Now that I've moved to a machine with a 24-bit monitor, here's a little
bit about the sorts of things the strong color contrasts in the Hallas' photo
reveal from simple inspection.  It will help in following my chit-chat to have
a small-scale naked-eye star atlas handy.
     First there's a striking lesson in star temperatures and their effect on
the interstellar medium.  Have a look at the circular glows around the stars in
upper Scorpius:  gold, blue, and "raw meat" (I'll call this pink), and maybe
two glows with blue/pink mixed.  We all know that Antares is a cool supergiant
star, with essentially no ultraviolet photons to fluoresce the gas in the
rho Ophiuchi cloud (the bright nebulae in the head of Sco, and the three-forked
dark clouds extending to the northeast).  So all Antares does is light up the
dust with its dominant yellowish color.  The stars flanking Antares, sigma and
tau Sco, are early-B type stars, hot enough to fluoresce the gas, and (bingo)
their clouds glow raw-meat-hydrogen-alpha pink.  Just above Antares is the
blue cloud around rho Oph itself.  This is a pair of stars of type B2, just
barely _not_ hot enough to fluoresce the gas, so we see a reflection nebula.
Likewise up above, stretching on either side of nu Sco (type B2IV), another
blue reflection nebula.  beta Sco appears to have no nebulosity associated
with it, but both delta Sco (B0.2IV) and pi Sco (B1V) are hot enough to give
at least some emission.  Their clouds look somewhat purplish on my screen, so
perhaps the reflection component is strong enough.  Just from the colors of
the nebulae and the spectral types of the involved stars, we now know where the
cutoff is for an emission object---you gotta have stars earlier than B2.  The
other lesson that's obvious is that dark/reflection/emission nebulae are _not_
different types of objects in the astrophysical sense:  the make-up of the
clouds, their composition, is _identical_, and the only differences are the
temperatures of the stars (if any!) in them.  No stars and dense = dark nebula;
stars of type B2 and later = reflection nebula; OB stars = emission nebula.
All three are exemplfied in the same cloud near rho Ophiuchi; here's the sky
telling us how the World works!
     Now the rho Ophiuchi cloud is only a couple hundred parsecs away, and
any interstellar extinction you see in the area is caused by it, not something
in the foreground.  Thus the pink color of the emission objects represents
"unreddened nebulae".  Look around elsewhere in the photo at the various
nebulae:  the big region south of the Norma starcloud (at the bottom of the
picture) is a lot redder---but could this be because that area was reddened
by Earth's atmosphere?  (As the joke goes, most of the "interstellar medium"
is within the first mile of the telescope.)  The zeta Sco nebula is a healthy
pink (unreddened), but our friends NGC 6337 and 6357 are also dark red, and we
know from the recent 'amastro' chat that these objects are indeed obscured
behind those dark clouds that surround them:  reddened red nebulae.  But just
next door, that nebula just west of M6---which I never knew existed---is bright
pink again, and unreddened.  The main cluster there, NGC 6383, contains
several O and early-B stars, which are doubtless responsible for the UV
photons.
     One can play the same game with star colors, in this case the grand mean
color of the starclouds.  It looks as though Tony and Daphne have rigged the
color balance so that roughly solar-color stars are white, which if there are
enough stars to make the image dense, you see as a neutral starcloud color.
The region near 45 Oph, just left of center in the photo, shows a nearly
unobscured part of the galactic bulge; similar unreddened regions appear near
M7, and from there off toward the east edge of the photo.  (Indeed, a nearby
cluster of galaxies is visible near 45 Oph, since we can see (barely) right out
past the center of the Milky Way here.)  But now look at the color gradients of
the starclouds in the obscured regions, notably along the 'back' and under the
'rump' of Richard Berry's "galactic dark horse" in the upper left corner.  Wow!
There aren't any blue stars there unless they're in the foreground, limiting it
to a few of the naked-eye stars.  This is right in the direction of the nucleus
of the Milky Way, which has been measured to have extinction of at least 30 and
possibly 50 magnitudes (very patchy in any case), so sure enough you don't see
it---even a supernova would be completely blocked.
     The large network of dark material in the bottom of the photo seems to
form a ragged circular outline encompassing the zeta Sco region and the Norma
starcloud:  not by chance?  (I don't know.)  I recall Bart Bok writing
somewhere that the filamentary nature of some of the nebulae here were
suggestive of a supernova remnant, but I think the characteristic emission
lines are not present.
     Another thing worth pondering is the differences in spatial scale in this
image---always a tough problem in conveying astronomy to the uninitiated.
The upper Scorpius stars and circular nebulae are very much in the foreground.
Normally I (at least) think of a star's region of influence being always quite
small on the sky, no bigger than the star images in this small-scale photo.
But here we see that it's actually a few degrees across for these guys that
are even 250 parsecs away.  But down around zeta Sco, something like seven
times more distant, the affected area is tiny by comparison (a factor of 50
in area if the relative distances are as I said).  The rho Ophiuchi clouds are
at around 250 parsecs, but the unobscured starclouds of the galactic bulge are
something like 5 or 8 kiloparcsecs away (just considering stuff closer than
the galactic center).  Antares, a couple hundred parsecs; M4 sitting right next
to it, two kiloparsecs!

\Brian