Man, I really hope Koji Igarashi (and Keiji Inafune, among others) play Odin Sphere. Because I want them to experience the bitter taste of eating their own words.
For years, IGA has been telling us that making a console-based 2D Castlevania would be prohibitively expensive, and Inafune said something similar regarding classic-syle Mega Man games when we interviewed him last month. It’s nice that they’ve explored moving to 2.5D fixed-perspective polygons as a compromise between cost and gameplay (although given the sales of the PSP Mega Man titles we’ll quite possibly never see anything else along those lines for that particular series). And for the most part, I think everyone’s been content to accept their word, because — after all — they are the guys dealing with the realities of game design and budget management.
But then Odin Sphere comes along and makes liars of them all. Here’s a 2D game whose visuals totally shame everything pretty much ever. You may not care much for the style (given the abundance of morons guffawing “HAW HAW LOLI PEDO FURRY PORN HUURRR” any time a screen shot is posted); but even a complete Internet retard can’t deny the sheer technical panache that Odin Sphere displays. It’s pretty much completely amazing to see in motion. Yet — it’s all 2D. Hand-drawn, two-dimensional, sprite-based gameplay. So how did a little never-heard-of-them developer like Vanillaware pull it off where major studios like Konami and Capcom throw their hands up in despair and settle for churning out the likes of Curse of Darkness and Mega Man X7?
I think the trick is twofold. One, Odin Sphere makes abundance use of jointed characters.
Basically, every body in the game moves like an action figure — limbs are connected independently, and a lot of the animation is simply the motion of individual body elements in relation to one another rather than multiple hand-drawn frames. It’s not a new concept! In fact, it’s basically little more than a new application of a fundament of sprite-based gimmickry. Konami’s been dabbling with it for years in games like Lifeforce and Contra — create the illusion of large-scale, sinuous motion with small, linked sprites. Treasure pretty much made a cottage industry of it.
If you ever go ROM-diving to look at sprites, you’ll even discover that the sprites of old NES characters like Mega Man are actually made of multiple smaller sprites. The NES could only generate 8×8 pixel sprites, so something 24 pixels high and 16 pixels was assembled of an array of sprites two wide and three high. Even sprites for later systems, which weren’t hampered by such harsh size restrictions, generally built characters from smaller components so that not every single body part had to be redrawn for every frame of motion.
Jointing characters is simply an extension of that concept, except that than building characters out of smaller pieces serves as the entire foundation of the character’s animation rather than simply being a cheat that allows them to fit them into memory or lets the spriter save himself a little extra work. Symphony of the Night was the first time I ever saw this concept in action — which isn’t to say this was the first time it was ever done, of course. The result was that certain large characters like Galamoth and the Hammer/Blade/Sword animated beautifully even though each of their component elements only had a single bitmap frame. Their animation was in the motion of their limbs parts rather than in those limbs’ changing details. It worked well since it was limited more or less to armored or undead characters whose bodies naturally consisted of static, jointed pieces.
The technique is pretty common in Flash now, but it generally looks awful, like a cheap digital puppet show. Or it looks extremely stylized, as in the Paper Mario games. Odin Sphere does not look cheap, and it’s stylish but not strictly stylized.
That’s simply a matter of quality control. The characters’ joints blend seamlessly, and their animation occurs on two levels: The gross motion of body segments, and the bitmap art within those segments. The animators also make use of subtle tricks like warping and distorting sprite chunks to give a semblance of movement for things like breathing and the like. It’s very sophisticated, and executed with such quality that you really have to look closely to see the man behind the curtain.
So why can’t we have a new Castlevania or Mega Man or Mario (or whatever) that looks like this? Therein lies the second trick: Odin Sphere is a tiny game that no one cared about except its creator. I’ve heard that it’s been in the works in various forms since Princess Crown was released, which pegs this as a decade-long labor of love. Generally speaking, major publishers aren’t crazy about letting their franchises lay fallow for ten years. Hell, Capcom doesn’t seem to want to let Mega Man lay fallow for ten months.
