CompuServe Thread

#Text Morph Probs.

14 messages in this thread
#81688From: Jeff ALuFeb 6, 1994 6:25 PM
Hello, I'm attempting to morph some text, and am having a few problems. I'm using Barrel.fnt, but when I morph the text, things look pretty bad. I know that each morphed object must have the same # of verticies, so I am adding verticies using "refine" in the 2-D Shaper, lofting with optomise off in the Lofter, and I do end up with the same # of verticies using this method, but it is the numbering of the verticies which seems to get messed up. I've tried messing with "choosing" the first verticie in the 2-D Shaper, but have had no luck. Is there a method to the madness of morphing text with a different # of verticies? Thanks! Oh, I'm trying to morph on the "letter" level, not the "word" level… Jeff
#81757From: John K. JordanFeb 6, 1994 11:28 PM
Jeff, >>text morphing… Generally, the best way to make foolproof morph objects is by modifying an original object in the 3D editor until it is in the shape of the desired morph target. You can loft different shapes to morph between, but it requires more care. I just got done with some beveled text morphing last week. I don't know if what I did will help you, but just in case: I had to morph between the word "NEWS" in a straight block font to the word "News" in a curved stylized font (both fonts came from CorelDraw thru the AI method). I brought the shapes for "NEWS" into the shaper and froze them. I positioned a copy of the "News" shapes approximately over the "NEWS" shapes. Then, one vertex at a time, I repositioned the vertices of each letter of "News", adjusting the segments until "NEWS" was traced precisely. By lofting the original "News" and the adjusted copy "News" (now in the shape of "NEWS"), I had sets of objects which morphed perfectly. (Good news, but confusing, huh?. <g>) Two things to watch out for using this method: First, when a curved segment is repositioned, even if adjusted to make it appear linear, it must remain a "curve" so the number of vertices will come out right. (Also, each segment which started out as linear in one font must remain that way.) This may require planning ahead by creating some linear-looking curved segments on the source shape. Note that when you adjust a vertex between a linear and curved segment, the linear segment may accidentally be turned into a curve and must be re-"linearized". Second, I had to morph between a lower case "e" (with a hole) and an upper case "E" (with no hole). This required cutting the morph target "e" at an appropriate point and spreading it to create the morph object. Note this is not a particularly quick or easy thing to do! It took me hours from start to finish, including some bumbling around. I was using two shapes (an outline) for each letter so I could loft bevels, so I had to do each letter twice! Cutting the beveled "e" was especially entertaining. <g> JKJ PS: Sunday eve disclaimer: There may be a better idea for this but I didn't have it.
#81781From: Jeff ALuFeb 7, 1994 1:57 AM
John, thanks for the info. Sounds like an involved process, but it makes sense. I'll try it out. Some of the morphs I need to do are to words of different lengths, posing still more problems. Sheesh. Jeff
#81784From: John TissavaryFeb 7, 1994 2:20 AM
Have you thought about doing the morphs on an image level? You'd be surprised at the quality-versus-ease of use ratio. John Tissavary (La Luna cie)
#82031From: Jeff ALuFeb 7, 1994 10:23 PM
John, I've got a Windows morph program, but haven't had the chance to really check it out yet, so now looks like the right time. Jeff
#81829From: Yost GroupFeb 7, 1994 9:54 AM
As JohnT says, if you're morphing words of different lengths, you should probably do the morphs on the image level. 3D morphs require the face/vertex lists to match (as JohnJ described), which can make morphing between dissimilar objects incredibly difficult. – G
#82050From: Jeff ALuFeb 7, 1994 10:28 PM
Yeah, 3-D morphs can indeed be difficult. Any plans to add dissimilar Verticie #s morphing in r4?
#82250From: Jonas Ruikis [ADESK]Feb 8, 1994 12:44 PM
Jeff, << morhing dissimilar # of verts in R4 >> I'll add it to the wish list. jonas[adesk]
#82279From: Yost GroupFeb 8, 1994 2:22 PM
I know of at least one IPAS developer who's working on this for an IPAS3 plug-in.
#81834From: Feb 7, 1994 10:22 AM
Sounds like a 2D morphing program would make your life a whole lot easier right now! We use BlackBelt WinImages:Morph, and love it! May be the best $199 you spend this month! <g>
#82049From: Jeff ALuFeb 7, 1994 10:26 PM
Greg, yeah, I'm going to check out the 2-D Morph thing. BTW, just stopped by Pyros the other day to get Nancy Fulton's book. Seems really informative. Anything else on the press I should know about? Jeff
#82227From: Feb 8, 1994 11:53 AM
Jeff: Sorry I missed you when you stopped by! Are you coming to the open house tomorrow? (Wednesday, 10:00 -> 5:00) Inside 3D Studio should be shipping later this month, 3DS Tips and Tricks series is excellent, maybe I should add descriptions of these to IPAS.TXT. In fact, I think I will! Greg Pyros
#81883From: MARTIN G FOSTERFeb 7, 1994 1:50 PM
sounds like a good technique. I hope you were payed well….
#82061From: John K. JordanFeb 7, 1994 10:48 PM
>>paid well… Oh my! If I started getting paid well this would all start to seem like work instead of fun. Ack!