ALT_AZ.CPP: Code to convert RA/dec to alt/az, and galactic coordinates to RA/dec.

ASTEPHEM.CPP: The main function for the code to compute asteroid ephemerides. It demonstrates accessing orbital elements from the Guide disk, and then shows how to compute an RA/dec, magnitude, elongation, and other data based on using those elements to compute the asteroid position, and VSOP to compute the Earth's position. The most significant thing omitted right now is the topocentric offset.

ASTFUNCS.CPP: Some basic functions to set up asteroid/comet elements and to compute their locations (solving Kepler's equation and so forth.)

BIG_VSOP.CPP: Code to compute planetary positions from the full VSOP series, which is stored in binary form on the Guide CD-ROM. The logic is very similar to that used for the "truncated" VSOP (see VSOPSON.CPP), except that the function works by reading in pieces of a file; VSOPSON.CPP assumes the data file has been read into memory.

COLORS.CPP and COLORS2.CPP: Given, say, a B-V color value, it's possible to compute a "pretty good" V-R value. Similar conversions are possible between most other color indices, with varying degrees of accuracy. The general form is a polynomial, say,

index2 = a0 + a1 * index + a2 * index2 + a3 * index3 +.... 

These resemble Taylor series, though I suspect they are actually curve-fit polynomials. I got them from some Fortran snippets supplied by Brian Skiff, and they have been converted to C here.

Certain conversions are handled by inverting the polynomial. The need to do this indicates a very unreliable color conversion, to be used only in cases of dire emergency.

Also, code to convert Tycho B and V to Johnson B and V is given, along with a little piece of test code. (That Tycho-to-Johnson algorithm is one based the book that was distributed with the Hipparcos data, "Introduction and Guide to the Data." COLORS2.CPP is a more recent, and probably better, Tycho-to-Johnson piece of source code.)

COM_FILE.CPP: This is pretty much Guide-specific code. Guide goes through some odd hoops to draw comets as speedily as possible. The tricks involved were absolutely necessary back when computers were slower than they are today, but are probably not all that helpful on modern PCs.

COSPAR.CPP: Code to compute the "planet orientation" matrix for the Sun, all planets, and quite a few satellites. A few years ago, the IAU set up a committee to define pole positions and longitude systems for these objects. The pole position is usually defined to be a linear function in RA/dec, with the meridian intersecting the J2000 plane as another linear function. Often, some small trig terms are added to those linear functions.

At present, COSPAR.CPP includes many, but not all, of the satellites shown by Guide. Some (such as the "captured asteroids" of gas giants) have no defined orientations. Others (small inner objects such as Amalthea) just haven't been tackled by me yet (no big job to do, but not much interest, either).

DELTA_T.CPP: Code to compute Delta_T = TD - UT for any date, using a lookup table for years 1620 to 2002 and extrapolations beyond those ranges.

DIST_PA.CPP: Code to compute the distance and position angle between two spherical coordinates. Theoretically, this ought to be a straightforward task. In practice, there are an amazing number of ways to get roundoff/truncation errors and divides by zero, mostly involving how you deal with cases where the distance between the two points is small. I am reasonably sure now that this code evades all such cases.

EART2000.CPP: Code to compute the Earth's location relative to the Sun, in J2000 coordinates, using VSOP.

EASTER.CPP: Code to figure out the date of Easter for a given (Gregorian) year.

GETPLANE.CPP: Code to compute the location of any planet, or the Moon, relative to the Sun. The resulting vector is given in five different systems.

JD.CPP: Produces an "example" program showing the use of the functions in DATE.CPP. You can type in, say, "jd 10 7 1999", and be told what day 10 July 1999 corresponds to in assorted calendars. (You can also get the actual JD corresponding to that date, or type in, say, "JD 2451324.879" and get the calendar date... which was the original reason I wrote the program.)

MISCELL.CPP: Code for basic matrix functions (inverting an orthogonal matrix, rotation, setting identity matrices, etc.) Also code to handle the odd system used for variable star designations, and to provide a version of the C-language ctime() function that takes a JD and can work over a full time range (ctime only works between 1970 and 2028.)

NUTATION.CPP: Computes the nutation of the earth's equator.

OBLIQUIT.CPP: Computes the obliquity of the earth's equator for dates from -8000 to +12000.

PRECESS.CPP: Code to build the matrix used to convert coordinates between epochs, plus code to use those matrices to convert vectors, RA/dec coordinates, and ecliptic coordinates between epochs.

REFRACT.CPP: Contains functions to convert an observed altitude (one affected by refraction) to a true altitude (one that would be seen on an airless planet), or vice versa. Two different methods are provided. (See next paragraph for a third method.)

ROCKS.CPP: Code to compute positions for 25 faint, inner satellites of the gas giants. These are all modelled as precessing ellipses.

SSATS.CPP: Code to compute the positions of eight of Saturn's satellites (Mimas, Enceladus, Tethys, Dione, Rhea, Titan, Hyperion, Japetus), using the theory due to G. Dourneau.

VISLIMIT.CPP: This code computes the limiting visual magnitude, sky brightness, and extinction coefficients. It's basically lifted/ported from Brad Schaefer's article and code on pages 57-60, May 1998 Sky & Telescope, "To the Visual Limits".

