Audio Baking Fails in Blender 5.0 – 'Action' object has no attribute 'fcurves' (Both Generated & Existing Geometry)

[nonexistent3921-web]

Description

In Blender 5.0, the Simple Audio Visualizer add-on throws an AttributeError: 'Action' object has no attribute 'fcurves' when attempting to bake audio.
This occurs in two situations:

When generating cubes through the add-on

When applying the add-on to existing geometry (manually created objects, imported objects, etc.)

Case 1 — When generating cubes
Python: Traceback (most recent call last):
File "...\simple_audio_visualizer_init_.py", line 252, in execute
create_and_bake_cube(O, i, height, L, h, intensity)
File "...\simple_audio_visualizer_init_.py", line 239, in create_and_bake_cube
bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier)
File "...\simple_audio_visualizer_init_.py", line 34, in bake_sound_to_channels
fcurves = action.fcurves
AttributeError: 'Action' object has no attribute 'fcurves'

Case 2 — When using existing geometry
Python: Traceback (most recent call last):
File "...\simple_audio_visualizer_init_.py", line 329, in execute
bake_sound_to_channels(
File "...\simple_audio_visualizer_init_.py", line 34, in bake_sound_to_channels
fcurves = action.fcurves
AttributeError: 'Action' object has no attribute 'fcurves'

Steps to Reproduce

Case 1 — Generated Geometry
Install add-on in Blender 5.0
Load an audio file
Run “Generate Visualization”
Error occurs immediately

Case 2 — Existing Geometry
Create any mesh (e.g., Add → Cube)
Select it → run the add-on’s “Bake Audio” function
Error occurs before baking starts

Expected Behavior
The add-on should:
Create animation data when absent
Create a valid Action for the object if missing
Insert a keyframe before audio baking
Access F-curves only when they actually exist
Bake the audio and generate F-curves cleanly

Actual Behavior
In Blender 5.0:
obj.animation_data is often missing on existing meshes
sound_bake no longer creates or assigns Actions automatically
Actions returned by Blender may be empty or uninitialized
The add-on attempts to read action.fcurves before they exist
Both workflows crash with AttributeError

[YimuQrrr]

Me too

[nonexistent3921-web]

Me too

Ya, just re downloaded Blender 4.5 so I could work on ya personally project with it. I use this add-on a lot so it would be nice if it was later updated to work with 5.0

[sluedecke]

Disclaimer 1: I asked an AI to fix it and it works for me. It is probably not ideal and I don't know about copyright issues here but still.

EDIT: AI says it's ok if it stays GPL which it does, no need to cite it as author, but it was "Grok by xAI" for transparency and I was astounded that it got it right first try.

DISCLAIMER 2: never trust sources pasted into comments, please check it before running, no warranties!

Note: After importing the soundfile into the video sequencer, I had to set the end of the animation to the last frame of the audio BEFORE I run the script. Else any animations created stop at frame 250 (which seems to be the default in my blender).

You need to replace the __init__.py in $HOME/.config/blender/5.0/extensions/blender_org/blender_org/simple_audio_visualizer with this content:

# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTIBILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
import bpy
import time
import os.path
from bpy_extras import anim_utils  # Added for Blender 5.0+ channelbag support

def bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier,
                           anim_loc, anim_rot, anim_scale, anim_x, anim_y,
                           anim_z, additive_mode=False):
    """
    Bakes sound data to specified transform channels of a given object.
    This function handles creating animation data, switching to the graph editor context,
    baking the sound to f-curves, and applying an intensity multiplier.
    """
    # Ensure animation data exists
    if not obj.animation_data:
        obj.animation_data_create()
    action = obj.animation_data.action
    if not action:
        action = bpy.data.actions.new(name=f"{obj.name}Action")
        obj.animation_data.action = action
    
