Audiosurf II is in desperate need of a demo, as I suspect this one will be a little polarizing. The game takes the sublime music-riding puzzle experience that has captured so much of my attention over the years and adds an entirely new game mechanic--wakeboarding. It's fantastic, converting every audio track into a points-driven series of gambles. But it's also incredibly confusing, sapping away that particular exploratory charm that made the first game so wondrously unique.

Let's back up. If you've somehow missed out on the first Audiosurf, you should actually just go ahead and get it; it's on Steam for ten bucks and has a free demo. In brief: the game converts your MP3 files into rollercoaster-esque thrill rides, rivers of light and sound undulating in time with a music track. Colorful bricks sit in the "stream," and you'll pilot something like a rocket ship to line those bricks up inside a grid and rack up points. Actually, you'll mostly be firing up new tracks to try, giddy with wonder as old favorites you thought you knew so well take on a brilliant new form.

A sequel makes things complicated--all of Audiosurf's "levels" are procedurally generated based on your music, so the only real way to go is to add new features. Which brings us to this new wakeboarding mechanic.

It's a seemingly simple twist on the formula: you're cruising along the same river of beats, dragged along by two boats whose speeds are synchronized with the track you're playing. The aim is to ride the boats' wake, veering into them just before a large "bump" in the track to jump into the air and perform tricks; bigger jumps means more airtime means more points.

And then things get really interesting: Wakeboarding begets Wakeboarding Grid mode, which bolts the new game mechanic onto the old game mechanic. You'll now be collecting those colored bricks to build up a score multiplier, which will increase the value of every jump. Fill your grid with colored bricks and you'll net a few token points but reset your score multiplier. This  encourages a wary sort of vigilance as you scan the track ahead for the right time to jump, weaving between bricks to fill up your grid incrementally without collapsing the entire stack.

Sound fun? I think it does. But here's (part of) the problem: it's not Audiosurf. See, much of the beauty of the first game was in discovering an entirely new way to visualize your music. I'd spend days wading through MP3s, building up a varied collection of tracks spanning countless genres just to see what they'd look like--Neutral Milk Hotel's "Holland, 1945" remains my favorite. Wakeboarding muddles all of that: instead of feeling the ebb and flow of a song, you're watching out for optimal jump points, as the game rewards spending more time off of the track than on it.

Those of us who've played the same tracks hundreds of times probably won't mind a change of pace, but the new mechanic still feels a little half-baked. The game analyzes a song at the start of every run, and lists every spot with the potential for a massive jump. It's admittedly rather impressive, but also problematic. Since the game knows how much air time you have with every jump you attempt, and knows how long each stunt takes to perform, it'll "helpfully" prevent you from starting tricks you won't have time to finish. This is of course maddeningly infuriating. You'll mash keys or spin your mouse wheel in vain, oblivious to that bit of text informing you the particular trick you're attempting would've required more air time, so you aren't allowed to even try it.

Worse still, wakeboarding isn't even necessarily the centerpiece of Audiosurf II. You're just forced to play through it when you first fire up the game, restricted from diving into the classic modes until you've earned a token number of points playing the new mechanic. As Audiosurf relies on your MP3 collection to generate tracks, that means thumbing through audio files until you find a song with just the right elements, to allow for the sorts of jumps you'll need to earn enough points to unlock the modes you actually want to play. If I were new to all this, I'd be throwing up my hands in frustration.

And then there are the more serious problems. This is an early access game, and the interface is... rough, which is being generous. Finding your music is a tedious affair as you'll need to wade through Windows' byzantine file structure. Even then, truncated filenames and difficulty parsing large collections is problematic. This is especially confusing because the first game handled media browsing just fine--I suspect updates will remedy this shortly.

There's also the matter of content: back in 2008 I had a massive hoard of (legally acquired) MP3s. But those hard drives have long been relegated to a storage closet, as a Spotify subscription tackles my voracious audio appetite a heck of a lot more efficiently. The few gigabytes of media I keep around is plenty for giving Audiosurf II a trial run, but without some means of hooking up with more streaming media outlets (Soundcloud is supported), I suspect that the game's potential will be diminished.

Oh who am I kidding, it'll likely just encourage a heck of a lot more piracy.

I remain wary of the early access phenomenon, but the original Audiosurf holds a special place in my heart. Besides, the heart and soul of the experience is in place, and once you get past that wakeboarding roadblock and the difficult interface the world is your oyster. Better still, the game is eminently moddable, and enterprising users have already began tossing up entirely new game modes to try. Fifteen bucks gets you in the door; you can find Audiosurf II on Steam, but I recommend giving the demo of the first game a spin first.