Why the source for Charon is being posted: I do not now, nor have I ever, regarded astrometry software as a commercially profitable idea. By a slight margin, I'm able to keep myself fed and indoors through sales of Guide; trying to sell Charon is not apt to help very much. By posting it, I hope to accomplish the following:

-- People will be able to improve/modify the source code to accomplish tasks of their own. (The basic idea behind almost all open source, of course.) With any luck, some of that code will return to me and I'll post it here for others to use. I know that some people have written Charon-like code and functions for their own project; I'm hoping that we can stop re-inventing wheels.

-- The number of users of Charon has risen to the point where I'm not doing too well at keeping up with requests for improvement. I've become a bottleneck to progress.

-- Posting the source code will probably encourage somebody to start work on making a version of Charon that does not require a Guide CD. If it can access catalog data from the Internet, then it would require no CD-ROM at all. In this case, people could do astrometry with Totally Free Software. This might encourage some people to enter the hobby.

-- Verifying the fact that the code does what I tell people it does becomes easier. Some people build their own astrometry software because they would rather not have to place trust in me and my software. Such trust would be well-placed, especially since the results of the software have been examined by dozens if not hundreds of people. (The worst risk in 'rolling your own' astrometry software is that it will usually be tested only by you, and not by dozens if not hundreds of people.) But being able to check the source code ought to relieve some anxieties.

Current state of the Charon source code: All of the rest of the source posted on this site has been cleaned up nicely and documented well enough to make it useful to somebody. Right now, this is not in the least true of Charon. You can use the source to satisfy curiosity about parts of Charon, and maybe to grab bits and pieces from the code. But there are some things I need to do to make it usable.

Charon is compiled with WATCOM C/C++, the only compiler I've got that can produce 32-bit DOS code that can access graphics. (Micro$oft's compiler can produce 32-bit DOS console applications, but none that access graphics.) I'm in the process of getting the make file into a usable form; right now, it runs across assorted directories and is not 'ready to go out of the box'.

There are parts of the code that run for hundreds of lines without the faintest trace of the residue of a comment. The code isn't particularly ugly (except for the enormous main() and parts of matcher.cpp), but figuring out what does what is a frightening prospect right now.

Overall structure of Charon: The program is overdue to be broken up into certain components. To a minor extent, this breakup is present now, but it's incomplete and Charon is really one big monolith at heart. This is really a shame, because astrometry software lends itself naturally to being broken into a set of component tasks.

Breaking Charon into components would make following the structure of the code much easier. It would let us rip out certain parts, replace them, or use them by themselves. The components (or, if you prefer, "near-components") are:

Load an image from a file: A piece that loads an image, be it FITS, SBIG, or other, into a standardized in-memory format. This is currently done using the load_image() function in load_img.cpp. A side note: it's possible that, in preference to writing one function that can load a zillion image formats, there should be a separate routine that can convert a zillion image formats into FITS. If I had it to do over again, I'd do it that way.

Find and build a list of stars in an image: Takes an image in the aforementioned standardized in-memory format and return a list of stars in it, with their pixel coordinates and brightnesses. The code in findstar.cpp does this right now. (I'm not totally thrilled with the job it does on noisy images; it often finds hordes of spurious stars in noise. Ideally, the code I wrote might get pitched in favor of SExtractor. I started playing with SExtractor and gave up, though; I had no real success in puzzling out what was supposed to do what.)

Get the RA/dec of an object at a given time: Given an object ID such as "V4191 Sgr" or "P/Linear S4" or "1997 XF11", and a date/time and observer position, this figures out the RA/dec of that object. The code in find_obj.cpp does this right now. If somebody tells Charon, for example, "Here's an image of asteroid 4179", Charon can figure out a rough RA/dec for 4179 at the time the image was taken, so it knows where to look for a pattern match.

In some ways, this is just a convenience. People can and do feed Charon images and just say, "Here's an image taken at about 14h51m12s, +10 43' 56"... go process it." But the 'convenience' is a major one.

Get stars from a catalog within a given radius of a given RA/dec: Given a catalog name and an RA/dec (normally provided by the preceding piece, but not always) and a radius, this piece grabs catalog data for that area and returns it in a standard way. Right now, assorted functions in grab_gsc.cpp do that. They can extract ACT, GSC, or GSC-ACT data from the Guide disks, or USNO Ax.0 or SAx.0 data from the CDs distributed by USNO, or GSC 1.2 data from the files that come from the STScI server. All the functions feed through the grab_catalog_stars() function in charon.cpp.

For Guide users, this is pretty nice; it means they can get all the data they need (comet/asteroid orbits, plus the GSC, plus other handy catalogs) all on one disk. For an open-source project, though, it'll be nice if the 'catalog grabbing' code can 'grab' from the two GSC CD-ROMs distributed by STScI, or via Internet. The first, at least, ought to be very straightforward. The second will be slightly trickier, but opens the program to anyone connected to the Internet... this would be a huge plus. I note that VizieR, for example, provides a way to ask for a given list of data from a catalog, specified by RA/dec and radius, using a single URL... almost exactly what we'd need for a stunt like this.

Right now, people have to shell out some money to buy the Guide disk, or the two GSC disks, to get into astrometry. Neither is really pricey, as astronomy products go ($89 and $69, respectively). But if trying out astrometry software was totally free, it might draw in some fresh blood.

Assorted minor functions: A whole slew of small pieces to do things such as write out MPC headers and reports, figure out which image star or catalog star is nearest to a given cursor position, provide ways to adjust settings, handle the list of MPC stations...