Double click on sampled ellipsoid

CesiumJS
1

Hi, We were using an old cesium version (1.10) loading the scenes using CZML and we decided to update it to 1.29. We found that one of our examples didn't work with the new version. I've beeing looking for additional info but I didnt find any similar problem in the forum (sorry if it is dupplicated).

We create an ellipsoid that changes its position and radii on an interval. Once the CZML is loaded in cesium, it points to a random positon on the space. I click the "home" button to move the camera to the earth and I can see the red point of the CZML. When I double click on it, the camera moves again to a random position of the space and I see nothing.

If I use the same czml without the ellipsoid and I repeat previous steps, the sample works fine. I also checked in sandcastle that the ellipsoids with no sampled data also are zoomed fine. So the problem seems to be the use of the ellipsoid + time dynamic data.

Any ideas? Thanks in advance.

Check the code here:

var czml = [{
    "id":"document",
    "version":"1.0",
    "clock":{
      "interval":"2015-11-03T00:00:00Z/2015-11-04T00:00:00Z",
      "currentTime":"2015-11-03T00:00:00Z",
      "multiplier":10,
      "range":"UNBOUNDED",
      "step":"SYSTEM_CLOCK_MULTIPLIER"
    }
  },
  {
    "id":"SAT",
    "name":"SAT",
    "point":{
      "color":{
        "rgba":[
          255,0,0,255
        ]
      },
      "show":true,
      "pixelSize":5
    },
      "position":{
      "interpolationDegree":5,
      "interpolationAlgorithm":"LAGRANGE",
      "epoch":"2015-11-03T00:00:00Z",
      "cartesian":[
        0,41135569.83604452,10529445.08775081,-2592340.399843352,
        9693,24802054.373201564,34533648.14522849,-1254190.962921509,
        19386,-3119320.0214920063,42400209.400282934,670092.773882628,
        29080,-29578699.66305307,30434388.12677617,2280914.529707506,
        38773,-42183770.93257903,4217403.0070346575,2823482.7169600846,
        48467,-35016137.616095856,-23975133.74977094,2042545.558970656,
        58160,-11433606.240267385,-40929571.96347262,303927.2016867829,
        67854,17507171.946269713,-38712023.81421206,-1577394.5068920546,
        86400,41691026.277549595,8031384.871797217,-2657222.265938431
      ]
    },
    "ellipsoid":{
      "radii":{
        "epoch":"2015-11-03T00:00:00Z",
        "cartesian":[
          0,952.1529108575653,160.27592173538292,23.918865122711754,
          9693,901.9691398345186,159.36635646200654,26.238895837307286,
          19386,881.8835050647224,158.5274798119592,27.357939070630113,
          29080,852.6432917149185,157.75854515111405,29.336547217628674,
          38773,814.4931393637447,157.06122428665418,31.654726580888436,
          48467,787.109314109322,156.43847267687002,33.66166228842491,
          58160,752.8332929461376,155.8895157977463,36.13946995177457,
          67854,725.923364204505,155.41721491778128,38.46002659858811,
          86400,693.012092067969,155.02046376495323,41.25095480674983
          ]
      },
      "fill":true,
      "outline":false,
      "outlineColor":{
        "rgba":[
          0,255,0,255
        ]
      },
      "slicePartitions":25,
      "stackPartitions":35
    }
  }];

var viewer = new Cesium.Viewer('cesiumContainer');
var dataSource = Cesium.CzmlDataSource.load(czml);
viewer.dataSources.add(dataSource);
viewer.zoomTo(dataSource);

2

Something appears to be broken in the bounding sphere calculation for ellipsoids. The bounding sphere as computed is clearly wrong:

center: Cartesian3

x: -4799832008.766061

y: 18985081429.637413

z: 842458361.9376999

radius: 19600567306.54599

Position of the entity at this time is:

Cartesian3 {x: 41132641.25709635, y: 10540859.619063426, z: -2592022.591873995}

and radii is:

Cartesian3 {x: 952.132460181332, y: 160.27555107322385, z: 23.919810571729002}

3

Thanks Scott. It looks like this was broken in 1.27.
I’ll write up an issue.

In the meantime, you can upgrade to Cesium 1.26 and your example should work fine. We’ll post a follow up when we fix the problem.

Best,

Hannah

4

For reference, here is the GitHub issue: https://github.com/AnalyticalGraphicsInc/cesium/issues/4866

-Hannah

5

Thank you for the answers. I wanted to update cesium to solve the issue with translucency in firefox 50 (https://github.com/AnalyticalGraphicsInc/cesium/pull/4762), however I will probably consider your advice of updating to 1.26 to use a more recent version until this gets solved. Thank you very much.

6

This was just fixed in #4907, which will be in Cesium 1.30 next week on February 1.

Patrick