Adding a bunch of unrelated defaults together is not manageable. You can create a separate topic for the pie menus. The other ones are cases where a design was chosen intentionally, and for that reason are not suitable as paper cuts.
I find it odd that the annotations overlay is disabled by default in the sculpting preset… especially now that we have annotations tools in the toolbar in sculpt mode.
It seems to be pertinent defaults to me.
Having an accumulation of brush effect at plane of symmetry does not seem to be the most wanted behavior.
And if you consider that user will begin to test brushes before starting a sculpt session, that does not look weird to have symmetry off.
But directly starting a sculpt session with symmetry with decent brushes makes sense, too.
Mirror ON by default is prone to errors and newbies don’t expect that behaviour. Sculpting symmetric models is the unique case so it don’t have any sense to activate by default. If that have some sense the same logic obligate us to put on by default the mirror in edit mode. Also me with 20 years in the industry I find a lot of times how I have lost a lot of work because I didn’t see the symmetry.
Feather is the normal behaviour that any user wait of the symmetry.
If you start with a primitive as a cube or a sphere, there is a strong probability that you will start to complicate it still by making a symmetrical basemesh shape and only at the end of the work deactivate symmetry.
Most of objects of our world are globally almost symmetrical.
Most of time, an artist decomposes a complex unsymmetrical object into symmetrical parts easier to treat.
If it was not the case, we would have lots of unsymmetrical primitives.
But as you say that is better for newbies to test brushes to have symmetry OFF.
Anyways, I agree that Feather option should be ON by default. It was probably set OFF to avoid a corner case but I don’t remember which one.
Fast Navigate=OFF
It is annoying to work with Fast Navigate=ON. I think the only reason for this enabled by default is the low performance of the new Multiresolution modifier. When modifier has better performance, Fast Navigate should be set to OFF by default.
I’m wondering: has Sculpt Mode always shown triangles only, or is this a recent 2.81 thing? Somehow it only recently occured to me that quad models in Object Mode are displayed showing all triangles in Sculpt Mode:
When you use quadriflow you search a fixed number of faces, for that reason use the face count like parameters is the obvious way
When you use voxel remesh you don’t know what number of faces you want, you know the detail of the model that you need to keep. For that reason the obvious way is use the voxel size. Also that for the workflow of a sculpting artist the logic is keep the parameter of detail in the sculpt process. Not that if your model have now the double of volume that five minutes before you need to recalculate the number of faces that now you need to keep the same detail.
I don’t see any problem with that. It seems more logical to look at size of detail you want to preserve.
Polycount only makes sense for an ultimate remeshing when remesher is able to maintain shape.
It is the case for Quadriflow. But it is not the case for Voxel remesher.
Voxel remesher is faster because it is not working like Quadriflow.
The voxel size value is indicated in unit dimensions. You usually don’t keep track of an object’s size when you’re sculpting, so you need to spend time entering multiple values and remeshing again until the resulting detail is satisfactory.
In my humble opinion a poly count approximation is more intuitive, because you can easily estimate what polygon amount is necessary for your sculpt’s current level of detail. On top of that you can easily keep track of your sculpt not becoming too heavy in polygon count.
If you prefer the current method, then I suggest making the poly count value an option.
I don’t think Remesh is faster because it doesn’t use a polycount approximation. It’s because of a totally different algorithm.
That is basically what will happen with your proposal. Because actually except the first time you won’t change the detail of the voxel. But with face count you won’t know what will be the good number because you can’t extrapolate voxel to faces.
5000 faces are not the same solution in one model than in another, and you will need to change this every time that you made a few changes in the mesh.