Comparison Tycho, Tycho reference and Tycho 2. The quest for the ideal planetarium star database by Bill Anderson:  (billXXX.andersonXXX @ pcmail.XXX , remove XXX)

Dated: 2000-3-8

As with many people in the computer planetarium users group, I awaited the arrival of Tycho 2 data with some anticipation. I had found that Tycho1, released back about 3 years ago, to be an improvement over the time-honored SAO as far as coverage over the general sky. The greater magnitude coverage of  Tycho 1 over the SAO suggested better star density at smaller fields of view, so if I wished to have a finder chart for DIMTHING 2010 for my large 80 mm finder scope, I would have such a chart available. Many of the producers of computer planetaria across the world jumped on the bandwagon and included the data from Tycho 1 as part of their data to fill in the enormous gap between the SAO and the ubiquitous Guide Star Catalog (the GSC).

Incidentally, there are two versions of the Tycho catalog. The first one is a part of the set of data files with the Hipparcos set and is found on the net at the Astronomical Data Center. It is the Hipparcos and Tycho Catalogs, Astronomical Data Center Catalog number I/239, (ESA, 1997). This catalog contains 1,058,332 stars.

The second catalog is the Tycho Reference Catalog, Astronomical Data Center Catalog number I/250, Hog +, 1998. This catalog contains 990,182 stars. There may be some confusion about the difference between the two

catalogs. The first, and obvious difference is that the Reference Catalog has 68,150 fewer stars. While this may not seem to be a lot of stars, I have found that it represents the difference in the display between a clearly visible Perseus Double Cluster and a slight bunching of stars in the general area. When it first came out, the GSC was hailed by many of us to be the ultimate answer for the densest star fields possible. The presence of the GSC data on the CD suggested "state of the art" for planetaria producers. If the program did not have the GSC, it was "old" and "out of date" and "archaic".

However, most did not heed the basic idea behind the GSC in the first place. It was designed to be used by the Hubble Space Telescope and was not intended to be the answer to planetaria producers. The "square black holes" and the doughnut voids around the brightest stars are things we have grown accustomed to as we have adopted the GSC into our tools for the computer.

Still, we (I) looked forward to the arrival of Tycho 2 as the ultimate answer for the empty middle ground for star fields. A source for the GSC is (or was):

Guide Star Catalog version 1.1, Astronomical Society of the Pacific CD's.

This catalog contains well over the 15,000,000 stars indicated.

When Tycho 2 came out a month or so ago, I spent some time downloading it and with much eagerness I set up my little planetarium program to display the new dataset. Some years ago, in the days of the old XT's that used 8086 processors, I adapted a program that was in the public domain for use for myself to use bigger catalogs than came with the program. At first it was the SAO from which I got soul-satisfying displays of the constellations that could be used for constellation identification and crude object finding.

When Tycho 1 came out, I converted the data to permit more density for the star plots. This same program was then used to display the GSC data when it came out. That was a trick because the GSC data, as most are aware, is quite a disk hog. Even with a lot of conversion and fancy programming, the data still takes up several hundred meg of disk space. Then, when Tycho 2 came out, I adapted the data over to use the new dataset and with much eagerness I displayed my favorite test area with the Tycho 1, Tycho 2, and the GSC data. Also, I added a view of the data from the Tycho 1 Reference Catalog as well.

Incidentally, a source for Tycho 2 is:   The Tycho 2 Catalog, Astronomical Data Center Catalog number I/259, Hog +, 2000. This catalog contains about 2,500,000 including the stars listed in the Supplement 1 containing the stars presumably left out of Tycho 2 that are in Tycho 1 (or, more likely, Tycho 1 R).

My test area is the Double Cluster in Perseus. The displays used 2.34 hours and 57 degrees as the field center with a field of view of 3 degrees diameter. I used ALL the data from each of the respective catalogs to make four separate displays. These I then captured with an old DOS screen capture utility and converted the results into .JPG files. Anyone with a

favorite viewer can see the comparison easily between the four catalogs, Tycho1, Tycho 1 R, Tycho 2, and the GSC.

(Incidentally, an absolutely outstanding viewer is one that is freeware and available from which supplies Irfanview, or IVIEW. With this program, or anything else of your choice, the difference between the three catalogs can easily be seen.)

If you wish to see what I saw, there are four JPEG pictures representing screen captures from this old DOS program. The details are:   OBJECT: the Double Cluster in Perseus, NGC 869 and NGC 884 boundaries (in degrees) of RA: 30 to 45 degrees  and DEC: 50 to 60 degrees. All pictures show exactly the same view in width and height, so "blink comparison" is very simple.

Tycho 1 - 5,217 stars, DC1.JPG:

Tycho 1 showing NGC 869 and NGC 884

Tycho 1 R - 4,367 stars, DC1R.JPG: (This is the Reference Catalog)

Tycho reference showing NGC 869 and NGC884.

Tycho 2 - 10,907 stars, DC2.JPG

Tycho 2 showing NGC 869 and NGC 884

GSC - 26,086 stars, DCGSC.JPG

GSC showing NGC 869 and NGC 884

The results were, for me, absolutely astounding, although less than satisfactory. One would expect the GSC to produce a fine display, and it does. If one does not get into a region that has a plate error or a square black hole or whatever, the GSC can produce a nice display. That is to be expected. The difference between Tycho 1 and Tycho 1 R data display is striking. The loss of 68,150 stars seems to really make a difference. The difference between Tycho 1 R and Tycho 2 is significant, but not surprising at all. The fact that Tycho 2 has 2.5 times as many stars makes the difference obvious. The Supplement 1 that comes with Tycho 2 was added back in so the result logically has all the stars Tycho 1 R has. The really striking difference comes with the comparison of Tycho 1 (NOT TYCHO 1 R) and Tycho 2. Even after inclusion, the difference between the two is still striking. I don't think I left anything out of either catalog, but the results show that Tycho 2 does not even permit the viewer to see the Double Cluster to be easily recognized as a cluster at all.

After seeing this, I sat back with some kind of a stunned numbness that said, "What do I do now?" After the wait, the release of Tycho 2 that promised to be the answer turned out to be just another big question. How do I get back to Tycho 1 with the increase of star density contained in Tycho 2? That problem is not as simple as combining the two catalogs. The slightly better precision of Tycho 2 could mean that the slight difference possible in the coordinates may produce a whole bunch of close binaries that do not exist. That would render the whole combined catalog totally useless.

Some kind of test routine would have to be utilized that would try to eliminate the close binaries and make the combination feasible. What kind of test routine would do this and not take up my computer time for the next many months? I still have not answered this problem.

Han Kleijn, the author of the outstanding freeware computer planetarium "HNSKY" has taken up the challange of doing this combination and I watch with eagerness to see how he pulls it off. I wish him the best of luck in this endeavor. What results may be the answer. We will see. For now, I am thrown back into a wait and see mode to see how others handle this problem. Perhaps someone else can see a technique to solve this problem (if my interpretation of the data is correct). Anyway, I feel a little cheated right now that I still do not have the perfect database.

{00-3-24 note han Kleijn: Bill did the job himself, (faster then me). Since all stars have a (GSC) number ,the job could be done in a day or so using a sorting program and some basic programs. Based on the result of Bill, I made a supplement file of it, see section 8) in the software page. }

2001-7 Now you can download the combined Tycho-2 and 1 called the TYCHO-2+

I guess this last observation shows that there is still a big gap between the amateur and professional astronomer.

Bill Anderson:  (billXXX.andersonXXX @ pcmail.XXX , remove XXX)

Phoenix, Arizona