I have two objects in the scene: geo_Franklin and geo_Standard. they are object merged objects stemming from geo1. In geo1 I attempted at SOP level to sblend between the two geometries which have exactly the same number of points and faces but look stylistically different. However I can't seem to get the sblend to work properly between the geo_Franklin and geo_Standard. What I have in the current SOP network of geo1–the sblend– is not the result I want. Can anyone help me out?

The attribute transfer is getting in the way…What are you trying to achieve with it?

I thought I needed the attribute transfer in order to properly morph between two geometry shapes, which is my original intent. The geometry looks a bit different in Franklin than it does in Standard but because they have the same number of points and polygons I thought a sequence blend would be relatively easy. I found that this was not the case.

Sincerely,
markerline

Does it work if you delete the attribute transfer? I just noticed your file is made with 9.1…

Yeah, I use 9.1 because 10 has a bug in it with Adobe Illustrator EPS importing. Not that that matters in this file here. I just noticed that the two point sops, point1 and point2 in geo1 were bypassed. Those should be enabled and the attribute transfer should be in place in order for any proper morphing to come close to the result I want. But it's still not right. I would in essence like to morph the geometry coming from null_FRANKLIN to the geometry coming from null_STANDARD in geo1 which is in essence the same as the two object level geometries: geo_FRANKLIN and geo_STANDARD. sblend4 is close, sblend2 is a completely wrong blend.

you'll not get perfect blend simply because the topologies don't match
so there is no way you can get one from another simply by moving points

if the topologies would have matched but the points were in wrong order then Match Topology SOP can help, but in your case this will not work

but is quite impressive that you managed to have exactly the same point/face/vertex count with different topologies

you'll get better result if you blend before subdividing, but still very far from what you want

a quick question by the way: what is the general houdini approach to morph blending objects of different geometry?

Thanks tamte from Slovakia. I can't take credit for managing to have the same point/poly/vertex count for both soccerballs–it was a chance occurrence, or perhaps if you believe like I do that there are no coincidences it was perhaps for a different reason that I managed to model the two pieces of geometry that way. And to the other poster there is a method to morph two pieces of non-similar geometry which I had asked about previously. It involves something called Volume Mix and right now I am not at the workstation to load the file and give you the specifics that another user discovered but the results don't really work for complex geometry. I wish I knew exactly how it is done professionally. I have heard that people who are experienced modelers know how to constrain their geometry to make the topologies similar.

I guess you could get by with using simple geometry in the outset and then using a switch sop to go to more complex geometry and having the simple geometry as a middle ground where the visual illusion takes place.

First off the Match Topology SOP is the tool to use when some geometry has shuffled geo orders but these two soccer balls are substantially different toplogies.

Many approaches to doing blending of geometry and quick camera cuts in your shot sequence are a good way to get in and get the heck out.

You can super-res up one of the balls and then ray intersect it with the other soccer ball and pick up the color attributes. Now blend the same soccer ball.

See the attached file.

General approach to morph blending? There isn't “one” sorry to say. There are many many ways to do this. Study all the SOPs and get creative.

Well Jeff that solution was indeed a creative one. I was on the right track using attributes but didn't realize I could go for Cd. I further refined the result in mantra by increasing your subdivision from 2 to 3 and then appending a subdivide before the ray4 on the right-hand side of the green netbox2. This seemed to give a better illusion and I know you were just giving an overview with your file. I guess the higher the resolution of the geometry the better your tv commercial will look. I just saw one ad last night where there were green gel-like monsters or cleaning agents in humanoid form that seemed to morph from just spheres and they had limbs and all. I guess metaballs was probably used in that ad since they jiggled slowly as they stood still. You're right, Houdini def offers more than one approach to a solution.

Regarding the jelly character, you could use the jiggle chop to really good effect for action like this.