Brian wrote: > In re Berto's comments: if one were to digitally scan sky patrol plates >and calibrate them photometrically, it is certainly the case that one could >dig out a lot more variables and lightcurves for them thean simply blinking >pairs by eye. There's a lot of information stored on photographic plates and >films, but digging it out to the highest accuracy is a challenge. We've thought about scanning the Harvard plate collection here, since we are rapidly running out of surveys that we *haven't* scanned, but there are politics involved. More to the point: there is a meeting in Sonneberg in March titled "TREASURE-HUNTING IN ASTRONOMICAL PLATE ARCHIVES" (see url http://vsnet.stw.tu-ilmenau.de/~web/workshop/thapa.html for more details), in which many of these data mining issues will be discussed. I personally have never even looked at a Harvard or Sonneberg plate; my experience with large plate surveys other than POSS type ones is the minor planet plate collection from Indiana University. I thought the quality was pretty low; lots of trailed plates, poor development, etc. Certainly good enough for the project they were used for, but if the other archives are of similar quality, I'd question how faint and how reliable searches would be, even if the plates were digitized. Someone with more experience might want to comment. Arne