This item originally posted to the TASS mailing list.
Both Hipparcos and Tycho systems used equivalent detectors, and so have roughly equivalent limiting magnitudes. The main difference is that Hipparcos used a broad-band V-like filter, whereas Tycho used two narrower filters (roughly B & V). This means Hipparcos is limited at about V=12.5 and Tycho is limited at V=12 (one sigma).
However, 'magnitude limited' and 'completeness' are two entirely different concepts with these instruments. Hipparcos is complete to V=7.5. Fainter than this, all stars were part of an 'input catalog' and were placed there according to their astrophysical value. In no way can you call the Hipparcos catalog a statistically complete catalog. For example, Hipparcos *only* looked at stars contained in its input catalog. It contains about 100,000 stars, with extremely high accuracy astrometry (1mas), proper motions, parallaxes, and mean magnitude.
Tycho is complete to V=10.5. The input catalog for Tycho contained the 3 million brightest stars, but Tycho actually measures *everything* it sees, and the input catalog is just there to find faint, marginally detected objects based on some a priori knowledge. The 1 million stars in the first catalog release are the brightest stars plus a few fainter ones that have astrophysical interest. A reanalysis of the data will permit another million or so stars to be released in a couple of years. Tycho, because it used the star tracker and was not designed as an astrometric instrument, gives good but not great astrometry (25-50mas), proper motions and parallaxes, and good mean magnitudes (0.05mag errors at V=10.5, but up to millimag accuracy on brighter stars).
The ASCII versions of the catalogs are quite large (for example, Tycho is 370MB), but these catalogs are easily distilled into binary records (a Tycho catalog with most fields retained becomes 40MB). I don't know the copyright restrictions as to whether the catalogs could be put on an Internet server.
How do these catalogs fit into the TASS concept? Hipparcos has little value because most of its stars are brighter than measurable with the mark III systems. Tycho is a much better catalog, since the majority of its stars are nicely exposed on mark III images. Its astrometry is better than GSC, and more than sufficient for our uses. I'd consider Tycho as the primary astrometric catalog for TASS. Its photometry can be transformed nicely into the standard Johnson system, and so you have good V measures, against which you could set the V-band zeropoint and also use as input into the automatic star selection parts of the astrometry/photometry modules of Star, Sextractor, etc. There is no R or I band information, so you still have to use Landolt standards for those passbands. The equatorial stripe of Tycho should be pretty small and managible. As far as variable stars are concerned, Tycho individual measurements are worse than any TASS image. I think you can use the variability flag to check for bright variables, and then TASS data to actually search for periods. You could also use the variability flag to eliminate objects from the master photometry list that might possibly be variable.
More information about Hipparcos and Tycho are in the S&T article, plus several issues of Astronomy and Astrophysics were dedicated to the mission if you want more scientific information.