To make a CMD, we had to measure how much light came from each star in both the B and V filtered images. Measuring this light is called doing photometry. We used a package called DAOphot for this.
We showed DAOphot examples of bright, well-isolated stars in our fields (this is called picking PSF stars ... because the shape of the "hill" for each star is called the point-spread function) and then we let DAOphot go to work. It did the best it could (well, maybe it could do better, but we're not DAOphot experts) at measuring all the light "under" each "hill" of a star. One way to see how good a job it did is to ask DAOphot to measure all the light and subtract off the stars... here's a before-and-after example of an M5 image and then the image with all the stars fit and subtracted away: You can see we did OK.
First before subtraction:
Then after subtraction:
After doing this for a B frame and a V frame, we have magnitudes (the V values) and colors (the B-V values ... do you remember why we subtract the magnitudes to get colors?) but there were a couple of details to take care of ...
The stars in one frame had brighter magnitudes than in the other not necessarily because they were really brighter, but because one image was exposed for longer. We had to correct for that, so we added 2.5*LOG(exposure time) to all the magnitudes. This is called converting to instrumental magnitudes. Then we calibrated our instrumental magnitudes to "real" magnitudes by comparing with the standard stars we observed. These stars are like light bulbs with a wattage we know, so we know what magnitude they really are. We offset our magnitudes to line up with those stars, and then we were in business ... time to make a CMD!