objective in using the digital images. And the answers often
end at the realization that bandwidth, storage capacity, cpu speed, i/o
channels and speed, display devices color control/matching, software
capability and performance, --in other words, almost every conceivable
part of an image-use enabling architecture-- are still sorely deficient
and hostile to our objectives.

So, we often "punt" as I noted above. We do with what we've got in hand,
and can easily afford, which til now has been most often at the very,
very low end on the scale of serviceability.

You often hear the recommendation "Digitize at the highest resolution
possible" [read affordable]. This is (good) advice that attempts to address
the issues described here without having to really get into being specific
or knowledgeable about the particular subject matter and its potential uses.

Those who can afford to indulge in the leading edge technology (in all
the dimensions listed above) need to get more aggressive about investigating
the questions of how far can we go with the use of images, and in
particular, as many of us as can, should push to capture images at
high quality marks (it is my contention that for many exercises, the
most costly element is the labor in getting to the film storage site,
assessing, selecting, cleaning, and otherwise preparing the film,
scheduling and setting up the digitizing system work session, returning
the film materials to storage, cataloging the digitized image information,
etc, and not on the actual digitizing at low or high resolution (color
manipulation might have a mitigating influence on this argument, I admit,
but even that is getting easier).

So, we shouldn't be trying so hard to save time or storage (tape is
cheap, and good enough for deep storage of images until things get
better; in the meanwhile, we can make some of those images available
for exploring and experimenting with all these issues).

Of course, if I make my high quality images available to the world,
the next topic becomes interesting.

Rights to Use the Images

I want to put to use every ounce of intellectual talent we have on this
earth today, so that we can come up with some fast and solid solutions
to the problems that are leading to the total destruction of our
environment. I want scholars, environmentalists, and anyone else with a
bright idea to share, to have access to information without
limitation.  The right to use information will be the great mediating
force that affects the degree of access.

This topic puts a lot of emotion in what is otherwise a more objective,
technical subject (hmmm, is the topic of scholarly uses objective or
emotional?).    ;>)

Terms like "intellectual property rights," "copyright," "ethical use,"
"financial impact," "rights to first use," "stategic advantage,"
"censorship," and other forms of "controlled access," "my pictures,"
all load the cannon with shot for many a discussion.

Lawyers and businesses know a lot about these issues, and what they'd
like to do to address them. What do scholars have to say about them?
Some traditional forms of expression, such as book publications, film,
and shared numeric and textual data have raised all these issues for
the scholar. But, these traditional media have never had the
opportunity to be as readily copied, manipulated, and made more
valuable (both from value-added packaging, and from value-adding study
of the material) as will digital scholarly materials, including images.
The challenges to supply information in ways that control for all these
"rights" is probably going to hamper the rapid, widespread distribution
and access to images more than any strictly technological barriers to the
simple production and transmission of valued images.

Well, Jim, so what was your question.....?
Peter