• Which the release of FS2020 we see an explosition of activity on the forun and of course we are very happy to see this. But having all questions about FS2020 in one forum becomes a bit messy. So therefore we would like to ask you all to use the following guidelines when posting your questions:

    • Tag FS2020 specific questions with the MSFS2020 tag.
    • Questions about making 3D assets can be posted in the 3D asset design forum. Either post them in the subforum of the modelling tool you use or in the general forum if they are general.
    • Questions about aircraft design can be posted in the Aircraft design forum
    • Questions about airport design can be posted in the FS2020 airport design forum. Once airport development tools have been updated for FS2020 you can post tool speciifc questions in the subforums of those tools as well of course.
    • Questions about terrain design can be posted in the FS2020 terrain design forum.
    • Questions about SimConnect can be posted in the SimConnect forum.

    Any other question that is not specific to an aspect of development or tool can be posted in the General chat forum.

    By following these guidelines we make sure that the forums remain easy to read for everybody and also that the right people can find your post to answer it.

LOD editor creates invalid model - and bug reporting broken ;)

jimkeir

Resource contributor
1) Open an FSX .MDL file
2) Export it again as FSX .MDL, to prove that the file is valid
3) Use LOD creator, Vertex clustering, Calculate
4) Export as FSX .MDL

Error: An item with the same key has already been added.

Second, the bug reporter crashes. Both bugs present in "stable" 1.4 and in development version dated today :(

System.Web.Services.Protocols.SoapHeaderException: Error Type: APPLICATION WARNING #300, Error Description: String "view_state" not found. at System.Web.Services.Protocols.SoapHttpClientProtocol.ReadResponse(SoapClientMessage message, WebResponse response, Stream responseStream, Boolean asyncCall) at System.Web.Services.Protocols.SoapHttpClientProtocol.Invoke(String methodName, Object[] parameters) at Futureware.MantisConnect.MantisConnectWebservice.MantisConnect.mc_issue_note_add(String username, String password, String issue_id, IssueNoteData note) at Futureware.MantisConnect.Request.IssueNoteAdd(Int32 issueId, IssueNote note) at ASToFra.ErrorHandler.SendError.AddComment(String email, String comment) at ASToFra.ErrorHandler.ErrorHandler.button1_Click(Object sender, EventArgs e) at System.Windows.Forms.Control.OnClick(EventArgs e)
 

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Hi,

I actually got two such bug reports in the last half hour. I guess they are yours then.

Thanks for reporting how to reproduce them, because until now I never knew how to reproduce it. I'll check tonight.
 

jimkeir

Resource contributor
I suppose they will be, I did try to use the in-built reporter. Let me know if you need the exact model I was working with.

(Edit) Forgot to mention, I remember there was a small bug in the animation system regarding the order of transforms, I can add some details for that if you want? (Reminder to self - B748F)
 

jimkeir

Resource contributor
Many thanks for the LOD fixes, I can confirm that I can now save out models with LODs correctly!

I think this might have thrown up another couple of problems though. First, it appears that whenever I add a new LOD, the number of triangles in the source LOD 100 goes up by a similar amount, even if I don't use LOD 100 as the source. Also, if I create a third LOD, the triangle counts on the others already created are changed upwards by a smaller amount too. Please see the screenshots, having created LODs 99, 98 and 1 in that order.

Next, the resulting triangle counts don't seem to make much sense. As I understand it, lower LOD numbers should be lower detail, and similar LODs should be similar detail. Starting with LOD 100 (6532 tris) and using Vertex Clustering with Ground Clamp off, I create LOD 99 as a test and end up with roughly 1/3 of the triangles (2139, with LOD 100 now reporting 7948) when I would have expected that going from 100 to 99 would have a minimal effect. LOD50 also generates 2139 triangles, as does LOD 5 if they are generated with no other LOD levels present.

It appears that it's ignoring the "To LOD" value and always calculating 40% of the source, since if I change the source from LOD100 to any other, I do get a new LOD with a different, smaller number of triangles.

