Isosurface picking #3858
Isosurface picking #3858
Conversation
There was a problem hiding this comment.
Very impressive! The hover effect is awesome and makes this feature look and feel very futuristic, great job 👍
Out of interest, how did you solve the problem you had during development, where the hit test did not work all the time? With the clippingOffsetFactor?
| @@ -183,6 +183,7 @@ export function rotate3DViewTo(id: OrthoView, animate: boolean = true): void { | |||
| const pos = voxelToNm(dataset.dataSource.scale, getPosition(state.flycam)); | |||
|
|
|||
| const aspectRatio = getInputCatcherAspectRatio(state, OrthoViews.TDView); | |||
| const clippingOffsetFactor = 100000; | |||
There was a problem hiding this comment.
Could you add a comment explaining why this is needed? :)
| ); | ||
|
|
||
| raycaster.setFromCamera(mouse, this.cameras[OrthoViews.TDView]); | ||
| const intersections = raycaster.intersectObjects(isosurfacesRootGroup.children, true); |
There was a problem hiding this comment.
Why do we need to intersect recursively? Is this because the segments for one id are in a group or is there another reason?
There was a problem hiding this comment.
Yes, it's because the segments are in a group and the group is only traversed if recursive is true. I'll add a comment!
| @@ -137,6 +143,69 @@ class PlaneView { | |||
| } | |||
| } | |||
|
|
|||
| performIsosurfaceHitTest(): ?THREE.Vector3 { | |||
There was a problem hiding this comment.
Did you experiment with throttling this function? To me, during testing, the TD view felt a little sluggish and I was wondering whether this could help to make it feel more performant :) I do understand that the "instantness" of the hover effect contributes to the "coolness" of the future and I do want to keep that, so maybe just throttling it a little is enough.
It's not a blocker though as the performance hit is only noticeable when isosurfaces are loaded.
There was a problem hiding this comment.
You are absolutely right. I definitely needed to add throttling, as it doesn't make sense to compute this on every render. I opted for 150 ms as a throttling delay, which felt like a good trade-off between performance and delay to me.
URL of deployed dev instance (used for testing):
Steps to test:
Issues: