Load MD frames

[NAEV95]

Hi! I really like this project and just discovered it yesterday. I was wondering whether it would be possible to have all the frames and edit each frame when loading an MD. E.g I would like to have all the states superimposed to show the overall movement as a picture. Is there a way to do this?

[BradyAJohnston]

This feature could certainly be added. As it is, it would be a little tricky to pull it off with the nodes that are available, but when Blender 4.0 comes out I should be able to add this as a feature without too much trouble.

[NAEV95]

I made a script for this if it can be of help:

import bpy

def duplicate_for_each_frame():
    # Specify the object name of your MD system
    object_name = "Name_of_Your_MD_System"
    
    obj = bpy.data.objects.get(object_name)
    if not obj:
        print(f"Object {object_name} not found!")
        return
    
    # Define the range of frames, starting from the second frame and ending at 250
    start_frame = 2  # Start from the second frame
    end_frame = 250  # Set end frame to 250

    # List to store the duplicated objects
    duplicated_objects = []

    # Modify the range to select every other frame
    for frame in range(start_frame, end_frame + 1, 2):  # Step of 2 for every other frame
        # Set the current frame
        bpy.context.scene.frame_set(frame)
        
        # Duplicate the MD object
        bpy.ops.object.select_all(action='DESELECT')
        obj.select_set(True)
        bpy.ops.object.duplicate()
        
        # Get the duplicated object and set it as the active object
        duplicated_obj = bpy.context.selected_objects[0]
        
        # Rename the duplicated object based on the frame number
        duplicated_obj.name = f"{object_name}_Frame_{frame}"
        
        # Apply all modifiers on the duplicated object
        for modifier in duplicated_obj.modifiers:
            bpy.ops.object.modifier_apply({"object": duplicated_obj}, modifier=modifier.name)

        # Add to the list of duplicated objects
        duplicated_objects.append(duplicated_obj)

        # Deselect the duplicated object
        duplicated_obj.select_set(False)

    # Join all the duplicated objects into a single object
    bpy.ops.object.select_all(action='DESELECT')
    for obj in duplicated_objects:
        obj.select_set(True)
    bpy.context.view_layer.objects.active = duplicated_objects[0]
    bpy.ops.object.join()

# Call the function to execute
duplicate_for_each_frame()
 

This is the final look I got in the end

The issue is that I get extremely large file doing this and the rendering becomes quite long! (5GB)

[BradyAJohnston]

Duplicating objects will be the reason you are getting such large file sizes. This will be able to be done through geometry nodes through a more memory efficient manner, but will likely need to wait for Blender 4.0

[BradyAJohnston]

This can now be achieved in 4.0 using the 'Repat Zone', but only after importing the trajectory with In Memory ticked so that it all imports. This will loop through and for each iteration of the loop, get a new frame of the trajectory and then create a style, and join them all together as instances at the end.

It is quite performant, but beware with more frames realized as you increase the number of iterations it will get computationally more intense.