In a video on the effects of music on the brain Oliver Sack's brain is shown when listening to Bach as opposed to listening to Beethoven:
Jokes are made from the off: "Sorry, Beethoven, there's not much there."
But I assume that the images are misleading (as many fMRI images are). Of course there's a lot going on in Sack's brain when listening to Beethoven, but fMRI just works like this. It can only highlight regions with some higher activity (compared to other stimuli or regions), not absolute activity levels all over the brain. In other words: fMRI can only yield images with very sharp contrast, showing only tips of icebergs, so-to-say. Finer details are levelled.
My question is: Are there simulations that look like fMRI records but give a sense for the true activity levels in a working brain (on a sensible scale)? Probably such images would need some sharpening of contrast because otherwise one would see nothing because the differences are so small.
Though this is mainly a reference request for specific examples, an answer with theoretical considerations (about feasibility, accuracy, and benefits of such simulations) would be appreciated, too.