I’ve had a couple situations pop up this morning with MatchSrf giving a nonsense result with the “match edges by closest points” option in V7 SR32. It worked fine in the WIP…but oh, it also worked fine after restarting V7? Ugh that’s not helpful.
bad matchsrf.3dm (76.6 KB)
1 Like
Checking it, thanks.
-Pascal
1 Like
I’ve never found the behavior of that tool to be very good, with any version, unless the isocurves are already lined up really well.
This is part of the reason I just use network surface methods, because if isocurves have to be flowing and lined up so well then might as well network everything anyway.
1 Like
Wha…?
NetworkSrf doesn’t give very good surface quality, it’s of limited use. The way I work these days is almost everything is just brute-force point-pushing, with the odd use of MatchSrf to nail down the seams.
1 Like
guess depends how you use it.
not sure how that constitutes better quality. I guess it depends on the geometry. Yes, sometimes ‘trimmed’ data is inevitable, but that doesn’t mean networking data isn’t the hidden underlying foundation of surfaces anyway.
also, who says a netwrk’ed srf can’t be “point pushed” around?
all surfaces are comprised of a network 
yes, sometimes this is inevitable too.
1 Like
Hey if you like surfaces that are all unnecessarily complicated–and therefore hard to point-tweak-- because every input is a fitted approximation and you’re trying to force-fit surfaces to edges instead of getting intersecting surfaces that capture ‘design intent’ better, fill yer boots. 
1 Like
Everything is relative. I only use ‘trimmed’ or ‘oversimplified’ srf’s when it’s inevitable. And, I only use ‘complicated’ ‘forced-fit’ when it’s inevitable.
Yes it all depends on ‘design intent’. Some designs require very high compound curved organic shapes, and some don’t – and/or combinations thereof.
There’s pros n’ cons either way. Yes, the design intent, and user talent or tendencies of methods will ultimately dictate the path of least resistance.
Sometimes however, if one path causes obstructions in workflow, and a user learns to hurdle over it, like say ‘learn how to handle complicated networks more easily’ then breakthroughs may happen that can antiquate past issues associated with things like ‘match srf’ being used on every seam.
What do you do when you have two trimmed srf’s edges that you need to be a nice seem and you can’t match either one to another? Guess it all depends on design intent and ‘tolerances’.
Maybe we should compile example geometry to demonstrate different scenarios where these issues arise, and design intent has to be compromised, etc. – maybe there’s already threads on this idk.
I am speaking with old premises to be honest, cause sub-D’s have changed everything. Yes, we used to have T-splines but sub-D is better imo. Sub-D’s are a whole new level I’ve still not adapted to 100%, but I’m enjoying every step of the way.
Even still, it does take some effort and strategy to switch between different geometry formats – depending on intent, of course.
1 Like