ABCs AND PQRs OF CCDs
-------------------
John Pazmino
NYSkies Astronomy Inc
nyskies@nyskies.org
www.nyskies.org
1992 February 1
[This was a two-part article published in 1992 January and February.
Here they are assembled into on piece under the later date.]
One highlight of the [fall 1991] AAVSO Cambridge meeting was the
CCD workshop convened on Friday afternoon 25 October [1991]. AAVSO
from time to time stages workshops on new technologies or major
situations in observational astronomy. This workshop was the first on
CCDs and a second one is planned for the spring 1992 convention in
Columbus OH.
The workshop was a survey of CCD apparatus for the home astronomer
with the potential for use in photometry. The discussions were carried
in several short papers and demos and by a panel fielding questions
from the floor. The lecture room, at AAVSO's HQ, was filled to its 60-
seat capacity.
A CCD, standing for 'charged couple device', is a computer chip
that converts light impinging on it into a computer digital signal.
The chip has a grid of cells, or pixels, that cover the focal plane
area of a telescope. A typical CCDchip for home use has 192 X 256
pixels on a patch about 5mm square. Each pixel receives light (and
other EMR) from a tiny but specific place in the focal plane and the
whole image is then assembled by computer from the signals from all
the pixels. The brightness of a pixel in the reconstituted image on
the screen is consonant with the signal strength from the
corresponding pixel on the chip.
Because the image is a stream of computer signals, it can, besides
be displayed evanescently on the screen, be captured into a discfile.
In this form the image can be sent by mail to other astronomers, sent
by telcomms to other computers, saved for later review, and even
combined into other images. Also because the image is in a computer
form it can be massaged in astounding ways: remove flaws, correct
distortion, add color, sharpen detail, to cite only a few.
While the CCD technique has been employed at observatories for
about 15 years, it was barred from home use by the necessity for heavy
computing power and the overall expense of the apparatus. The
breakthru for home enjoyment of CCDs came in about 1985. By then the
personal computer became cheap, powerful, and standardized enough to
be regarded as a home appliance. Today CCD equipment is sold without
the attendent computer, for anyone entering the field reasonably has
one.
Home astronomers during the 1980s tried to acquire CCD experience
by tinkering with camcorders and security/surveillance cameras. Results
were encouraging yet half-baked and, in general, work was abandoned
after even valiant and heroic efforts.
Now there are two main suppliers of home CCD equipment: Santa
Barbara Instrument Group and SpectraSource Instruments. The former,
from Santa Barbara CA, beefed up a satellite star tracker into a
workable CCD outfit. The latter, from Westlake Village CA, developed
an allnew CCD apparatus under the brandname Lynxx. Both equipages come
with the CCDchip encased in a protective head, computer software,
power supply, and assorted cables and fittings.
In its simplest elements the CCD captures a 'television' picture
of the focal plane and the computer displays this 'television' image.
It is on the whole easy to capture good and pretty pictures of the
celestial objects. It is far simpler than regular astrophotography
CCDs are to regular cameras as camcorders are to movie cameras.
One telling feature of this workshop -- and of AAVSO meetings in
general for the last 8 or 10 years -- is the routine tapings by
camcorder. In all the prior history of AAVSO, back to its inception in
1911, tho cinecameras were readily available, there are pitifully few
films of AAVSO procedings! One surely will eat guano before going back
to cinefilm again, right?
In the early days of CCDs there were many computer systems, each
out of touch with the other. Many astronomers were stymied when they
tried to use images or accessories from a one system with an other. By
now everything is based on the IBM system. The few remaining off-IBM
systems accommodated to IBM with emulator boards or software. There
are few genuine complaints of system incompatibility anymore.
AAVSO is into CCDgraphy, as it may be termed, for its potential in
photometry, the assessment of the illumination received from the
stars. AAVSO has a long-running program of electrophotometry employing
analog photomultiplier heads and circuit electronics. In fact,
electrophotometry in the United States was largely promoted and
developed thru AAVSO.
But this photometry never matured and remains today a wildly
capricious endeavor. AAVSO accumulated an alltime total of barely
5,000 electrophotometric observations, versus over six millions from
eyeball inspection of the stars.
Despite the boon and bonanza of going on the road of CCD, many and
major issues were uncovered in the workshop. Right off the bat, the
astronomer must be well grounded in computer operations. It can be
argued that this is a skill not ordinarily found among the public.
Yet, given the brief history of home computers, there are orders more
people litterate in computers than in electronics. Or, for that
matter, ordinary photography. Computers are taught (however ineptly)
in the primary schools; electronics and photography are not.
A truly solidly mounted and stable telescope is required, just as
for photography and electrophotometry. One can not make do with a rig
assembled for each session on a rooftop. Some believe that with the
very short exposure (integration) time a less-than-substantial mount
is adequate. Not so; the image is blurred exactly as for photography.
