Anyone who was ever a fan of Final Fantasy VII has by now I'm sure seen the FF7 HD "tech demo" that was put out in 2005 and which they have utterly failed to develop into an actual HD remake. I mentioned before that the remake seemed to be stalled based on the sense by the developers that they never really recaptured the magic, and when I think about the differences between modern action-heavy game aesthetics and those of the original PS1 period, it's clear they need to think very heavily about tone.
I mean, obviously, that tech demo is meant to show off the engine and their model for Cloud, but let's compare it to the original, specifically the moment when Cloud jumps off the train. The original can be found here, for reference, and the modern version is here.
The most obvious difference is that in the actual game, Cloud is not the center of attention. Because that's outside the scope of the tech demo, it's hard to compare apples to apples; however, the opening cinematic is very much us being dragged along by something a lot bigger than we are. In the game, it's unclear what's even going on; sure it's thugs getting off a train, but to what point and purpose? Until your first battle, you never even know that Cloud is an ultra-badass with a sword as big as he is. This is the first part of the PS1 aesthetic: there is a difference between the battle character and the story character.
Let's explore that. Because of game logic, you know Cloud carries that giant-ass sword with him literally wherever he goes. He carried it while crossdressing in Wall Market, for chrissakes--why would Don Corneo pick a woman carrying around a giant sword? Why would the guards let her in? The story character does not have the sword as part of his model, and the tone of the story would be inconsistent if he DID have it as part of his model. You need some sort of conceit, whether it's the sword disappearing, or just game logic. I would argue that the best choice would be to leave it at game logic and separate the two worlds as in the existing game, but what really happens is up to the developers.
Now, let's look at a game staple--the tutorial. Final Fantasy VII had an interesting take on it, where there are tutorial rooms where the main character, instead of being lectured about how to do things, lectures other people. It includes completely un-subtle nods at the fourth wall, including a bouncing save point. It is campy fun, and honestly, that's another part of the PS1 aesthetic: not taking yourself too seriously. The two tutorial rooms (Beginner's Hall as they're called) are basically the developers, in a move more common in that age than the current one, basically saying to the players, "Hey, there's no way for us to make this make sense in context. A giant tutorial menu would be too much, but there's no way to fit this into the actual game without ruining the tone. We tried, but we decided having it separate is best. So, we'll have some fun with it, I hope it's good for you too."
Which, tangentially, is something that I think a lot of industries are missing, software development high among them: a personal touch. There is a service that I used to use, which lived off of donations, and they posted every month their target number, and the sum total of how much had been donated as well as by whom. The target number of course was low; they were just trying to stay in business, and when they didn't get enough donations, it came out of the owner's pockets. In addition, there was a chatbox on site and it was encouraged that people use it. It all adds up to a situation where you know how sincere the admin is being; if he was asking bazillions of dollars for nothing, or if he never seemed to care about the site, etc, you would know and would be able to find out. In modern corporate culture, there is a strong divide between the game staff and the real world consumers, which I am going to argue is part of modern culture that needs to go away: call it a sense of shame about creating games. Games are made to make money, but the people who make them are not wanting to make a crappy product--in those situations where a crappy product is the result, they have to decide on a corporate level whether they're going to own up to fixing large sections of the game, and make the game into something that makes people happy, or whether they are going to cut losses and accept a shitty product.
As far as that goes, you could call it the difference between a tailor's shop and a Wal-Mart-style clothing store; if a tailor offers you a shirt that does not fit, you take it back and they tailor it to you. If a button pops off, they can sew it on. That sense of pride in the result is a dedication on their part to doing the job right, even after they thought the job was already done. Wal-Mart, by contrast, if they sell you a shitty shirt, will sell you another shirt if you don't like it. That's the modern games industry; certainly you will see patches to fix bugs in a game, but you will not see people go back to games that were broken and un-break them. Once the game is done, it either goes in a pile of shame or a pile of glory.
(You will notice this goes back to my previous article's mention of revision as a part of the artistic process. There are games that you can tell needed revision and didn't get it. A decision was made to keep pushing in the current direction until it ends up in the shame pile, even though you can tell from a thousand feet away that that's where it's going. Programmers, designers, really any part of the staff who are part of that--if they loved games to begin with--will end up being ashamed of being part of the games business, and that's bad--when you bring those same programmers to new projects, they won't give it their all because they're disillusioned about what happens to games with good potential. Games projects need to have revision baked into their budgets, both for time and money, and baked into their dev process to avoid this.)
