Not sure if you guys come back to these question threads, but I was wondering if it is necessary to do this intermediate retopology or if it was mostly for the sake of the course. I understand that higher-poly retopo is good for beauty renders and animations, but what if you only need the game-ready model and want to save time?
Thanks
The point of this intermediate retopology is specifically due to the hardsurface nature of the robot character. The sculpture is simply too "hand-made" with bumps and lumps everywhere. By retopologizing intermediately, we generate a perfectly smooth, machined quality to our hardsurface shapes.
It's the only way to get super crisp hardsurface details to come through in the final game model + texture bakes.
Does that make sense?
That makes sense. I just figured that a game-ready retopo would have just as crisp, hardsurface edges as a high-poly retopo since it is made of straight edges, but it makes sense that the high-res one is needed for good texture baking. Would you still recommend the intermediate retopo for a purely organic character with no hardsurface features?
No I wouldn't recommend it for the same reason. I would recommend it for this reason: If you wanted to sculpt super-high-frequency details, like skin pores. Because an intermediate retopology transitions a sculpt from dyntopo (does worse with high-polycounts) over to multires (does better with high-polycounts). That's the best way for squeezing the maximum amount of sculpt detail with Blender in my experience.
If you stick purely with dyntopo, Blender will bog down at skin pore detail - even become unusable. But I've been as high as 24 million polys with multires sculpting and it still be workable. Slow and choppy, but workable.
If you don't need to sculpt that high-frequency detail, no intermediate retopo would be necessary.