The CCDchip has a tiny area, around 5mm square. This is an awfully
tiny part of the focal plane! The scope must have means to aim and set
and track on that small area. Even with an offaxis or a flip-mirror
guider the task of getting the target onto the CCDchip can be
positively maddening
An other consequence of the tiny receptor in the CCDchip is the
allegedly lousy images displayed on the computer screen. Everything is
fuzzy and the stars are gross blobs. What's going on here?! The piece
of the focal plane captured by the chip is enlarged and zoomed to fit
on the screen.
So the Ring Nebula, to pick one example, may stand perhaps 20
degrees across. To achieve this same angular extent for the Ring
Nebula in an eyeball view thru the telescope requires a optical
magnification of about TWELVE HUNDRED! (The Ring is about an arcminute
in true diameter.) Go pump up a homesize scope to 1200 power and how
does the Ring Nebula look? Lousy.
The computer is an integral part of the CCD circuit and ideally it
should stand at the telescope. The astronomer must work the computer
during the CCDgraphy session and the CCDhead must be cabled to it.
This is often not practical or possible if the machine is a regular
desktop. A new other computer may be required, a laptop, small and
compact and lightweight.
The images are stored on magnetic media, which are vulnerable to
damage, both physical and digital. The usual media available to home
astronomers are not viable for true archiving of data. Also, there is
the fear, not unrealistic given the recent experiences of NASA and
USGS, that in future years the media will be undecipherable. For the
machines that write and read them will be irretrievably abandoned.
There are many and incompatible file formats for the digital
images. The Santa Barbara files, for instance, are useless for the
SpecraSource software. There are emerging some utilities to convert
among formats, but in any conversion some information is lost.
So far the operator is at the mercy of the equipment company to
furnish the software. There is not yet an independent software
industry for CCDs. Never the less, there is a fount of general purpose
image processing applications that can handle selected file formats.
Already astronomers are pressing into service applications developed
for the various space missions.
CCDgrams are black-&-white images. Each pixel holds one of the,
normally, 256 grayscale values. In computer terms, each pixel occupies
one byte of memory. All those gorgeous colors in textbook and magazine
repros of CCDgrams are artificially inserted by the CCD astronomer.
They are arbitrarily assigned to the grayscale values according as
the intent and purpose at hand. Hence, it is ridiculous to interpret
these colors with no initimacy with the original experiment by with
the CCDgrams were captured. They are, for the disinterested audience,
really just false, but often fabulous, colorations.
There are only few and poor catalogs, charts, atlases of the sky
at the deep faint limits of the home CCD. A CCDgram from a 200mm scope
can record 15th magnitude stars. There just are no complete surveys or
census of the heavens at 15th magnitude. One can litterally be lost
among the stars.
There was hope that the Guide Star Catalog would be the roadmap
into CCDgraphy. Alas, as the travails of Hubble brought out, the GSC
is too full of knots for reliable astronomy.
Paralleling this is the lack of AAVSO standard stars and magnitude
sequences. Without these it's tough to assess the magnitude of a
variable star with any competence. True, one could cite the boxing in
by brightness of the target star between other stars. But the
comparison stars taken by opportunity may be unstable in light or have
spectral complications.
As yet there are no commercial applications for photometry with
CCDs. There is some bundled with particular CCD outfits. This is not
transferible to other CCD rigs. Everything on the market is for
collecting fullfield pictures, wonderful and beautiful, yes, but in
themselves hardly valid for photometry.
In the workshop the slides were of hit-or-miss quality. The
speakers took pictures, via regular camera, of the computer screen.
Sometimes the edge of the display unit showed around the edge of the
picture! That's because there is still no cheap, quick, and satisfying
way to get a hardcopy replica -- not a crummy printer output -- of the
displayed image. (The presented results also do highlight the lack of
camera litteracy in the populance!)
It is realisticly infeasible to build a CCD rig from scratch. The
cost and trouble will so far excede that of getting a commercial unit.
And there is little more instruction or knowledge gained this way over
studying the storebought unit.
This is quite different from telescope building. Altho hardy a
means of saving costs, building a telescope yields understanding and
insight to optics and mechanics not easily obtained from merely
working a storebought instrument.
Finally, from this workshop, there is the glatt immaturity in the
instructions and documentations accompanying the CCD outfits. Altho
the apparatus may be well designed, built, and crafted, the paperwork
is useless. It really takes a well-read and wisely astronomer to dope
out what is going on. Many noted round after round of calls to the
company for resolving questions. Others recounted the obligation to
seek for themselves other people expert in CCDgraphy to help them.
The AAVSO workshop was a signal step toward appreciating this new
animal, the CCD system, and everyone is eagerly looking toward the
next workshop at the Columbus meeting.