    # Blender 5.0+ migration: Ensure action slot and channelbag
    slot = obj.animation_data.action_slot
    if slot is None:
        slot = action.slots.new(id_type='OBJECT', name=f"{obj.name}_slot")
        obj.animation_data.action_slot = slot
    channelbag = anim_utils.action_ensure_channelbag_for_slot(action, slot)
    fcurves = channelbag.fcurves
    
    # Determine which channels to animate
    props_to_animate = []
    if anim_loc: props_to_animate.append('location')
    if anim_rot: props_to_animate.append('rotation_euler')
    if anim_scale: props_to_animate.append('scale')
    axes_to_animate = []
    if anim_x: axes_to_animate.append(0)
    if anim_y: axes_to_animate.append(1)
    if anim_z: axes_to_animate.append(2)
    if not props_to_animate or not axes_to_animate:
        return # Nothing to do
    
    # If not in additive mode, remove any existing f-curves for the channels
    # we are about to bake to ensure a clean bake.
    if not additive_mode:
        for fc in list(fcurves):
            if fc.data_path in props_to_animate and fc.array_index in axes_to_animate:
                fcurves.remove(fc)
    
    # Find or create the f-curves to be animated
    fcs_to_bake = []
    for prop in props_to_animate:
        for axis in axes_to_animate:
            fc = fcurves.find(prop, index=axis)
            if not fc:
                # Blender 5.0+ migration: Use fcurves.new instead of keyframe_insert
                fc = fcurves.new(prop, index=axis, group_name=prop)
            if fc:
                fcs_to_bake.append(fc)
    
    if not fcs_to_bake:
        print(f"Warning: Could not create any F-Curves for {obj.name}")
        return
    
    # Switch area to Graph Editor for the sound-to-fcurve operator
    original_area_type = context.area.type
    context.area.type = 'GRAPH_EDITOR'
    try:
        # Deselect all f-curves in the action
        for fc in fcurves:
            fc.select = False
        # Select only the curves we want to bake
        for fc in fcs_to_bake:
            fc.select = True
        # Bake sound to all selected F-Curves
        bpy.ops.graph.sound_to_samples(filepath=path, low=lo, high=hi, use_additive=additive_mode)
        bpy.ops.graph.samples_to_keys()
        # Apply intensity multiplier and deselect
        for fc in fcs_to_bake:
            if intensity_multiplier != 1.0:
                for kf in fc.keyframe_points:
                    kf.co.y *= intensity_multiplier
                    kf.handle_left.y *= intensity_multiplier
                    kf.handle_right.y *= intensity_multiplier
            fc.select = False
    finally:
        # Restore the original area type to not disrupt the user's layout
        context.area.type = original_area_type

class SimpleAudioVisualizerProperties(bpy.types.PropertyGroup):
    columns: bpy.props.IntProperty(
        name="Number of cubes",
        default=20,
        min=1,
        description="Number of created objects. Higher values will increase calculation time."
    )
    path: bpy.props.StringProperty(
        name="Audio File",
        subtype='FILE_PATH',
        description="Set the path to the audio file."
    )
    mirror_mode: bpy.props.BoolProperty(
        name="Mirror Mode",
        default=False,
        description="Offsets the origins of the cubes to create a mirror effect."
    )
    intensity: bpy.props.FloatProperty(
        name="Audio Visualizer Multiplier",
        default=1.0,
        min=0.0,
        soft_max=20.0,
        description="Multiplier for the animation's strength"
    )
    animate_location: bpy.props.BoolProperty(
        name="Location",
        default=False,
        description="Animate Location"
    )
    animate_rotation: bpy.props.BoolProperty(
        name="Rotation",
        default=False,
        description="Animate Rotation"
    )
    animate_scale: bpy.props.BoolProperty(
        name="Scale",
        default=True,
        description="Animate Scale"
    )
    animate_axis_x: bpy.props.BoolProperty(
        name="X",
        default=False,
        description="Animate X Axis"
    )
    animate_axis_y: bpy.props.BoolProperty(
        name="Y",
        default=False,
        description="Animate Y Axis"
    )
    animate_axis_z: bpy.props.BoolProperty(
        name="Z",
        default=True,
        description="Animate Z Axis"
    )
    additive_mode: bpy.props.BoolProperty(
        name="Additive",
        default=False,
        description="Add animation from the previous frame to the next, instead of replacing it. Recommended for use with the Rotation channel."
    )
    # Dummy properties for UI tooltips
    info_generate: bpy.props.BoolProperty(
        name="Generate audio visualizer objects in your scene that will be animated based on the selected Audio File. Uses a logarithmic frequency distribution.",
        description="",
        default=False
    )
    info_animate: bpy.props.BoolProperty(
        name="Apply animation data to selected objects in your scene by specifying which transform channels and axes will receive the data. The frequency bands are distributed between the number of selected objects and sorted in alphabetical order.",
        description="",
        default=False
    )

