Suzanne Vega ought to be the hardest thing in the world to find an MP3 artifact in, since "Tom's Diner" was the primary litmus test used in the development of the psychoacoustical model.
I'd argue the opposite, "voice is the hardest to get right so we chose a vocal track" -> voice is hard to get right, so they got it least right..
For me, Tom's Diner was the easiest of the lot to pick out.
https://en.wikipedia.org/wiki/Tom%27s_Diner#The_%22Mother_of...