>>3341115I had the same problem from a different lab.... If anything, mine tended towards slightly overexposed.
It's either some sort of auto-correction, based upon how the first or last slide in the roll was exposed. Someone upped the brightness on this to the point that the whole thing looks a lot better in a slide viewer.
I can't say I've never gotten a good scan of a slide, mind - I've only ever gotten one set scanned and processed. But I'd imagine these are a bit harder to scan than a negative.