But we've gotten away from the central point of this article, which is a discussion of tone. A big problem with full 3d models of a character, which you didn't have with the PS1 approximate models, is that you very quickly hit that accursed "uncanny valley" problem, and that can absolutely kill the tone of a story. Perhaps the most iconic bit of 3d modelling in FF7 was the "Cloud Shrug", where Cloud turns, faces the camera, and gives an exaggerated shrug. This is emblematic of a key part of the game's tone; because the backgrounds were pre-rendered, the camera was fixed at theatrical differences, by which I mean you were looking at the scene as though you were in a theater looking down on a stage. The characters on the stage, because they were too far away for you to see them, made large, sweeping movements.
Compare this to the PSP game Crisis Core, in which Zack's model remains constant the entire game. Zack still tries to do the Cloud Shrug (and more notoriously, squats) in spite of the fact that those actions are not really all that literal in FF7. Those large movements are means to an end, and that end is conveyance. In a world as shitty as the Shinra-led Midgar, sometimes heroes have to throw up their hands and say "I'm not stepping into that mess"--or other variations on the same--a single move can mean a lot of different things under different circumstances. Which is, again, what makes simple, theatrical movements far more complicated than full 3D models.
If it were me, I think, I would prepare the game to have different levels of detail for any particular scene, and aim to have a consistent tone between them, even if that means giving up focusing on making the full 3D models perfect. For example, offer a 3D chase camera as with modern third person games (to show off your totally hot 3D surroundings), but also offer fixed scenic camera angles if the user so chooses. Design the levels so that every fixed camera angle available works on every map (puzzles excepted), but so that the 3D camera (when offered) is never something you regret using. An illustrative example are the air ducts in Midgar when escaping the trains tunnels; the fixed camera angle works, but you probably don't want to allow the free-rotating 3D camera because you are unlikely to see anything other than someone crawing in a duct. Similarly for the Shinra boardroom spying duct--you can offer multiple fixed cameras, and arguably you can allow the camera to wander, but why?
The other part of the mixed closeup/fixed scenic camera is distance. It is an understandable conceit that you want the 3d models for everything to be amazing for your uber high resolution displays, but if the camera angle chosen is scenic and the character is just a small portion of the screen, in order to properly convey the character's actions, you still need large theatrical movements. It may seem odd to the texture modelers when, in close-up, your character makes exaggerated movements, however, if those movements match the exaggerated scenic movements, then the audience will easily draw the parallel. Alternately, you can have a separate set of high-definition movements for when you are using a closeup camera, but be aware you will still want to match the tone between the two sets of emotes.
Wrapping up, let me get back around to the beginning. Tech-demo Cloud jumps off of the train, sword in hand, with a very serious look on his face, exaggerated movements, and so on, but the camera is trying to focus on it to make him seem badass. That's appropriate for a tech demo, but not appropriate for an ff7 remake, because it substantially shifts the tone. The first two people to get off the train are NPCs, and each takes down a Shin-ra guard. Cloud also being able to take on a guard or two does not suggest that he is a superior fighter, as each of them takes out a guard in a single move, same as Cloud does in your first battle. The characters then tell the audience what will soon be shown--that there is something special about Cloud. They have mixed reactions, which is again appropriate for people who are about your peers in combat.
The Cloud that jumps off the train in the HD render is the Cloud we know by the end of the game, all abuzz with hints that he will eventually be peerless. Except he won't ever be peerless--as a player character his advancement is in step with every other player character, including the ...indescribable Cait Sith. He does not look like what we all know he is--a Final Fantasy character. The entire series is full of characters with various weapons of various kinds, and every player character eventually levels up into a badass, no matter what weapon they use, from giant swords to tiny little arm-mounted returning boomerangs that make no sense (Rinoa), to bare fists (Tifa, Zell, others), to rapiers, to guns (Barret among many--how do you even improve gun damage when you level up?) to inflatable sports balls (Wakka). The weapon does not make the badass. Instead, the weapon is part of the character, in looks and in tone. However getting back to what I said, arguably large sections of the game outside of combat prefer the character's tone as it is when he is not actively toting the sword around.
In the end, to understand Final Fantasy VII, I think you need to accept that the developers acknowledge the fourth wall and the cracks in it. Cloud has his sword on him even when he doesn't, because it's a game. He turns towards the camera and shrugs sometimes because he's sharing a moment with us, the viewers, about how the whole thing is sort of "whatever". Not because the game developers were ashamed of the game they had made--but instead because they got to a point where they were having fun but needed to move on, and with a nod to the viewers, did exactly that. And that's good; it's good to be on the receiving end of an inside joke, and it's good to feel a bond with the character, even when it's more like an actor on stage apologizing to the crowd for the silliness before going back to the play. It's fun, and that's what we came here for.
I love you, Final Fantasy, and I hope you get through your existing problems and figure out what you are missing. We've had a lot of fun in the past and I still hold high hopes for you.