class SimpleAudioVisualizerOperator(bpy.types.Operator):
    bl_idname = "object.simple_audio_visualizer"
    bl_label = "Create Objects"
    bl_description = "Creates new cube objects and animates them to the audio"
    bl_options = {'REGISTER', 'UNDO'}
    def execute(self, context):
        sav_props = context.scene.sav_props
        if not sav_props.path:
            self.report({'ERROR'}, "Please specify an audio file path first.")
            return {'CANCELLED'}
        path = os.path.abspath(bpy.path.abspath(sav_props.path))
        if not os.path.exists(path):
            self.report({'ERROR'}, f"Audio file not found: {path}")
            return {'CANCELLED'}
        start_time = time.time()
        columns = sav_props.columns
        intensity = sav_props.intensity
        total_objects = columns
        spectrum_Start = 10 # Frequency in Hz
        spectrum_End = 21000 # Frequency in Hz
        mirror_mode = sav_props.mirror_mode
        maxheight = 6.0 # Default height value
        # For iterative frequency ranges
        l = 1
        h = spectrum_Start
        base = (spectrum_End / spectrum_Start) ** (1 / columns)
        print(f"STARTED (Cubes_Spectrum={columns})")
        bpy.ops.screen.frame_jump(end=False)
        # ---------------------------------------------------------------------
        # Create Visual Elements
        # ---------------------------------------------------------------------
        def create_and_bake_cube(x, y, zmax, lo, hi, intensity_multiplier):
            bpy.ops.mesh.primitive_cube_add(location=(x, y, 0))
            obj = bpy.context.active_object
            if mirror_mode:
                # In mirror mode, the object is centered on the ground plane (Z=0)
                # and scales symmetrically. The default cube creation already
                # places the origin at the center, so no change is needed.
                pass
            else:
                # Default origin at base of the cube
                bpy.context.scene.cursor.location = obj.location
                bpy.context.scene.cursor.location.z -= 1
                bpy.ops.object.origin_set(type='ORIGIN_CURSOR')
                # Move object up to place its base at z=0
                obj.location.z += 1
            # Scale and apply transform
            obj.scale.x = 0.4
            obj.scale.y = 0.4
            obj.scale.z = zmax
            bpy.ops.object.transform_apply(location=False, rotation=False, scale=True)
            bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier,
                                   anim_loc=False, anim_rot=False, anim_scale=True,
                                   anim_x=False, anim_y=False, anim_z=True)
        # ---------------------------------------------------------------------
        # Generate the Cubes Along the Spectrum
        # ---------------------------------------------------------------------
        for i in range(columns):
            l = h
            h = round(spectrum_Start * (base ** (i + 1)), 2)
            print(f" c: {i} l: {l} Hz h: {h} Hz")
            height = maxheight
            create_and_bake_cube(0, i, height, l, h, intensity)
            # Optional progress display
            done_percent = (i + 1) / columns * 100
            elapsed_minutes = (time.time() - start_time) / 60
            print(f"{done_percent:.1f}% done in {elapsed_minutes:.2f} minutes")
        print(f"CREATED {total_objects} OBJECTS")
        # Reset 3D cursor to the world origin
        bpy.context.scene.cursor.location = (0, 0, 0)
        # Deselect the final object to leave a clean state. This is done
        # without operators to be context-agnostic.
        if bpy.context.active_object:
            bpy.context.active_object.select_set(False)
            bpy.context.view_layer.objects.active = None
        return {'FINISHED'}

