Jump to content

MAV_FOUND_DEGENERATE_TRIANGLES


EnCore Mayne
 Share

You are about to reply to a thread that has been inactive for 1860 days.

Please take a moment to consider if this thread is worth bumping.

Recommended Posts

[First, it looks like you made a prim build, exported and then uploaded as mesh.] Sorry I was confused by all those extra edges :D.  I don't think that is actually this issue and there is NO strikethrough in the forums any longer (boo hiss).

There are several steps that need to be completed to have that work "well".  Your problem is that you are trying to let the uploader make the physics model which is almost never a good plan -- certainly not with your model.

The error message you are getting is most likely because the uploader hates tiny triangles and you have OH SO many up there on that top landing (not sure why).  

Below is an example of what the stairs should look like to be optimum.  I selected the stairs and they are in orange. So see how simple they need to be to work as they should?  Selecting your stairs and using Alt + J_key will get rid of most of those extra edges  -- if not you can remove edgeloops. Then remove the doubles and you will likely have better luck letting the uploader make the physics model. Remember to change your stairs to PRIM after uploading. 

 

Here is a video that might help:

image.png.030e5b15aee7938fab69338a2fe5add8.png

Here is a more complex set of stairs:

image.png.8dd27fea239750243d0adbd5617e107a.png

 

Edited by Chic Aeon
Link to comment
Share on other sites

6 hours ago, Chic Aeon said:

[First, it looks like you made a prim build, exported and then uploaded as mesh.]

nope, built from the ground up, or top down, i've forgotten which. i did build it (in Blender) from a downloaded primset but that was just for sizing.

6 hours ago, Chic Aeon said:

Your problem is that you are trying to let the uploader make the physics model which is almost never a good plan

that's not the case here. although, i've tested a number of combinations for LOD and physic models, both system "generated" and from my own files, it's a throw of the dice every time. one time the uploader will sing my praises allowing my model to exist only to find it forgot to apply the physics model it had previously okayed. kinda makes ya wonder.

6 hours ago, Chic Aeon said:

The error message you are getting is most likely because the uploader hates tiny triangles and you have OH SO many up there on that top landing (not sure why). 

most likely? i'm far from trusting that these error messages mean anything that serve the hapless user. more like the house's "good luck sucker". as to the horrifyingly tiny triangles, taken into scale with the total size of the build (avatar scaled) they just may be. that top landing's odd shape consists of vertices at each corner that have to be joined somehow. the stand alone stairs i started with were all nice and lovely before i had to integrate them into a 45 degree stepped wall. why? i'm asking that same question.

as bad as they look, i've managed to get them uv mapped semi decently but far from finalized. the outside structure (i've just illustrated the underside model here) has the same geometry and depending on just how i angle my tongue, it goes up without a hitch. you'd think the Lab would have some documentation on their alpha toys by now. oh wait, Beq Janus has it:

too many vertices in the physics model. i've got 561 verts (without system Analyzing). oops, disregard all of my ignorant undignified complaints.

Link to comment
Share on other sites

39 minutes ago, Kyrah Abattoir said:

The Vismesh (I assume it's the vismesh) is wasting a lot of triangles. Here is a quick example of what it should look like (ideally)

ridicule me all you like, Kyrah. how else am i expected to learn dammit. i'm no pro. i learn a great deal through the forum interactions. they might seem totally retarded to you but i take "there's no stupid questions" to heart. forgive my inattentiveness to earlier lessons. i no learn good sometimes. that doesn't stop me from wanting to be where i know good people can and do help. a good critical slap is always respected.

by vismesh, i'm assuming you mean a visual model. and yes, the example i posted here was uploaded with the normals out. if the uploader image shows anything discernable, it should show the Blender mesh (image below it) with the same model used for both the vismesh and physics. i was getting frustrated and it seemed like everything i tried was failing. (come to think about it, when has that changed?)

nonetheless, i did make and try numerous low poly models. some fairly similar to your own. part of the problem is my unskilled unpaid "plan" is to grow it organically as the whims come upon me. in its present form there are stairs with a platform and a main building (platform to ground in height) to be added behind the square floored stepped walled platform. under the platform and inside the stairs is to be an avatar accessible space/room. so, the illustration i'm using here is for under the main stairs/platform which will be joined to the main building.

as it's coming along, the physics are proving to be quite a lovely treat for poking my eyes out. my goal is to eventually have the entire vismesh stairs, platform, understairs and building all one object with accessability/useability inside and on top. i'm happy. it's not like it's ruining my life. i'm in no hurry. peace.

i altered the above model (now that i know there's a 256T limit for physics models) down to 52Triangles and will be trying that for the platform/stairs physics. who knows, it just might come together, God willing.

Link to comment
Share on other sites

The same way the uploader "offers" to generate LOD models for you, it also "offers" to generate a collision mesh from your visible mesh. By that I mean the results are typically horrible and their value is more in the fact that the uploader will "accept" the result rather than the result having any value.

A good collision mesh is typically as simple as possible, with as few/no small/narrow triangles and as convex as possible.

It also doesn't have to be a single fused shape and it can sometimes be better to have a few loosely intersecting cubes than a more complex model with no overlaps.

  • Thanks 1
Link to comment
Share on other sites

To be clear -- and for others - physics models need to be as SIMPLE as possible.  The physic model for the foundation and steps (the more complex one) that I posted earlier is here. The steps part has four vertices. There are of course times when one might want individual steps identified in the physics mesh (so to get the physics sounds when walking up) but I opted not to do that on these shallow steps.   

So the general answer to your issue is that your model is still way too complex.  Again, mostly for others reading this. There is no reason at all to have all those extra edgeloops. I am still unclear how you ended up with all that complexity on top when it is a simple cube shape.  So maybe thinking more simply will help. 