1) Load an FSX model with a single LOD
2) Generate LOD 99 from LOD 100
3) Observe that the number of triangles in the *source* LOD has changed, and that the number of triangles in the *destination* LOD is reduced far more than would be expected
4) Generate LOD 50 from LOD 100
5) Observe that the number of triangles in LODs 100 and 99 have changed, and that the number of triangles in LOD50 is the same as the previous number in LOD99
6) Generate LOD 1 from LOD 100
7) Observe that the number of triangles in LODs 100, 99 and 50 have changed, and that the number of triangles in LOD1 is the same as the previous number in LOD99 and LOD50 when they were created
 

Attachments

  • LODs_01.jpg
    LODs_01.jpg
    22.4 KB · Views: 8
  • LODs_02.jpg
    LODs_02.jpg
    23.8 KB · Views: 7
  • LODs_03.jpg
    LODs_03.jpg
    24.5 KB · Views: 8
  • LODs_04.jpg
    LODs_04.jpg
    27.1 KB · Views: 7

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Hi,

I'll try to reproduce it.

The amount of triangles in the lod is determined by the settings. So if you generate a lod 99 and a lod 50 with the same settings, they get the same triangle count.

So for the lower lods you need to set a bigger grid size in the vertex clustering settings.
 

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Hi,

I can't reproduce the issue that the LOD100 increases in triangles here. I tried a few object, but when I calculate a new LOD, the LOD100 is not changing on those objects.
 

jimkeir

Resource contributor
I can't reproduce the issue that the LOD100 increases in triangles here. I tried a few object, but when I calculate a new LOD, the LOD100 is not changing on those objects.
I've attached one of the models I'm using. It happened on the first four I tried and I stopped there. I'm using 1.5.0.0 4e8c1fc7 DEV 27/02/2021 . Open the model, open the LOD tool, click calculate. LOD100 goes from 6331 tris to 7339. My guess would be that it's counting the number of vertices but including lower LODs because if I remove one of the generated LODs, the triangle count returns to the initial value.

So for the lower lods you need to set a bigger grid size in the vertex clustering settings.
Ah - sorry, my mistake, I've tried what you said and it's working nicely.
 

Attachments

  • b739.zip
    97.5 KB · Views: 7

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Thanks let me try that model.
 

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Thanks, I have been able to reproduce the issue now. The problem was caused by some nested modelparts, these were not cloned correctly. In the next development release it will be fixed.
 

jimkeir

Resource contributor
Thank you, all working nicely :) Now if I can just get the wheels to stay moderately circular...
 

arno

Administrator
Staff member
FSDevConf team
Resource contributor
That work probably work best with the quadratic error method. For the vertex clustering method for the wheels to stay kind of circular the grid size you have to be at least 4 times smaller than the radius of the wheel.
 

jimkeir

Resource contributor
Thanks for the advice - but you win some, you lose some :)

With quadratic, the more curved parts are great. Textures, not so much, well within material boundaries. That kind of distortion is visible from a long way out. Entire parts are disappearing pretty quickly too. If I'm just expecting too much from it please do say so, models are really not my thing...
 

Attachments

  • JFAI_737_900.zip
    875.8 KB · Views: 7
  • Original.jpg
    Original.jpg
    235.5 KB · Views: 12
  • Quadratic 80% 5061 tris.jpg
    Quadratic 80% 5061 tris.jpg
    235.5 KB · Views: 12
  • Vertex 0.3 4085 tris.jpg
    Vertex 0.3 4085 tris.jpg
    235.9 KB · Views: 12

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Hi,

The quadratic error method tries to take the texture mapping into account, but it is not an easy thing indeed. I'll see if I can spot something with the model you attached.

Not sure if you expect too much. Years ago I heard somebody say about a similar LOD tool that it is a great way to make damaged and destroyed models :D
 

arno

Administrator
Staff member
FSDevConf team
Resource contributor
Hi,

Oops, my memory was wrong. The feature to take the texture mapping into account is still on the todo list.
 
Top