class VisualizeSelectedObjectsOperator(bpy.types.Operator):
    bl_idname = "object.visualize_selected_objects"
    bl_label = "Visualize Selected Objects"
    bl_description = "Animate the selected objects based on the audio."
    bl_options = {'REGISTER', 'UNDO'}
    @classmethod
    def poll(cls, context):
        return context.selected_objects
    def execute(self, context):
        sav_props = context.scene.sav_props
        selected_objects = context.selected_objects
        if not sav_props.path:
            self.report({'ERROR'}, "Please specify an audio file path first.")
            return {'CANCELLED'}
        path = os.path.abspath(bpy.path.abspath(sav_props.path))
        if not os.path.exists(path):
            self.report({'ERROR'}, f"Audio file not found: {path}")
            return {'CANCELLED'}
        has_channel = sav_props.animate_location or sav_props.animate_rotation or sav_props.animate_scale
        has_axis = sav_props.animate_axis_x or sav_props.animate_axis_y or sav_props.animate_axis_z
        if not has_channel or not has_axis:
            self.report({'ERROR'}, "Please select at least one transform (Loc/Rot/Scale) and one axis (X/Y/Z).")
            return {'CANCELLED'}
        start_time = time.time()
        columns = len(selected_objects)
        intensity = sav_props.intensity
        spectrum_Start = 10 # Frequency in Hz
        spectrum_End = 21000 # Frequency in Hz
        # For iterative frequency ranges
        l = 1
        h = spectrum_Start
        base = (spectrum_End / spectrum_Start) ** (1 / columns)
        print(f"STARTED (Visualizing {columns} selected objects)")
        # Sort objects by name for consistent frequency distribution
        # Jump to the first frame to ensure sound baking starts from the beginning
        bpy.ops.screen.frame_jump(end=False)
        sorted_objects = sorted(selected_objects, key=lambda o: o.name)
        original_active = context.view_layer.objects.active
        for i, obj in enumerate(sorted_objects):
            context.view_layer.objects.active = obj
            l = h
            h = round(spectrum_Start * (base ** (i + 1)), 2)
            print(f" Animating: {obj.name} l: {l} Hz h: {h} Hz")
            bake_sound_to_channels(
                obj, context, path, l, h, intensity,
                sav_props.animate_location,
                sav_props.animate_rotation,
                sav_props.animate_scale,
                sav_props.animate_axis_x,
                sav_props.animate_axis_y,
                sav_props.animate_axis_z,
                sav_props.additive_mode)
            done_percent = (i + 1) / columns * 100
            elapsed_minutes = (time.time() - start_time) / 60
            print(f"{done_percent:.1f}% done in {elapsed_minutes:.2f} minutes")
        context.view_layer.objects.active = original_active
        print(f"FINISHED Animating {columns} objects")
        return {'FINISHED'}

