Good, and thanks for taking the time to explain; gives this sub-tangent a bit of structure.
Because if PD are texture mapping the way you say they are, there is a major obstacle in the way of us getting a livery editor. The work required for them to go back and create decent full coverage UV maps for 1200 cars is large, and I think the probability that would happen is low.
Two things come to mind:
- A livery editor has long been a planned feature. It would surely be something PD is working towards and has been planning for (and experimenting with, if they're sensible).
- The cars are to be re-used on PS4.
These would indicate to me that PD would better produce the (hypothetical, but expected / required) full coverage maps for a livery editor alongside the (hypothetical, but usual) selective maps at the same time, when the model is made, rather than "go back" over the cars. It's probably no more than a few days' work per car, especially with some of the nifty unwrappers available. Some of these are optimised for minimal distortion and maximal contiguity (connectedness, size of "UV islands") for ease of painting in 2D, others to optimise the packing into the map to maximise texture detail - painting in 3D probably has its own requirements, and that may or may not simplify things: as there's no need to intuitively know where the 2D space maps to 3D, the map doesn't need to be so "human intelligible", I'd wager.
Actually, some (most?) cars would only need a selective map for the interior and non-paintable areas like headlights, grilles, reflectors, badges etc. Wheels and tyres seem to be handled separately. That might cut the amount of work down even more. It still leaves a potential quality issue, if the liveries in GT5 are selectively mapped.
Recall that UV maps are usually dimensionless (U and V scaling from 0 to 1, or equivalent; that's also why they tend to map from square textures, as it's easier to "understand"), and that means you can plug in virtually any texture, of any resolution (feasibly).
If they author the PS3 and the PS4 textures at the same time, too, they can simply carry over the models, with the maps saved within them (as is usual) to the PS4 and simply swap in the PS4 textures. Actually, they can make the new textures whenever they want, but it would make sense to author the same content to two detail levels at once (akin to LoD modeling anyway, literally)
However, if that's in doubt then there's still some reasons to think that a livery editor may be possible. If they already use full car UV maps like other racing games do, then it's not exactly a trivial matter to implement a livery editor but it's much, much easier. Within the realm of the possible, you might say.
For the best possible quality, I'd expect they'd want some kind of magic algorithm to generate the map in a semi-optimised way according to what parts of the car have been painted on by the player (it may be easy enough to mask off only those polygons with vector shapes etc. projected onto them, if that's how the system is going to work, say). I have no idea as to the real feasibility of that, especially given you not only have to transfer the livery (preferably as the collection of the descriptions of shapes used, not the texture) but also the mapping as well, in order to share it. It has the massive benefit that you don't need to make separate maps for custom liveries.
I still wonder if the texture quality achievable with or without "magic" is high enough that PD won't be wary of releasing the game with it that way. I'd rather they just gave us what they can do, but you know some people would be quite opinionated if it doesn't live up to the expectations created by other editors. "We" can probably handle that, but would Sony / PD see it that way?
This is the point. If what you say is true, it makes a livery editor highly unlikely. I think that's a statement worth clarifying, and not something merely for idle speculation.
I can't see it being so clear-cut. The main issue for me is not that they wouldn't have time to make the maps (even if they're still struggling, they could do what HKS suggested, God forbid, and just do a handful, or some arbitrary number and not worry about it), but whether it actually looks any good. The quality of the Premiums could be acceptable to most (according to my perception of the general feeling towards them),
if they can achieve that for every single livery a player can create, and not just those that are like the ones we have in the game.
If it's still in doubt what methods PD use, then we are less able to draw conclusions about the likelihood of a livery editor. You might say that it's more possible, but the real truth is that while we still don't have enough information we haven't got anything to make us give up hope either.
Relevant to the thread? I think so.
I don't think we should give up hope at all (quite the opposite, if we can; gingerly), but it's important to ground your expectations according to potential limitations. I don't personally see any reason for PD not to include a livery editor, even if it looks like the Standard cars. I bet many would disagree, and that for me is the real issue if the maps made for our liveries are not equivalent to those made by PD for their liveries.
Anyway, I need to get up-close and personal with some cars in photomode. Initial impressions are that a system like GRiD 2's would work very well, as PD seem to use a lot of decals. But imagine the ability to draw your own base layer using vector shapes, and have the map magically optimised for it, then to project other stuff on top from an array of pre-made decals. Maybe PD will add a decal editor, too. I'd suggest that this wouldn't work well for some types of liveries, though (e.g. Manga / "art" cars etc.) But now, surely, that's too optimistic.
