Page 1 of 2

Constellation lines

Posted: 01.06.2002, 01:51
by Matt McIrvin
In the hope that someone will find it useful, I'm expanding Celestia's constellation maps to include the full set of 88.

The exercise is illuminating. People keep making jokes about how the lack of resemblance of constellations to their names implies that "the ancients" were hopped up on goofballs, but I think that Lacaille, who named much of the Southern Hemisphere, was the biggest goofball freak of all, or just had a mania for populating the sky with the equivalent of Easter eggs.

Of course, the stars in a constellation can be and are connected in many ways. My choices are based on my own tastes, but it's easier to say what my own tastes are for the constellations I've actually seen. I guess if I post the thing publicly and a bunch of Australians tell me I'm crazy, I'll know what's up.

Done!

Posted: 02.06.2002, 00:26
by Matt McIrvin
It's finished and available on Bruckner's addons site:

http://bruckner.homelinux.net/addons.html

Follow the instructions for asterisms.zip.

Done!

Posted: 02.06.2002, 00:42
by t00fri
Matt McIrvin wrote:It's finished and available on Bruckner's addons site:

http://bruckner.homelinux.net/addons.html

Follow the instructions for asterisms.zip.


Since some weeks I have checked into the CVS tree a carefully done complete set of 88 constellation figures that will be part of the next version.

Of course my new asterism file may be downloaded from there.

Bye Fridger

Oops

Posted: 02.06.2002, 00:52
by Matt McIrvin
Well, I had fun making mine anyway.

Screenshots:

Ursa Major and environs:

http://world.std.com/~mmcirvin/ursamajor.gif

Modern constellations in the south polar region:

http://world.std.com/~mmcirvin/southpolar.gif

Oops

Posted: 02.06.2002, 11:21
by t00fri
Matt McIrvin wrote:Well, I had fun making mine anyway.


Sure, any further attemps should be very valuable since there are /no
IAU-standards/ unlike the constellation boundaries.

The "global" expert on these figures seems to be Chris Marriott ( SkyMap Pro
8), http://skymap.com/. The constellation figures in XEphem are also by him. On
http://skymap.com/constellations.htm, there are various configuration files for
free download that were submitted by SkyMap users and contain the stick-figures
corresponding to well known atlases, like:

1) Based on the popular Nature Company Guide books
; "Skywatching by D. Levy",
; "Advanced Skywatching by R.Burnham, A.Dyer, R.Garfinkle, M.George, J.Kanipe,
D.Levy"

These figures are /complete/(88/89) and associated with a quite experienced
astro-popularizer: D. Levy. They have some appealing simplicity in
favour. Yet a number of constellations seemed /too/ simplistic;-). My
new design for Celestia is substantially based on that set and, of
course, on a large amount of hand-editing. Also, my 12 year
association with XEphem cannot be denied...

2) On the other end of the spectrum is an amazingly fancy set, but
it may well be too fancy, I think (involving too weak stars, etc):

; Constellation figures based on H. A. Rey's "The Stars", published
; by Houghton Mifflin Co., Boston
;
; Adapted by Tracy Williams (tcw@iname.com), January 1999, with
; assistance from Robert Wade's Sky Map Pro 4 figures, where applicable.

3) "The Observer's Sky Atlas" by E. Karkoschka.

2) From the constellation maps from Anton?n R?kl from "Souhvezd?", published
by Artia, Prague in 1971 and from second book with unknown original
title (spanish translated title is "Estrellas y Planetas") published 1988
by Artia, Prague in 1988


As a basis for my designs, I have coded a Perl script (asterism.pl)
that 'elegantly' converts any desired set from
http://skymap.com/constellations.htm into Celestia format.

Anyone who is interested to play with different|further sets including
the above, may get it from me by email. Or, if preferred, I also have
of course the already converted asterism*.dat files: asterisms-hareyv5.dat,
asterisms-skywatch.dat, asterisms-rukl.dat, asterisms-karkoschka.dat, ...
( note that asterisms-hareyv5.dat sometimes refers to 'unnamed' stars;-), but
this is not too bad)

Of course, my final design in the CVS tree may also be obtained by email.
If a larger number of people turn out to be interested, we can also
put the new asterism.dat file to Bruckner's.