class SimpleAudioVisualizerPanel(bpy.types.Panel):
    bl_label = "Simple Audio Visualizer"
    bl_idname = "GRAPH_EDITOR_PT_simple_audio_visualizer"
    bl_space_type = 'GRAPH_EDITOR'
    bl_region_type = 'UI'
    bl_category = "Simple Audio Visualizer"
    def draw(self, context):
        layout = self.layout
        sav_props = context.scene.sav_props
        # Global settings
        layout.prop(sav_props, "path")
        layout.prop(sav_props, "intensity")
        layout.separator()
        # Box 1: Create new objects
        box_create = layout.box()
        row = box_create.row()
        row.label(text="Generate New Objects")
        row.alignment = 'RIGHT'
        row.prop(sav_props, "info_generate", text="", icon='QUESTION', emboss=False)
        box_create.prop(sav_props, "columns")
        box_create.prop(sav_props, "mirror_mode")
        box_create.operator("object.simple_audio_visualizer", text="Create Visualizer")
        # Box 2: Animate selected objects
        box_visualize = layout.box()
        row = box_visualize.row()
        row.label(text="Animate Selected Objects")
        row.alignment = 'RIGHT'
        row.prop(sav_props, "info_animate", text="", icon='QUESTION', emboss=False)
        # Sub-box for toggles
        sub_box = box_visualize.box()
        col = sub_box.column(align=True)
        row = col.row(align=True)
        row.prop(sav_props, "animate_location", toggle=True)
        row.prop(sav_props, "animate_rotation", toggle=True)
        row.prop(sav_props, "animate_scale", toggle=True)
        row = col.row(align=True)
        row.prop(sav_props, "animate_axis_x", toggle=True)
        row.prop(sav_props, "animate_axis_y", toggle=True)
        row.prop(sav_props, "animate_axis_z", toggle=True)
        box_visualize.prop(sav_props, "additive_mode")
        row = box_visualize.row()
        row.enabled = bool(context.selected_objects)
        row.operator("object.visualize_selected_objects", text="Visualize Selected Objects")

# --------------------------------------------------------------------------
# Registration
# --------------------------------------------------------------------------
def register():
    bpy.utils.register_class(SimpleAudioVisualizerProperties)
    bpy.types.Scene.sav_props = bpy.props.PointerProperty(type=SimpleAudioVisualizerProperties)
    bpy.utils.register_class(SimpleAudioVisualizerOperator)
    bpy.utils.register_class(VisualizeSelectedObjectsOperator)
    bpy.utils.register_class(SimpleAudioVisualizerPanel)

def unregister():
    bpy.utils.unregister_class(SimpleAudioVisualizerPanel)
    bpy.utils.unregister_class(VisualizeSelectedObjectsOperator)
    bpy.utils.unregister_class(SimpleAudioVisualizerOperator)
    del bpy.types.Scene.sav_props
    bpy.utils.unregister_class(SimpleAudioVisualizerProperties)

[sluedecke]

Here is the output of diff -u:

--- __init__.py	2025-11-29 17:46:52.545383222 +0100
+++ __init__.py.orig	2025-11-29 17:27:01.656777852 +0100
@@ -10,10 +10,11 @@
 #
 # You should have received a copy of the GNU General Public License
 # along with this program. If not, see <http://www.gnu.org/licenses/>.
+
 import bpy
 import time
 import os.path
-from bpy_extras import anim_utils  # Added for Blender 5.0+ channelbag support
+
 
