Ah. I know there is work to be done.
Just figured X character would be faster than making a brand new character. With all the planning and work that comes with that.
Info drop incoming. This is gonna still be pretty surface and more for explanation purposes.
It is faster, but only because they already have the entire theme of the character, what they're weapon is and their style along with VA. But, you can think of it more like, they have an idea of what the character is already, they still have to flesh it out again as opposed to creating an entirely new idea from scratch.
Keep in mind, because its a new game, and they are rebalancing characters which means, some moves may end up being scrapped, retweaked or redone in order to fit into the direction they want the game to go. For example, they seem to have been scrapping A+K inputs so the problem becomes, do they just want to switch the inputs for the move or get rid of the move entirely. Is the attack moving to fast for the amount of damage it causes, versus reach and safety. If they want to add in new moves, how will they work with other moves already there. Will players believe that the move is possible and does it look "Cool" enough. In some cases adding changes to existing characters becomes more work because you're trying to preserve what player's find interesting and engaging about the character while still trying to add in something fresh and a slightly different spin on it. Playing it safe and not changing much will have people calling it a copy paste scam, where as changing too much will cause players to reject it.
In terms of the character models, I personally am not sure what PS's workflow is so I can only make a few guesses. I also don't know the engine they used for SCV and lost swords and if they have access to it, how much they have and what is reusable. Basically, because its an inhouse engine, if they don't have the original source files they are a number of things they could have done that could have made those assets unusable even before consideration for SCVI. For example, the file formats for their own engine could have taken in standard .obj files, converted them to a much smaller and faster loading custom format which is pretty standard, and now those files are only readable by the engine. If they lost or are unable to understand how to reverse the process because team members change, lack of documentation those files are practically gone. Additionally, there might have been a couple of things they did to the character models in order to make sure they loaded and looked best inside their own engine. In short a lot of cheats and hacks.
If they managed to get those character models, then they're is the whole process of what can they reuse. Remember, they still have to redesign the entire visual aesthetic for the character, even if there isn't going to be much external changes in the base character mesh. They can't really use older game models even if they do have them because....they're squares (more accurately twin triangles but you get the point). If they use the sculpt first approach to get a high poly sculpt of the character then it's much faster to just resculpt the character using the redesigned concept art as opposed to importing in someone else's model and trying to work off that. The old model could have its own set of problems that either Unreal will complain about, will be really annoying to try and fix (duplicate hidden geometry), proportions might not fit etc. There could also be problems with the way the character was modelling and the way the lighting in UE4 works that cause problems with shadowing and other visual effects. At the end of the day, the team can change it is sometimes much faster to resculpt/model the character because if you're doing every other step again, it'll be a better advantage if you have people that know every part of the asset inside out incase something goes wrong.
Textures are going to have to change, I don't really need to explain that. Same with colours that are used and making sure they aren't noisy, but still aesthetically appealing.
Rigs are going to change. They're going to need a new base rig that's used. I have some idea of how they are handling character creation rigs but I'd need more time to flesh out the entire rigging system.
Only then do we get to the actual animations. Which has to take into account things such as making sure they don't expose any problems in the character model. They don't cause visually cause problems for the player. They look and feel correct in regards to the character that is using them. We've all see those weird CAS Asta characters that grab enemies but they have a good cinderblock in between the hand the character. All those animations for each character need to be redone because retargeting from the old rig to the new one will have its own set of issues, on top of all the gameplay stuff I mentioned earlier. The audio needs to match the visual information. They have to build all the hit detentions for all the attacks. Depending on the method that they use, this can be extremely annoying to do in order to get very accurate results. Quick disclaimer, I haven't spent to much time with Motion Capture. I've been the person in the suit and i've looked into a lot of how to work with it but I haven't really imported raw data and worked with it so there could be a couple of things out of place that I am thinking of.
I just realized I'm writing a wall....so Ima cut if off here. Just a quick thing. Working with Unreal is a lot different than what devs might be used to because of all the flexibility but also rigidness of how the engine is built and how you have to work. You can't really copy paste so much stuff because of the way things work and link together.
Tl;dr: Old characters only give a solid theme and idea for the devs to use. Sometimes they can reference older models to speed up production but because they're using a new engine entirely they might opt for doing things over again. Even SCV and LS models might end up being used as a reference for gameplay movelists as well as techniques on the way some of the game works and is displayed instead of taking the character model and slapping a new outfit on it. Theoretically, there is less work that needs to be done but, I would also argue that sometimes it is more work trying to fit within a certain set of parameters. Particularly because this is a fighting game, a lot of asset reuse does not apply as much because of all the other changes that have happened. Might as well redo the stuff in order to bring them in line.