The basic strategy I applied was to try and find a good compromise
between a /conspicuously mnemnonic/ drawing (e.g. Gemini, Orion, Taurus,
Canis minor, Scorpius, Perseus, Hercules, Libra, ...), /popularity/ of
the figures (Ursa major/minor, Andromeda, Cassiopeia, Lyra, Virgo,...)
and /simplicity/, i.e. the restriction to sufficiently bright stars.

Any feedback is greatly appreciated...


Bye Fridger

Oops

Posted: 02.06.2002, 14:30
by Matt McIrvin
t00fri wrote:The basic strategy I applied was to try and find a good compromise
between a /conspicuously mnemnonic/ drawing (e.g. Gemini, Orion, Taurus,
Canis minor, Scorpius, Perseus, Hercules, Libra, ...), /popularity/ of
the figures (Ursa major/minor, Andromeda, Cassiopeia, Lyra, Virgo,...)
and /simplicity/, i.e. the restriction to sufficiently bright stars.


I had some similar thoughts while I was compiling my own figures and wrote down what I found to be the basic strategies followed when drawing constellations. It seemed to me that different authors had different goals in mind:


1. "Pointers" approach (original Celestia set, newspaper astronomy columns): Show
simple asterisms linking bright or noteworthy stars, and leave off
everything else.

Pros: Best for complete beginners to get themselves oriented, especially
in poor seeing conditions.

Cons: Not very useful for more advanced observers. Many dim
constellations will disappear entirely. Constellations usually don't
look like much of anything, with surreal exceptions like the Teapot.

2. "Transit Map" approach (seasonal sky maps in Olcott and Mayall,
"Field Guide to the Stars"): Augment the above with spidery, branching
structures that connect most of the naked-eye-visible stars to the
skeleton of the constellation.

Pros: Probably the most useful for advanced amateurs seeking to locate
objects for binocular or telescope viewing; helps locate individual
objects in a constellation by encouraging "star-hopping" along lines
that appear readily to the eye.

Cons: Spidery figures are ugly, similar-looking, and mnemonically
useless for finding whole constellations; confusing to novices.

3. "Cartoon Figures" approach (H. A. Rey, "The Stars: A New Way to See
Them"): Strain to find cute figurative representations of the
constellation names by connecting dimmer stars.

Pros: Star maps are charming and attractive to children and novices;
avoids "what were they thinking?" reaction; constellations may be
mnemonically useful to people blessed with clear, dark skies. (Rey's book was a great inspiration to me as a child, even though I couldn't see most of the figure he drew.)

Cons: People in the city won't be able to see those cartoons in the
actual sky; constellation lines may not be convenient for star-hopping;
pretty much dead in the water for all those southern-hemisphere
constellations named after scientific instruments.

4. "Here Be Dragons" approach (cave paintings, very old texts,
astrologers): Draw fanciful pictures on top of the stars instead of
connecting them with lines.

Pros: Makes pretty, archaic-looking illustrations for discussions of
star lore.

Cons: Bad for any astronomical use, since the actual stars get lost in
visual chaos. See also #3.



Some combination of 1-3 will be optimal for any given audience. You
probably want to lean toward Transit Maps if drawing for an audience
already well familiar with the night sky, or talking about deep-sky
objects; some combination of Pointers and Cartoon Figures will be better
for Jack Horkheimer-like pop use, with caveats about what the audience
will really be able to see in a given situation. A Here Be Dragons
drawing might be useful as an adjunct, but only if you're talking about
history or mythology, and I tend to think that anything Dragons can do,
Cartoon Figures can probably do better.

A possibly good approach that I don't think I've seen used on a large
scale is to use lines of differing brightness; you could draw the
Pointer asterisms in heavy lines and type 2 or 3 additions in lighter
ones.

Other ideas

Posted: 02.06.2002, 14:50
by Guest
I also thought of some other possible projects along these lines, such as an alternate asterisms.dat file consisting entirely of obsolete and unofficial constellations. Unfortunately, very detailed information on these is sometimes hard to find, and the problem of no standard for constellation lines is even more pronounced for them. But it would be nice just to be able to see their names written on the sky in the right places.