 def bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier,
                            anim_loc, anim_rot, anim_scale, anim_x, anim_y,
@@ -26,66 +27,65 @@
     # Ensure animation data exists
     if not obj.animation_data:
         obj.animation_data_create()
+    if not obj.animation_data.action:
+        obj.animation_data.action = bpy.data.actions.new(name=f"{obj.name}Action")
+
     action = obj.animation_data.action
-    if not action:
-        action = bpy.data.actions.new(name=f"{obj.name}Action")
-        obj.animation_data.action = action
-    
-    # Blender 5.0+ migration: Ensure action slot and channelbag
-    slot = obj.animation_data.action_slot
-    if slot is None:
-        slot = action.slots.new(id_type='OBJECT', name=f"{obj.name}_slot")
-        obj.animation_data.action_slot = slot
-    channelbag = anim_utils.action_ensure_channelbag_for_slot(action, slot)
-    fcurves = channelbag.fcurves
-    
+    fcurves = action.fcurves
+
     # Determine which channels to animate
     props_to_animate = []
     if anim_loc: props_to_animate.append('location')
     if anim_rot: props_to_animate.append('rotation_euler')
     if anim_scale: props_to_animate.append('scale')
+
     axes_to_animate = []
     if anim_x: axes_to_animate.append(0)
     if anim_y: axes_to_animate.append(1)
     if anim_z: axes_to_animate.append(2)
+
     if not props_to_animate or not axes_to_animate:
-        return # Nothing to do
-    
+        return  # Nothing to do
+
     # If not in additive mode, remove any existing f-curves for the channels
     # we are about to bake to ensure a clean bake.
     if not additive_mode:
         for fc in list(fcurves):
             if fc.data_path in props_to_animate and fc.array_index in axes_to_animate:
                 fcurves.remove(fc)
-    
+
     # Find or create the f-curves to be animated
     fcs_to_bake = []
     for prop in props_to_animate:
         for axis in axes_to_animate:
             fc = fcurves.find(prop, index=axis)
             if not fc:
-                # Blender 5.0+ migration: Use fcurves.new instead of keyframe_insert
-                fc = fcurves.new(prop, index=axis, group_name=prop)
+                obj.keyframe_insert(data_path=prop, index=axis, frame=context.scene.frame_current, group=prop)
+                fc = fcurves.find(prop, index=axis)
             if fc:
                 fcs_to_bake.append(fc)
-    
+
     if not fcs_to_bake:
         print(f"Warning: Could not create any F-Curves for {obj.name}")
         return
-    
+
     # Switch area to Graph Editor for the sound-to-fcurve operator
     original_area_type = context.area.type
     context.area.type = 'GRAPH_EDITOR'
+
     try:
         # Deselect all f-curves in the action
         for fc in fcurves:
             fc.select = False
+
         # Select only the curves we want to bake
         for fc in fcs_to_bake:
             fc.select = True
+
         # Bake sound to all selected F-Curves
         bpy.ops.graph.sound_to_samples(filepath=path, low=lo, high=hi, use_additive=additive_mode)
         bpy.ops.graph.samples_to_keys()
+
         # Apply intensity multiplier and deselect
         for fc in fcs_to_bake:
             if intensity_multiplier != 1.0:
@@ -169,40 +169,54 @@
         default=False
     )
 
