trigger script when viewport camera is moved

thomas

Hey!

I am working on a python plugin that sends renders to a webserver. It works well right now through a "render and send image" button in a dialog window the plugin creates.
I would love to send images when the viewport camera is moved or something changes in the scene with a visual impact, with a little timeout.
basically the same as the redshift IPR, it updates only when something relevant to the scene changes.

ChatGPT helped me find one way to do this by listening to messages.

class SceneChangeListener(c4d.plugins.MessageData):
    
    BASE_URL = "http://127.0.0.1:8000/inputs"  # Base part of the URL

    def __init__(self, plugin):
        self.last_render_time = 0
        self.plugin = plugin

    def CoreMessage(self, id, bc):
        # print(f"core message received: {id}")
        if id == c4d.EVMSG_CHANGE:
            print(f"my message received:{id}")
            current_time = time.time()
            if current_time - self.last_render_time >= 0.6:  # 300ms delay
                print(f"CoreMessage render triggered:{id}")
                self.last_render_time = current_time
                self.plugin.render(url=self.BASE_URL)
                # Trigger your render function here

        return True

this works but it gets triggered a lot for anything.

Next step is filtering a bit by messages. Ideally only when something that changes the main view, or I listen to only the messages I want to respond to, such as material manager, timeline, camera moves.

I played around with a few more messages I found here:
https://developers.maxon.net/docs/py/2024_2_0/consts/MSG_C4DATOM_PLUGINS.html. None of the few I tried were really that useful.

I guess these dont exist? Happy to try another approach.

i_mazlov

Hi @thomas,

You can use c4d.plugins.MessageData for creating the timer that would handle the "little timeout" part for you.

For the editor camera being changed you can access if through the BaseDraw as shown in How to get the active editor/object camera.

Generally speaking, I'd suggest using the SceneHookData plugin type and register it using the RegisterSceneHookPlugin function. This would give you more flexibility in checking different aspects of the document to treat them as a trigger for you to send data to the server (e.g. you could use EXECUTIONPRIORITY for plugging into directly where you need in the Scene Execution Pipeline, and the EXECUTIONFLAGS::CAMERAONLY could be something particularly of your interest). The downside is that scene hooks aren't exposed to python, so you need to develop C++ plugin to use it.

Cheers,
Ilia

thomas

@i_mazlov thanks so much!
some super useful insights here. I think I will have eventually to move to C++