There could also be files showing constellations used by other cultures. The interesting thing I found, doing some cursory looking into this, is that not all of them primarily use the Western concept of a constellation as a sort of star figure. Some have lore associated mostly with individual stars. Others (such as Chinese traditional astronomy, which is pretty complex) do have asterisms but more often speak of the sky divided into "mansions" somewhat resembling, though not quite like, Western astrological "houses." And the whole business is complicated still further by centuries of cultural cross-pollination; the Chinese ideas ended up affecting Indian and Arab astronomy. There's a lot of potential for interesting things to do.

Oops

Posted: 02.06.2002, 15:14
by Guest
I agree with most of your Pros and Cons. Exept, perhaps, that simple (Pointer-type) shapes for constellation figures are not very useful for advanced observers.

Well, I am an active amateur astronomer since I am a child and meanwhile, I know the sky /really/ well. Yet, as a quick orientation, I mostly adhere to quite simple minded search patterns that probably are the ones from my childhood. As soon I have arrived with my eyes at a given constellation, I know most visible /individual/ stars anyhow and the constellation figures become entirely unimportant...

As a matter of fact, I had a look at your asterisms.dat. I think the amount of "fancyness" is about
comparable to my figures, yet there are plenty of different figures in both. E.g. Ursa major. I chose the rather simple shape /very/ popular in Europe;-) while you have the one that most Americans like...Your's admittedly looks much more like a bear. Mine is even conspicuous in downtown areas;-).
On the other hand my Gemini shape shows /clearly/ twins, holding their hand;-) etc. We also differ markedly with Perseus and Andromeda. My Andomeda figure leads e.g. directly "up" to M31. Note also that Andromeda is not connected to Pegasus as your lines might suggest (the boundaries are standardized).

So at the end, it's all going to be a matter of taste. But fortunately, asterisms.dat can be very easily edited accordingly.

Bye Fridger

Boundary cases

Posted: 02.06.2002, 16:49
by Matt McIrvin
Anonymous wrote:We also differ markedly with Perseus and Andromeda. My Andomeda figure leads e.g. directly "up" to M31. Note also that Andromeda is not connected to Pegasus as your lines might suggest (the boundaries are standardized).


Right, in that case I violated the IAU boundary in a bow to the most popular American conventions. According to the IAU Alpheratz is unambiguously in Andromeda, but I didn't want to eliminate the "Great Square of Pegasus" asterism.

I thought that Beta Tauri was another case of this, but it looks to be precisely on the boundary with Auriga, though it is conventionally classed with Taurus. And then there's the confusing Ophiuchus/Serpens situation...

Other possible changes

Posted: 02.06.2002, 16:56
by Matt McIrvin
Also, I should say that not all of the constellations in my set were my choices; I kept some constellations from the original Celestia set which I probably would have drawn somewhat differently, particularly Virgo and Hercules. I made small changes to some others and added lots of lines to Orion and Ursa Major.

Boundary cases

Posted: 02.06.2002, 17:11
by t00fri
Matt McIrvin wrote:Right, in that case I violated the IAU boundary in a bow to the most popular American conventions. According to the IAU Alpheratz is unambiguously in Andromeda, but I didn't want to eliminate the "Great Square of Pegasus" asterism.

I thought that Beta Tauri was another case of this, but it looks to be precisely on the boundary with Auriga, though it is conventionally classed with Taurus. And then there's the confusing Ophiuchus/Serpens situation...


In case of Andromeda-Pegasus one often finds the use of dashed/dotted lines in order to avoid violations of IAU standards on the one hand and a very suggestive pattern on the other hand. Personally, I always virtually attach Andromeda to Pegasus;-)

Anyway, it is apparent that you enjoyed this work as much as I did...and that's most important after all.

Bye Fridger

Posted: 03.06.2002, 12:55
by marc
Ive found that if you define a constellation twice in the asterims file you end up with much brighter constellation lines. I know that its not very elegant but how about doing this to distinguish the major constellations from the minor ones?.

Constellations choices...

Posted: 03.06.2002, 14:46
by Kendrix
I will soon add to the code the possibility to render only the constellations you want... Probably with a list of checkbox...

Constellations choices...

Posted: 03.06.2002, 17:00
by Guest
Kendrix wrote:I will soon add to the code the possibility to render only the constellations you want... Probably with a list of checkbox...


Personally, I must confess that I do not see too much motivation for such a customization at present, but I may have overlooked something...

A more useful modification appears to me the option to include e.g. dashed or dotted lines that can be parsed from asterisms.dat.

Bye Fridger

Posted: 03.06.2002, 17:59
by chris
marc wrote:Ive found that if you define a constellation twice in the asterims file you end up with much brighter constellation lines. I know that its not very elegant but how about doing this to distinguish the major constellations from the minor ones?.

It would be very easy to add some sort of line style setting to asterisms.dat . . . is this a feature that a lot of people would find useful?

--Chris

Posted: 03.06.2002, 18:53
by Pixel
I find more usefull the constellation options available in Starry Night. I. e menu to select things aka:
Asterism mode
Astronomical mode
Rey's mode
Zodiac only mode
----------------------
Boundaries
line style...

Line styles

Posted: 03.06.2002, 22:13
by Matt McIrvin
chris wrote:It would be very easy to add some sort of line style setting to asterisms.dat . . . is this a feature that a lot of people would find useful?


I think it would be useful to give asterisms.dat the ability to specify, at the very least, dashed or solid lines, and maybe two different levels of brightness in addition to that. The dashed lines would be good for odd cases like the Great Square of Pegasus that do not obey official constellation bounds, and bright/dim lines could be used to distinguish bright "pointer" asterisms from the rest of a constellation, especially for cases like the Big Dipper and the Sagittarius Teapot.

I'm not sure whether it would be better for asterisms.dat to specify visual lines styles directly, or to specify some sort of "semantic" line categories (in the spirit of XML) that would just happen to be rendered as different visual styles. This is the sort of question that provokes holy wars, but in this case it might not make that much difference...

done !

Posted: 03.06.2002, 22:58
by Kendrix
Ok I have added a new line in the navigation menu...

Now you can choose to render all constellations (normal mode) or only one choosen in a combo list...

it is usefull to see how a particular constellation looks from a specific point in space ! (for ex : how Ursa Major looks like from betelgeuse)

I have also added some things like the actual max visible magnitude (just above the FOV) and a new camera mode : heliocentric...

So you can stand next to a planet and see it slowly move on its orbit. (for the earth it is 30 km/s so you really can see it !)

If some of you are interested just send me an email...

done !

Posted: 03.06.2002, 23:20
by t00fri
Kendrix wrote:Ok I have added a new line in the navigation menu...

Now you can choose to render all constellations (normal mode) or only one choosen in a combo list...

it is usefull to see how a particular constellation looks from a specific point in space ! (for ex : how Ursa Major looks like from betelgeuse)

I have also added some things like the actual max visible magnitude (just above the FOV) and a new camera mode : heliocentric...

So you can stand next to a planet and see it slowly move on its orbit. (for the earth it is 30 km/s so you really can see it !)

If some of you are interested just send me an email...


First of all the GUI is currently being redesigned, so modifying the GUI ad hoc at this time is perhaps not so worthwhile.

Moreover, I really do not see the use of switching on|off individual constellation figures. Even with all switched on, you may inspect how Ursa Major looks like from Betelgeuse. What's the difference??
You may zoom in such that you only see Ursa major on the screen...

Moreover, the max magnitude may be adjusted by means of the key ] and for Linux also in the GUI.

The heliocentric mode should also be obtainable in terms of the present facilities if I understood you right.

Bye Fridger

well...

Posted: 04.06.2002, 06:04
by Kendrix
For the magnitude it's only displaying the value now ! So you know where you are... ( because after many press on the [ or ] keys how to know what is the actual max visible mag)...

For the "only one constellation" I don't agree you... When you look a constellation from another star, it can be very stretched so you can't zoom on it to see only that particular constellation... And if you are really far from the solar system, you will see a bunch of pink lines but it's difficult to see only one of them...