IDs Buffer

[kitjmv (jmv)]

Hello,
I would like to know if there is a way to get an Ids buffer/visibility buffer.
I see for example that in the API there is the possibility to get the z buffer Rhino.Display.ZBufferCapture.
jmv

[kitjmv (jmv)]

Or, how can I create my own buffer with different objects?
My knowledge stops at the use of DisplayPipeline and ViewCapture.CaptureToBitmap.

[dale (Dale Fugier)]

Hi @kitjmv,

Perhaps you can step back and describe what problem you are trying to solve?

Thanks,

– Dale

[nathanletwory (Nathan 'jesterKing' Letwory)]

No doubt this is to get a selection map to make it faster to determine what object was clicked on in the viewport. See the custom rotation plug-in @kitjmv has been working on. It would eliminate the need to use GetRenderMeshes and do expensive ray hitting he has been asking a lot about recently.

Navigation plugin

Bonjour, I have seen on this forum several messages from users wishing to activate the Ctrl+Shift+RMB combination by default. I am also in this case. Waiting for version 8 and the Rhino.Options.View.RotateViewAroundObjectAtMouseCursor option, Here is a plugin that positions the camera target on the object under the mouse in order to interpret the Ctrl+Shift+RMB behavior. Installation This plugin is on the Yak server and is called AutoCameraTarget (Current version is 1.3.0). [image] Utili…

And some of the related posts RhinoObject.GetRenderMeshes, bug? , Render mesh topology , How do I always switch to wireframe when moving the view? . I thought there were more, but can’t find them now (at least a post with a huge amount of objects in and trying to get render meshes and do fast ray intersection and determining overlapping objects).

[kitjmv (jmv)]

Hello @dale,

Thank you @nathanletwory, you saved me many minutes of English translation!
That’s right, my goal is to quickly retrieve the object under the mouse cursor.
I don’t need a lot of precision, but it has to be fast, very fast.

Yesterday I found a solution with an “empty” display mode (DisplayModeDescription). No shadow, light and others:

And using a DisplayConduit + RhinoView.CaptureToBitmap

corbane/Libx.Fix.AutoCameraTarget/blob/master/Main.cs#L543

But the performance improvement started to show when 30 objects were under the mouse, producing (when meshed) over a million faces.

To avoid overloading the IdConduit, this technique requires me to keep calculating if the objects are visible in the view and extracting the meshes with RhinoObject.GetRenderMeshes.
It’s a shame because Rhino has the full vertex buffer to display objects on the screen.

So I want to know if there is a way as easy as Rhino.Display.ZBufferCapture to get map Ids (or visibility buffer).

With a depth map and identifiers, it would be easy to calculate an intersection. not precise, but fast.

In the end I wrote my long blabla in English… /

[farouk.serragedine]

Separate library edit