So basically, those classic series which used to represent the bleeding edge of console game design are doomed to future mediocrity because they belong to publishers who find it easier to strip-mine them for easy profits than to give their creators the time they need to create something truly amazing. As development costs rise, I suspect fewer and fewer series will be allowed the downtime it takes for a proper reinvention — if indeed those franchises are ever reinvented at all. Changing a successful formula is a risk, and for every Resident Evil 4 that remind people why a series was so popular to begin with, there are a dozen Breath of Fire: Dragon Quarters — games that take too many risks and despite their quality drive away the fans who just want to replay the same exact game they had last year (but with new graphics) and ultimately flop. We’re in a world of Final Fantasy XIIIs (it’s not a sequel, it’s a brand in and of itself!) and 30 Mario spin-offs for every one game that carries the Super Mario Bros. torch of innovation and creativity.
That’s why Keiji Inafune can create something radical and new with Dead Rising, but has to keep churning out Battle Network sequels on the Mega Man side. If a Mega Man game with the budget of Dead Rising flopped, it would kill the series forever. But low-maintenance repetition, while unsatisfying for everyone, at least keeps the name alive, and profitable. Which in a publisher’s mind is infinitely better than simply making the name prestigious. You can’t take prestige to the bank.
So cherish your Odin Spheres when they come around. George Kamitami and what I assume is a very tiny crew of dedicated auteurs have gotten away with someone here that most developers couldn’t, and most publishers wouldn’t allow. And who knows if they’ll ever manage to pull it off again? The same developer’s upcoming Grim Grimoire is reputed to be a great RTS… but its iteration of the jointed-sprite graphic style looks much cheaper and simpler than Odin Sphere’s. I guess everyone has to pay the bills somehow.
I was waiting to actually play the game before weighing in on all this, but you’ve basically covered everything I would have said. The next time you get to talk to Igarashi, please don’t hesitate to point out Odin Sphere to him.
“for the five or six of you who read both”
Interesting… do I meet the self deprication with encouragement or cower in shame for belonging to the nerdish few. Quite the dilemma.
It uh… is still prohibitively expensive.
Hey, no fair bringing old NES games into the argument. Back then, you had to innovate just to make a game in the first place. Not today, where… Wait. Never mind.
Heh… I also read both, but I read it here first, so I was about to point out that there was so much good stuff in this post that it seemed like the seed of a good 1up feature, or at least article. I guess blog post is close enough for now.
And, Tomm: So that begs the question, where did Vanillaware get bags of money? Genuinely curious. Or did they only pull it off because they had a small team willing to work for a decade for peanuts?
Way to read all the way to the end, Tomm.
Parish, I thought I was the only one who thought like this. But see, this is why I hang around your website (and Sardius’ and TheRedEye’s), because you people perceive that something weird about games that I love. You “know” it man. You get it. And I don’t want you to think I’m kissing ass here just because you work for some bigshot website thingie, that’s part of the reason I like to be anonymous. But man, you get it. It’s what I like to call the late 80’s Sunsoft/Konami/Capcom/(some)Tecmo feeling. It’s that something that can not be described that a game has that makes it just amazing. And, yes, Odin Sphere has it. Resident Evil 4 has it too. Metal Gear Solid has it. Snatcher has it. Bionic Commando has it. Mike Tyson’s Punch-Out! has it.
i had it, but i lost it.. what are we talking about again?
I used to be with ‘it’ until they changed what ‘it’ was. Now I’m not with ‘it’ anymore. And the same thing will happen to you! *points at camera and patiently waits for the dismissive ‘Psh! Whatever, man!’ responce*
Big city, hmm? Live. Work, huh? But. Only peoples. Peoples is peoples. No is buildings. Is tomatoes, huh? Is peoples, is dancing, is music, is potatoes. So, peoples is peoples. Okay?
given the abundance of morons guffawing “HAW HAW LOLI PEDO FURRY PORN HUURRR” any time a screen shot is posted
Honestly, you’re aren’t surfing 4chan, are you?
I should also mention the game The Rumble Fish (and it’s sequel) uses the technique you mentioned.
I did read all the way to the end. If Vanillaware had 10 years to work on it, that’s where the money came from – time. But Nich and I already had this discussion, so he can update you karmically.
Perhaps when they say “prohibitively expensive,” they mean that they think it won’t sell. Because that would be expensive. In other words, it’s the same argument that ‘s been against 2D since the 3D started: “People don’t like it!”
Yes, that’s all well and good, but you completely missed the fact that the lead-in was just a strawman complaint that I knocked down in order to say what you just wrote. Making a game that looks like this is too expensive and risky for entrenched franchises.
To the too expensive argument: Maybe they should be putting out less horse shit and concentrating more on the remaining games. But I know how it works. They’ll cut the good stuff and pour more rotting tripe down the line. Whatever. I’ll just enjoy the good ones when I can and try not to drown in the endless river of schlock. But I do wish all the crap they put out that makes me die a little inside would make them die a little inside too. By which I mean I wish they’d bleed internally.
Recently I’ve been introduced to the concept of a ‘skunk works’ (which is copyrighted, so it’s only called that when it doesn’t matter that you’re stealing other people’s words – which is basically a prototype shop, tasked with working out the feasibility of designs in as cheap and compelling way as possible. I know there’s places like the Experimental Gameplay project, but is there anything like that at the major publishers?
I ask because it seems that reinventing a game or a genre is naturally a risky proposition, and you would think that publishers had come up with some way to reduce that risk in order to get more than one surefire revenue stream. Normally, that’s a skunk works.
(I strongly suspect that part of gaming’s growing pains is that developers have seen people try to run game projects like engineering projects, seen it fail, and then draw the conclusion that games are nothing like other software engineering projects and can’t be managed the same way, which is somewhat analogous to exclaiming that a frog is actually a fish because it has to hop around to get anywhere.)
I’ve been working on a Contra type game for the past two years that uses the same techniques for Direct X 9. Basically the Head, torso, legs, and each arm are roughly 128×128 each. It supports about 50 of these characters at 1280×1078 @ 60fps, with 30 layers of parallax moving grass and a particle/decal system Requirements Geforce 3, P3 800. Unfortunately my hard drive is damaged and I risk losing the last 6 months of work that includes the procedural limb animation code among other stuff, like a separate sprite for a melee weapon. I am waiting on a quote from the data recovery guys ($600-$1500). Arg.
hi Jeremy, could you let us/me know if you tried Odin Sphere on an HD tv? does it do 480p? 16X9? how does it look on an HD set?
I played Odin Sphere on a Samsung LCD HDTV via component cable. No 16×9, but it does apparently support progressive scan because damn it looked good.
There is a very large interview in the most recent Play (it’s right next to the pretty bad review of the game) which states that the game’s production started in 2004, and it’s been finished since sometime in 2006. It only took two years to put together with a team of 14 people. The only things that had been completed before the start of the project were character designs and concept art. People should quit speculating so much: there is no reason that other developers couldn’t have done the same things. On top of that, in the interview it’s stated that the art was down sampled in order to fit on the PS2 and that they could have even made the game in HD.
“If you ever go ROM-diving to look at sprites, you’ll even discover that the sprites of old NES characters like Mega Man are actually made of multiple smaller sprites. The NES could only generate 8×8 pixel sprites, so something 24 pixels high and 16 pixels was assembled of an array of sprites two wide and three high.”
It also bears noting that Mega Man was 5 colors instead of just 3.
Despite Capcom’s insistence on annual rehashes, they still change things up every once in a while. Every 10 years or so, Mega Man seems to add another genre to the ol’ résumé and we end up with a spin-off series like Legends or Battle Network as a result. It seems like you can have your repetitive franchise and diversify it too. Maybe publishers can’t be convinced that it’s a worthy investment, but surely the prospect of striking fresh gold with an old brand has to be tempting.
I think part of the problem is finding a balance between good sprite animation and cost-effective budgets. There’s certainly going to be a happy medium which satisfies us spriteophiles while staying on budget. Certainly good planning and the ever-handy pallette swaps can shorten the development time.
Sprites, from a development standpoint, are a pain in the ass. They either work, or they don’t work and need to be partially (or fully!) redrawn. A 3D model can be morphed, changed, stretched, skewed, and otherwise be tugged around to fit the creator’s purpose. If you’ve created a model with arms which look just a touch too short, it’s simple to lengthen them. If you draw five hundred sprites and discover the arms look just a touch too short, well… It’s going to be a lot more work.
“A 3D model can be morphed, changed, stretched, skewed, and otherwise be tugged around to fit the creator’s purpose.”
A 2D object can be similarly morphed, especially if it’s drawn at a high enough resolution and the animations are done with a (2D) skeletal system and mostly-rigid joints. As long as you’re doing something smarter than traditional frame-by-frame animation, you can bring some of the convenience associated with 3D to 2D.