Open chris-b-pacon opened 1 year ago
Actually the standard morphs do have different regions attached to them. I was half way doing this already but then got side tracked. What I did was generate a new character from a new fresh install and then filter the morphs available on region. Then my idea would be to set a value != 0 so that the morph is active and save the appearance. You can then load this appearance file and check the json inside to see which morphs are all corresponding to that class.
In the future I'd have the app to have two regions: face and body and the corresponding standard morphs to both. There are still some morphs which influence both (Carmen morph or something, some of the standard morphs in VaM) but it would still be a great start.
Or we would try to find out which morphs actually influence BOTH head and body and fliter those out (but that is a lot of work, and requires manual tinkering). I think.
If you go with more regions (like head / breasts / glutes / arms / legs) for instance, they also have to be rated, which is a lot of work. So I think the best trade off would be head vs body.
Let's sink that in, keep it in the hidden shelves of our minds and come up with a really great idea as soon as possible. Let's keep this open, I still like the idea.
I am pretty much in favor of just splitting into head and body region. This is also the norm for most single morph characters, they have a head morph and a body morph. If you split into more regions handling the rating is going to be too cumbersome imho.
There was a person in the Hub who apparently knows something about morphs files I would like to know. Let's see what we/I can do with this info.
Description
Would be nice to have an option to limit the evolution to specified areas, like only head, or only legs.
I don't know how to implement that, because morphs itself give no explicit information, which area(s) they affects. The only cue is the morph's name, which can be checked to assume what it is meant for, like "morph/head/eye" or "navel".
I imagine two separate options for an approach:
In both approaches the genetic algorithm had to omit the filtered morphs.
Sounds this viable?