CompuServe Thread

#Vertex number order?

9 messages in this thread
#193500From: GARRY HARGREAVESOct 2, 1995 8:10 AM
Hi Guys, I am morphing some text into another piece of text (dissimilar shapes) so the vertices count should be the same. Whilst the first vertex is easy to choose, how are the other vertices numbered? (For example in a letter "A" is it clockwise from the 1st vertice..or ? (ie: If there is say 15 vertice in a 2D shape (letter A) and you "refine" a segment between vertex 1 and 2 is the new inserted vertex number 16? ….or does it become the new number 2? Is there any way to find out the numbering sequence of the vertice in a 2D shape? In this way I can get vertex 2 to morph with vertex 2 of the new morph target… ThanX GH
#193944From: David GouldOct 4, 1995 3:52 AM
<<Is there any way to find out the numbering sequence of the vertice in a 2D shape?>> In the 2D Shaper use Display/First. This will show you the first vertex in the 2D shape all the remaining vertices follow. So line up the two black, as shown by Display/First and the rest will follow. There is also a command to pick which face you want as the first. DaviD "morph by numbers" GouID
#194274From: GARRY HARGREAVESOct 6, 1995 5:37 AM
Hi DAVID, ThanX for the response…however I may not of explained myself enough? Try this..we want to morph from say a letter "A" to a letter "B". I, in the 2D shaper, we use NEWBLD.FNT we will get: an 11 vertex "A" and a 15 vertex "B". Now we know first vertices should be in line, however we also know that the two shapes should start out with the same vertice count (this will help the resultant lofted mesh morph better – keeping the path a constent). So we need to add 4 vertices into the letter "A". Move the letter "A" over on top of letter "B" – lineup and choose first vertices – but vertice number 2 of the "A" will want morph to vertice 2 of the letter "B". All I'm saying is how do you know the numbering system of the vertices? What clock-wise 1 to 11? AND IF YOU REFINE letter "A" (to have 15 vertices – add 4 more) do these new vertices become vertices 12, 13, 14 and 15 ? or because you refined (inserted) new vertices inbetween vertices 1 & 2 does the vertices count reset ? (ie: 1 becomes 1, 2 becomes 3 and the added vertice (refined) becomes 2 ……?????????? Long winded I know??? CU GH
#194305From: [F] Grant Blaha [Adesk]Oct 6, 1995 12:22 PM
>> ThanX for the response…however I may not of explained myself enough? Try this..we want to morph from say a letter "A" to a letter "B". << Garry, before you knock yourself out trying to pull this off, I just want to say that morphing is not going to work. It's not a matter of matching the vertices, there are also the faces to worry about. The faces that are used for an "A" are not going to work for a "B" even if the vertices are the same. Consider the steps need to make an "A" into a "B". Now there are many ways to do this, but what comes to mind immediately is the left side of the A going vertical, the top right of the "A" bows out, and the bottom right leg of the "A" bends and curves into to form the bottom loop of the "B". You're not going to get that to happen with morphing. Experiment with animating cross sections with Yost Skin or even Tuber2. Metareyes might also give the effect you want. I once did a morphing sequence of S->L->E->M->K. After failing miserably using the shaper, I hand constructed the geometry. I still used morphing, only I hand constructed a lot of the intermediate shapes. The M->K sequence involved making the letters out of cube primatives and moving them around in the keyframer. I actually did the L->E in reverse, because it was easier to make the vertical bars of the E shrink into the L. Think about the intermediate steps it takes to go from one letter to another and animate those, don't waste your time trying to make vertex counts match in the shaper. -Grant Blaha [ADESK]
#194417From: GARRY HARGREAVESOct 7, 1995 8:20 PM
Hi Grant, Hummmm, it did work at all or just didn't look any good? I thought if vertices counts where equal 3ds would do what ever is possible to take the vertice 1 of of one shape and move it to the position of vertice 1 in the other shape. I'm takin' about full 3d meshes, not just a 2d imported into the editor. Second line of defense is welding vertices or tessalating to get counts the same….The A to a B thing could be any letter. Obviously some letters will work better than others? Are you saying the meshes will not even be valid morph targets? CU GH in OZ
#194433From: Michele BousquetOct 8, 1995 12:50 AM
Garry, Reread Morphing & Animation. There's a whole thing on morphing between numbers. This should help. Michele <you ought to be ashamed of yourself>
#194465From: GARRY HARGREAVESOct 8, 1995 3:20 PM
Hi <M> >>>Reread Morphing & Animation. There's a whole thing on morphing between numbers. This should help. Michele <you ought to be ashamed of yourself><<< Actually Morphing & Animation is what stated this discussion? Mike Gibson tried to do your morphing different shapes to find the formula does'nt work! Is it possible he/we have an old copy of the book – one with mistakes still in it? Was it revised? Hope Moscow was cool <g> CU GH
#194570From: Michele BousquetOct 9, 1995 12:30 PM
Mistakes in my book! Impossible! Yes, an old edition (a very old edition!) had a mistake in the formula. I actually ended up taking that exercise out completely because it's really not the best way to morph. I meant, look at the exercise where create 2D shapes of numbers with the same number and orientation of vertices. Loft them with Optimization off, with always the same path, same steps etc. This creates decent morphable numbers. I found it actually worked best when I created an intermediate shape for each morph, eg. an intermediate 2D shape between the 1 and the 2, the 2 and the 3, etc. This avoided all the funny business that happens when you morph to a dissimilar shape. You can probably get Susan to send you the newest ones, but I don't think that exercise has changed much from one edition to the next. So put that in your pipe and smoke it 🙂 Michele PS. I'll be down in a month or so to walk you through the exercise if you still can't get it. (Oh boy I am really being obnoxious today) PPS. How is the cute little on these days?
#194587From: GARRY HARGREAVESOct 9, 1995 2:40 PM
Hi M, Mistakes? What mistakes….yeah I should try and get the updated version . I think it would take some intervention by the author to get a freebee My original question was based around how does 3ds number it's vertices. CW or CCW from the first AND what happens if you refine a segment. ie: if you insert a new vertice between 1 and 2, does the new vertice become 2 and 2 become 3? I believe this is how 3ds works…..see it wasn't a dumb question after all – just misled by incorrect DOC's – 🙂 PS: you didn't tell about your USSR trip – meet any nice ruskies? CU GH in OZ