"Gra-gra" wrote in message news: snipped-for-privacy@4ax.com... : On Mon, 12 Jul 2004 14:26:00 -0700, "David Janes" : wrote: : : >
: >"Gra-gra" wrote in message : >news: snipped-for-privacy@posting.google.com... : >: "David Janes" wrote in message : >news:... : >: > : "Gra-gra" wrote : >: > : I make a boundary surface, 2 curves in one dir, 3 in the other. I cut : >: > : a section through the resulting surface near one of the curves and : >: > : it's nothing like the curve, IOW the surface doesn't go through the : >: > : curve like it's supposed to. Anyone know why? : >: >
: >: > Try changing the order in which you pick the second direction curves. Pick : >them in : >: > order, one end to the other. If this is WF, when you've picked the middle one, : >you : >: > should see a surface form and then continue to the third curve when you've : >picked : >: > the last one. If you don't pick them in that order it will think you're done : >when : >: > you get to the end curve and ignore the middle one. : >: >
: >: > David Janes : >: : >: Thanks David. You have to pick them in order anyway or you get a weird : >: shape, don't you? Or do you mean 3,2,1 instead of 1,2,3? Tried that : >: and it makes no diff. : >
: >I assumed you were on WF or WF2. There they simplified the curve choice and pretty : >much eliminated 'Approx' type curves which roughly guided and helped shape : >surfaces but didn't exactly match them. If you are using the 3rd, middle curve as : >this type of 'Approx' curve, it will never match it exactly and it is not supposed : >to. : >
: >If it is just three second direction curves then they must all connect with the : >first direction curves. So generally you'll have points on the first direction : >curves through which your middle curve is constructed. Are we on the same page? : >speaking the same language? : : Oui, oui. Yeah, the curve in the 2nd dir was a through (2) points : curve, and the 2 points lie on the 2 curves in the 1st direction. I : wasn't using approx surface either, just the regular old boundary : surface. I've been doing this for years and I always seem to find the : inexplicable... : And try it and wrestle with it and fight it and it just won't work right no matter what. Then suddenly, for no reason, it does. Well, here's something else you've probably already thought of, but, for the less experienced in the audience, here goes: end point tangency of each of those second direction curves should to be set the same for the curves to 'take'. It most dramatically effects ability to set higher quality boundary condition on outside boundary curves but I've also had curves fail just because of that small difference. There's probably a troubleshooting check list somewhere for boundary blend surfaces because there's an awful lot like this that can go wrong. I've even taken to building them on other surfaces, making the curve type 'Quilt/Surface' instead of 'Free'. That turns those outer boundary curves into COS, capable of supporting a curvature continuous middle curve. Can't think of anything else at the moment. It'll probably work at some point and you won't even know why. And you'll be left thinking 'Well, all of those years of schooling sure paid off.'