Hello, I am a newbie here. The issue that I wanted to point out is:
-I have used CityEngine 2012 to create a model and have exported it as colada.
Now I want to use it with Colada for visualization purpose.
- When I drag and drop the Colada file in the online converter, my browser, Chrome crashes.
- The size of my colada file is about 44 MB after compression.
Is it owing to the size of the file? or is it because of some export glitches?
Hi Patrick,
I was able to successfully reduce the size of the collada using smaller textures from City engine, and the online conversion from collada to gltf ran like a dream.
But i have hit a few glitches :
1. The converted gltf doesn't get displayed in the 3D Model example of SandCastle. ( I tried the same procedure, placed the models in the sample data folder and changed the path to access it, yet to no avail)
2. Elsewhere, while integrating a wms layer from geoserver into the WMS example sandcastle, there is this problem that can be termed as partial globe as only one half of the natural earth texture is being shown in the globe.
I looked at this model. The .dae loads in Visual Studio, but I believe the scale is very small as navigation is awkward. In Cesium, I turned up the scale, but still didn’t see it. I also tweaked the shader and checked the render state (both common COLLADA export issues) but it still didn’t appear. I don’t have any more time to investigate.
This is either a CityEngine export or a COLLADA-to-glTF converter issue. I submitted #350. Can I share the model with the maintainer of that project?
Also, I’m not familiar with the WMS code so I suggest starting a separate thread for that question.
I hope it is ok if I add my issue here; it seems to be closely related to this one as well as to this one.
I’m also one of those guys who tries to switch from Google Earth to Cesium, so I’m a complete beginner. I tried the following workflow: Basic data from OSM -> modelling in CityEngine -> export as collada (.dae) -> convert to .gltf -> include/show in Cesium. My first try was successful and I was able to include many individual models in one “main model”. However, when I tried to select individual models (in my case buildings), I got only a transparent skeleton-model (see attached file model_image.png, on the left how it should look like, on the right how it looks).
Since the .dae looks good (Mac Preview, A360 Viewer) I think, referring to the Tutorial, that the problem is somewhere later (i.e. in the converting process).
I tried: a) the online converter ; b) converter build v0.8 and finally, due to this post also c) the “main” converter. In method b) and c) I used also nearly all available options. Due to this post I changed as well TEXCOORD_0 to TEXCOORD_2 since I thought that it could be maybe the other way around in my case ;). The effect is always the same and I get the shown result (see image). The only difference is that I receive this message in method c: “WARNING: unsupported opaque mode:RGB_ZERO fallback to A_ONE”. So maybe that helps?
Thanks for the detailed report. I submitted #2577. I’m not sure yet if this is a bug in the Cesium client or converter. I’ll try to take a closer look over the weekend since we want to support models exported from CityEngine.
Thanks for the fast response. If you need any more information or files let me know. I just added an screenshot of my first try with a model that contains several (i.e. all from the CityEngine-Scene) buildings. As you can see this works perfectly without any issues (apart from the wrong position on the map, but that has nothing to to with the model itself). I tried also to select only a few buildings from the CityEngine scene; sometimes it works, sometimes I get the skeleton-model… The .dae looks always good.
I was able to create models with a small workaround: Export Autodesk FXB (.fxb) files in City Engine; then convert the .fxb file to .dae with Autodesk FBX Converter 2013.3 x64. Then I converted this .dae file without any problems to a .gltf file. This worked fine for my first scene. However, I then created a second scene, following the same workflow I get again the following message “WARNING: unsupported opaque mode:RGB_ZERO fallback to A_ONE” with the result of a “skeleton-model”.