Hi,
Now I manage to make seamless continuous surfaces (many thanks to Pascal), but sometimes I get surface areas with unexpected big bumps on it. See example on the hood pictured below.
Q: Is there any tool / command for flattening such a surface (preferably gradually until I decide it’s good enough)?
// Rolf
Hi Rolf - MoveUVN is a good tool for this sort of thing. Also, if you’re doing this a lot, check out DragMode, especially the UVN and ControlPolygon options.Ans, while I am at it, Smooth. (note MoveUVN has some UV smoothing tools that are very helpful as well.
But @RIL, ideally your surfaces wuould not have so many control points - one thing to try, if it’s a Sweep2 is
RemoveMultiKnot
followed by
MakeUniform
to lighten up the surface (at usually a small cost in edge accuracy) and make it more tractable for point editing.
-Pascal
Thanks (again) for your quick reply.
… which led to this strange bump (I tried reducing the degree by using Rebuild, but then again the old ghost showed up, a glitch at the front edge).
Anyway, I’ll examine your advice to see if it applies in this case.
// Rolf
can you post the curves?
-Pascal
Yess.
I did as you suggested (“RemoveMultiKnot followed by MakeUniform”), and fixed a few remaining knots manually, and the result can be seen in the picture. I think it’s good enough. 
Thanks again for your excellent advice.
// Rolf
So… the question I have is, if the surface was an EdgeSrf rather than a Sweep2, and the result had multiple knots… what were the input curves like - can you post the curves in a 3dm file? EdgeSrf is a command that prefers matching curves on opposite sides to make a clean surface.
-Pascal
For the hood I (think I) worked my way forward from the sides, and used the surface edge to the right, and then using curves for the other three sides. As a general rule, I use the surface edges I have (an advice in the AK3D tutorial).
BTW, I realized a bit late that I didn’t really understand your question about what to post, but that’s only typical for newbies. 
My exercise model attached.
// Rolf
Audi R8 A 014.3dm (2.4 MB)
Hi Rolf - I don’t really know how that fender surface is supposed to be - it looks like maybe there is a crease that faces towards the front? Does that surface dip down near the top of the wheel arch? Anyway, I made it up as I went along a little but here are a couple of possible ways to get at this. Basically, using EdgeSrf and working across, to make two matching surfaces you’ll want all the curves in that direction to have the same structure- the red curves. The other direction can be two different structures, one on each surface but matching per surface - white and blue. Note the white curves are degree5/6 point and degree 3/4 point - in each case minimal point count for the degree: That counts as matching structure. The reds are degree 5 and 8 points - 8 because the most complex shaped curve, the marked one, I decided needed 8 points so the others match even if they could be fewer.
Note that on the edgeSrf version the fender surface is skewed forward by the currve at the edge of the hood - this is probably not right for how the surfaces are there - It’s clean enough but you see how reflections & zebra are skewed. On the other one, I started with Sweep2 with an ‘add slash’ at the top of the wheel arch to keep the points from skewing forward in that area. Then to simplify the result, RebuildUV in the U direction. I MatchSrf-ed to the hood, then moved points to break that match at the rear and re-introduce the crease, Anyway, see how the zebra is more controlled and not pulled forward so much above the wheel arch.
Anyway, no idea if that is at all useful, but for what it is worth, those are some ways to think about doing this - as I say I am not familiar enough with just how the shapes transition to be really confident that I’ve done anything good here.
Audi R8 A 014_PG.3dm (155.5 KB)
-Pascal
Wow, thank you for this information. This really increases my understanding, and suggest that I “ascribe more significance” to the point distribution along the uv directions.
It was very useful for me to study your model. I really think that I needed more ways to think about the surfaces.
I also found it interesting how you re-introduced the crease at the rear of the hood. Very useful tips!
Thank you very much for this model. Extremely clear and pedagogic.
// Rolf
