sounding good versus scoring good
Inspired after reading “This is Your Brain on Music” and getting through the Endless Setlist 2 on Rock Band 2, I found myself writing a long-ass post on the Rock Band community forums in response to a thread regarding the accusation that people don’t actually sound good when singing in the game.
I didn’t say so in my response, but I absolutely agree. I dislike the way I myself sound, and I try to protect my housemates from it as best I can. It begs the question of why I play it at all, which turns out to be perfectly apropos to this topic.
Simply put, it’s a challenge! Can I get a great score and sound cool? Some songs work great for me, and some songs are horribly unforgiving, score-wise, when I attempt to inject some style and tone. I’ve started recording myself when singing a wide variety of songs both in- and out-of-game, and in many cases I’m amazed at the dull flat drowning-cat noises that have won me perfect scores in certain songs.
But WHY is it a challenge to sound good AND score good? From this point on I have pasted what I posted in the Rock Band forum.
This topic is essentially all about what makes a “good singer”. As intuitive creatures we have a sense of what’s attractive and beautiful (“good singing”) versus what is robotic and off-putting (“monotone singing”).
So, what makes someone a “good singer” as opposed to someone who simply sings the pitches required to get good scores in a karaoke game but somehow manages to sound like a dying cat?
Here are my thoughts on this, in a somewhat-but-not-completely organized fashion.
I actually don’t think this is a hard problem. Obviously songs consist of logically arranged melodies (and that’s a huge topic, so I won’t get into it). But they also consist of “challenges” to what we expect to happen — twists in the plot, so to speak.
In terms of singing, those challenges come in the form of unexpected tiny delays, choosing thirds instead of fifths for the second time through a melody, using slides to get from one note to the next, and injecting spots of emotion into words that may affect the pitch slightly (even throw it “off”). Some of these things are captured in the Rock Band charting, but if they don’t “come from” the person singing them, they sound just as robotic as any other parts. (Sorry to be so imprecise, but we’re talking Aesthetics here.)
Good singing is pretty much characterized by flaws, because humans are flawed beings, and hearing someone perform “perfectly on-pitch” throughout an entire song feels alien to us. We call that singer robotic and monotone, whereas we call someone like Eddie Vedder an amazing vocalist even though he never sings his songs the same way twice (and arguably he cannot). And why should he? He should challenge our expectations and inject slightly different emotions to stay interesting (which he does) — not recite the melody like a playback machine.
My other main observation is that when you’re singing for the joy of singing and/or to produce something aesthetically pleasing, you are necessarily producing sounds that sound completely different from the vocalizations that occur when you are staring a screen with your thoughts bound up in the effort of getting that little arrow to match with that line. Just an attitudinal thing.
In other words, my opinion is that singing for the purposes of FCing songs in this game produces vocalization that is the antithesis of what our brains want to process as aesthetically pleasing.
Tangentially, I would love to see more original singers of songs in Rock Band playing their songs in the game. 
Just to wrap up, clearly it’s possible to sound really good AND appease the game algorithms enough to get 100%. I’m just providing explanations as to why (it seems) these conditions only rarely coincide.
–Naomi
* platinum RB2 vocals
* trained choral singer
* real-life rock band vocalist
