I have one as well - the rat shack analog model which seems to be the recommended version over the digital. I agree that it makes a huge difference over doing it by ear. I also found the material on calibration disks (AVIA, S&VHTT, VE, etc.) make a bigger difference than just using the speaker tones.
The intent of the exercise is to get all speakers at the same SPL at the sweet spot in your room at a reference level (I use 75db as my reference level, although 85db is recommended if you and the family can take it). The master volume of your amp should then increase/decrease all speakers equally across the range of volume settings.
That being said, I have found that the speaker levels will vary a tad based upon overall volume, and since I don't listen to material at the reference level, there are occasional noticeable changes, but very slight. Also, some material is mastered in a way that accentuates differences from calibration levels. In my room, I notice this in two areas - dialog and bass. In perhaps at most 5% of the cases, I'll tweak the center/sub level by .5db to 1db (you can see we're picking nits here) for the duration, then reset to calibration level when done.
There is also a difference in levels from Dolby Digital and DTS if you've ever run the tests using the different formats. I find the DTS material to be quite a bit "hotter" than the DD equivalent. However the overall speaker balance does not change, just the level.
Perhaps I do this to justify the price of the SPL meter, but if we add wall treatments or move things around in the room a bit, I'll sit down and do a quick level check to see if anything has changed. In some cases, there have been measurable difference in sound levels, especially on the surrounds.
Cheers,
Doug