Hope it all works out for you. It does take time for that "lightbulb" moment. Actually, for me, they never stop happening :D. 

image.png.b535420612f7d8f615904550d6188451.png

  • Like 2
Link to comment
Share on other sites

1 hour ago, Chic Aeon said:

There are of course times when one might want individual steps identified in the physics mesh (so to get the physics sounds when walking up) but I opted not to do that on these shallow steps.

There are a few more factors to consider when deciding whether to use ramp or step physics. They both have their pros and cons. But unless you are an experienced mesh maker, go for ramp physics in any case. You really need to know what you're doing to be able to make step physics with low LI.

 

2 hours ago, Chic Aeon said:

physics models need to be as SIMPLE as possible.

That should be made as a texture in BIG BOLD LETTERS and posted at every sandbox and builder's work platform in SL. ;)

The only purpose the physics model has, is to keep avatars (and occasionally physical objects) to walk, fall or jump through objects. Yet calculating physics takes up a tremendous lot of a server's capacity so we want to keep it as simple as possible.

Link to comment
Share on other sites

5 hours ago, Kyrah Abattoir said:

The same way the uploader "offers" to generate LOD models for you, it also "offers" to generate a collision mesh from your visible mesh. By that I mean the results are typically horrible and their value is more in the fact that the uploader will "accept" the result rather than the result having any value.

A good collision mesh is typically as simple as possible, with as few/no small/narrow triangles and as convex as possible.

It also doesn't have to be a single fused shape and it can sometimes be better to have a few loosely intersecting cubes than a more complex model with no overlaps.

and the physics/collision model has to have dimensions closely approximating the visual mesh. if not, the uploader tends to do an inhumane muck about it. ask me how i know....

should the collision mesh have some vertex to vertex simiarity or exactitude with the visual mesh?

Link to comment
Share on other sites

2 hours ago, Chic Aeon said:

To be clear -- and for others - physics models need to be as SIMPLE as possible.  The physic model for the foundation and steps (the more complex one) that I posted earlier is here. The steps part has four vertices. There are of course times when one might want individual steps identified in the physics mesh (so to get the physics sounds when walking up) but I opted not to do that on these shallow steps.   

So the general answer to your issue is that your model is still way too complex.  Again, mostly for others reading this. There is no reason at all to have all those extra edgeloops. I am still unclear how you ended up with all that complexity on top when it is a simple cube shape.  So maybe thinking more simply will help. 

Hope it all works out for you. It does take time for that "lightbulb" moment. Actually, for me, they never stop happening :D. 

 

it's a happening!!!!

proof.thumb.PNG.33c98fa721a9b0fae808a35674ea1c4e.PNG

Link to comment
Share on other sites

42 minutes ago, EnCore Mayne said:

and the physics/collision model has to have dimensions closely approximating the visual mesh. if not, the uploader tends to do an inhumane muck about it. ask me how i know....

should the collision mesh have some vertex to vertex simiarity or exactitude with the visual mesh?

Not needed to be exact (in planes physics one often just gets rid of lots of edgeloops and vertices but in cube physics (what I often use as was taught that way long ago) you are just replacing the area that you want to have COLLIDE with avatars and other objects. 

Be sure to apply location, rotation and scale to both model and physics. It sometimes works fine without doing that (and I just had an example this morning when not quite awake I forgot that step and it worked fine) but better to make that a habit.

Also most of these issues are ALL ANSWERED in that video made especially for SL and Opensim.  Just saying. 

proof.PNG

Edited by Chic Aeon
spelling
Link to comment
Share on other sites

1 hour ago, Chic Aeon said:

Be sure to apply location, rotation and scale to both model and physics.

well, aint that funny (in a tragic sorta way). just learned that the other day. there are so many components in this confusing design i forget i've not applied transformation in all layers if i move anything in one.

Link to comment
Share on other sites

On 2/20/2019 at 8:30 AM, EnCore Mayne said:
On 2/20/2019 at 1:14 AM, Chic Aeon said:

The error message you are getting is most likely because the uploader hates tiny triangles and you have OH SO many up there on that top landing (not sure why). 

most likely? i'm far from trusting that these error messages mean anything that serve the hapless user. more like the house's "good luck sucker". as to the horrifyingly tiny triangles, taken into scale with the total size of the build (avatar scaled) they just may be. that top landing's odd shape consists of vertices at each corner that have to be joined somehow. the stand alone stairs i started with were all nice and lovely before i had to integrate them into a 45 degree stepped wall. why? i'm asking that same question.

@Chic Aeon is exactly right. A degenerate triangle is one that is long and thin to the point where the length of the shortest edge is far smaller than the other two. 

You appear to be using an older Firestorm, but if you have a look at the upload using the latest firestorm it (should) block you from even attempting that upload and highlight the problem areas in RED.  The older versions of FS, use the same "highlighting" as other viewers, you can see the problems "highlighted" in your image, it just makes a very poor fist of showing them.

c9d36f82ed31443d18e69492a0ca5825.png

 

As has been noted, the solution is to simplify the shape as much as possible, a slope rather than steps is the typical solution.

As you observe, the bounding box MUST match that of the visible mesh, if not the physics model is stretched (or shrunk) to match. In the case where the stairs have a bannister or balustrade that it not part of the physics model and thus makes the visible model far taller, you have to place a single vertex (it is often easier to place a single trangle) in the physics model that fully corresponds with the visible mesh. This then prevents the stretching.

 

 

 

 

 

  • Like 2
Link to comment
Share on other sites

You are about to reply to a thread that has been inactive for 1860 days.

Please take a moment to consider if this thread is worth bumping.

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
 Share

×
×
  • Create New...