i have followed few courses from Gamedev, and i found that there are basically two ways to create characters:
There’s no best approach. Each project and each artist are unique, “the best” will depend a lot on what you’re trying to do and the type of character.
Some characters don’t require sculpting at all, some do partially and others a lot.
Keep your mind open to using all methods and tools at your disposal, as you acquire experience you’ll know when to use each tool.
There is also the proper way, lol, modelled directly with good topology.
But as GDTVPrince rightly says it is all down to the personal choices of the person making the object and what they are using or wanting the character for. Some crowd filling background characters will not want the same attention and detail of a man character that is the primary subject of any image or scene.
I think it depends on the level of detail.
Making a head, face, legs and arm with fingers. That can easily be done with the box (mesh modeling) method. But when it comes to small details, like muscles, wrinkles, lips, fabric, etc.
Then you use the box model as a low poly basic for sculpting the high poly.
For that process you need something like the multires modifier.
The truth is you can’t really get away from poly modeling. In most cases, you end up doing retopology after a sculpt which is done using poly modeling. There are arguments for multi-resolutions sculpting and using low poly sculpt instead of retopo or the arguments for auto-retopo tools. Often times those need clean up. So, I think it’s wrong to think of it as either or. Instead think of them as tools. Another way to think about it is as work flows.
Character sculpting workflow: is good for higher detailed(even if it’s just for normal map baking) organic characters. It’s more intuitive for beginners. Some argue that it’s more natural artistic flow. Con is 99% of the time you have to recreate the character with better topology. Requires better time management since it’s easy to move into smaller details to early and even working on details that will get lost. Sharp edges and hard surfaces are harder to achieve.
Character polygon modeling workflow: is great for hard surfaces characters like robots. Also good for low poly character. Have more control over topology. Con it requires more understanding of tools and planning. Not as intuitive.
i see, thank you for the answers