+
+
 class SimpleAudioVisualizerOperator(bpy.types.Operator):
     bl_idname = "object.simple_audio_visualizer"
     bl_label = "Create Objects"
     bl_description = "Creates new cube objects and animates them to the audio"
     bl_options = {'REGISTER', 'UNDO'}
+
     def execute(self, context):
         sav_props = context.scene.sav_props
+
         if not sav_props.path:
             self.report({'ERROR'}, "Please specify an audio file path first.")
             return {'CANCELLED'}
+
         path = os.path.abspath(bpy.path.abspath(sav_props.path))
         if not os.path.exists(path):
             self.report({'ERROR'}, f"Audio file not found: {path}")
             return {'CANCELLED'}
+
         start_time = time.time()
+
         columns = sav_props.columns
         intensity = sav_props.intensity
         total_objects = columns
-        spectrum_Start = 10 # Frequency in Hz
-        spectrum_End = 21000 # Frequency in Hz
+        spectrum_Start = 10    # Frequency in Hz
+        spectrum_End = 21000  # Frequency in Hz
+
         mirror_mode = sav_props.mirror_mode
-        maxheight = 6.0 # Default height value
+
+        maxheight = 6.0  # Default height value
+
         # For iterative frequency ranges
         l = 1
         h = spectrum_Start
         base = (spectrum_End / spectrum_Start) ** (1 / columns)
+
         print(f"STARTED (Cubes_Spectrum={columns})")
+
         bpy.ops.screen.frame_jump(end=False)
+
         # ---------------------------------------------------------------------
         # Create Visual Elements
         # ---------------------------------------------------------------------
         def create_and_bake_cube(x, y, zmax, lo, hi, intensity_multiplier):
             bpy.ops.mesh.primitive_cube_add(location=(x, y, 0))
             obj = bpy.context.active_object
+
             if mirror_mode:
                 # In mirror mode, the object is centered on the ground plane (Z=0)
                 # and scales symmetrically. The default cube creation already
@@ -215,30 +229,38 @@
                 bpy.ops.object.origin_set(type='ORIGIN_CURSOR')
                 # Move object up to place its base at z=0
                 obj.location.z += 1
+
             # Scale and apply transform
             obj.scale.x = 0.4
             obj.scale.y = 0.4
             obj.scale.z = zmax
             bpy.ops.object.transform_apply(location=False, rotation=False, scale=True)
+
             bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier,
                                    anim_loc=False, anim_rot=False, anim_scale=True,
                                    anim_x=False, anim_y=False, anim_z=True)
+
         # ---------------------------------------------------------------------
         # Generate the Cubes Along the Spectrum
         # ---------------------------------------------------------------------
         for i in range(columns):
             l = h
             h = round(spectrum_Start * (base ** (i + 1)), 2)
-            print(f" c: {i} l: {l} Hz h: {h} Hz")
+            print(f"     c: {i}   l: {l} Hz    h: {h} Hz")
+
             height = maxheight
             create_and_bake_cube(0, i, height, l, h, intensity)
+
             # Optional progress display
             done_percent = (i + 1) / columns * 100
             elapsed_minutes = (time.time() - start_time) / 60
-            print(f"{done_percent:.1f}% done in {elapsed_minutes:.2f} minutes")
+            print(f"{done_percent:.1f}%   done in {elapsed_minutes:.2f} minutes")
+
         print(f"CREATED {total_objects} OBJECTS")
+
         # Reset 3D cursor to the world origin
         bpy.context.scene.cursor.location = (0, 0, 0)
+
         # Deselect the final object to leave a clean state. This is done
         # without operators to be context-agnostic.
         if bpy.context.active_object:
@@ -246,49 +268,64 @@
             bpy.context.view_layer.objects.active = None
         return {'FINISHED'}
 
+
 class VisualizeSelectedObjectsOperator(bpy.types.Operator):
     bl_idname = "object.visualize_selected_objects"
     bl_label = "Visualize Selected Objects"
     bl_description = "Animate the selected objects based on the audio."
     bl_options = {'REGISTER', 'UNDO'}
+
     @classmethod
     def poll(cls, context):
         return context.selected_objects
+
     def execute(self, context):
         sav_props = context.scene.sav_props
         selected_objects = context.selected_objects
+
         if not sav_props.path:
             self.report({'ERROR'}, "Please specify an audio file path first.")
             return {'CANCELLED'}
+
         path = os.path.abspath(bpy.path.abspath(sav_props.path))
         if not os.path.exists(path):
             self.report({'ERROR'}, f"Audio file not found: {path}")
             return {'CANCELLED'}
+
         has_channel = sav_props.animate_location or sav_props.animate_rotation or sav_props.animate_scale
         has_axis = sav_props.animate_axis_x or sav_props.animate_axis_y or sav_props.animate_axis_z
+
         if not has_channel or not has_axis:
             self.report({'ERROR'}, "Please select at least one transform (Loc/Rot/Scale) and one axis (X/Y/Z).")
             return {'CANCELLED'}
+
         start_time = time.time()
+
         columns = len(selected_objects)
         intensity = sav_props.intensity
-        spectrum_Start = 10 # Frequency in Hz
-        spectrum_End = 21000 # Frequency in Hz
+        spectrum_Start = 10    # Frequency in Hz
+        spectrum_End = 21000  # Frequency in Hz
+
         # For iterative frequency ranges
         l = 1
         h = spectrum_Start
         base = (spectrum_End / spectrum_Start) ** (1 / columns)
+
         print(f"STARTED (Visualizing {columns} selected objects)")
+
         # Sort objects by name for consistent frequency distribution
         # Jump to the first frame to ensure sound baking starts from the beginning
         bpy.ops.screen.frame_jump(end=False)
+
         sorted_objects = sorted(selected_objects, key=lambda o: o.name)
         original_active = context.view_layer.objects.active
+
         for i, obj in enumerate(sorted_objects):
             context.view_layer.objects.active = obj
             l = h
             h = round(spectrum_Start * (base ** (i + 1)), 2)
-            print(f" Animating: {obj.name} l: {l} Hz h: {h} Hz")
+            print(f"     Animating: {obj.name}   l: {l} Hz    h: {h} Hz")
+
             bake_sound_to_channels(
                 obj, context, path, l, h, intensity,
                 sav_props.animate_location,
@@ -298,26 +335,33 @@
                 sav_props.animate_axis_y,
                 sav_props.animate_axis_z,
                 sav_props.additive_mode)
+
             done_percent = (i + 1) / columns * 100
             elapsed_minutes = (time.time() - start_time) / 60
-            print(f"{done_percent:.1f}% done in {elapsed_minutes:.2f} minutes")
+            print(f"{done_percent:.1f}%   done in {elapsed_minutes:.2f} minutes")
+
         context.view_layer.objects.active = original_active
         print(f"FINISHED Animating {columns} objects")
+
         return {'FINISHED'}
 
+
 class SimpleAudioVisualizerPanel(bpy.types.Panel):
     bl_label = "Simple Audio Visualizer"
     bl_idname = "GRAPH_EDITOR_PT_simple_audio_visualizer"
     bl_space_type = 'GRAPH_EDITOR'
     bl_region_type = 'UI'
     bl_category = "Simple Audio Visualizer"
+
     def draw(self, context):
         layout = self.layout
         sav_props = context.scene.sav_props
+
         # Global settings
         layout.prop(sav_props, "path")
         layout.prop(sav_props, "intensity")
         layout.separator()
+
         # Box 1: Create new objects
         box_create = layout.box()
         row = box_create.row()
@@ -327,41 +371,52 @@
         box_create.prop(sav_props, "columns")
         box_create.prop(sav_props, "mirror_mode")
         box_create.operator("object.simple_audio_visualizer", text="Create Visualizer")
+
         # Box 2: Animate selected objects
         box_visualize = layout.box()
         row = box_visualize.row()
         row.label(text="Animate Selected Objects")
         row.alignment = 'RIGHT'
         row.prop(sav_props, "info_animate", text="", icon='QUESTION', emboss=False)
+
         # Sub-box for toggles
         sub_box = box_visualize.box()
         col = sub_box.column(align=True)
+
         row = col.row(align=True)
         row.prop(sav_props, "animate_location", toggle=True)
         row.prop(sav_props, "animate_rotation", toggle=True)
         row.prop(sav_props, "animate_scale", toggle=True)
+
         row = col.row(align=True)
         row.prop(sav_props, "animate_axis_x", toggle=True)
         row.prop(sav_props, "animate_axis_y", toggle=True)
         row.prop(sav_props, "animate_axis_z", toggle=True)
+
         box_visualize.prop(sav_props, "additive_mode")
+
         row = box_visualize.row()
         row.enabled = bool(context.selected_objects)
         row.operator("object.visualize_selected_objects", text="Visualize Selected Objects")
 
+
 # --------------------------------------------------------------------------
 # Registration
 # --------------------------------------------------------------------------
 def register():
     bpy.utils.register_class(SimpleAudioVisualizerProperties)
     bpy.types.Scene.sav_props = bpy.props.PointerProperty(type=SimpleAudioVisualizerProperties)
+
     bpy.utils.register_class(SimpleAudioVisualizerOperator)
     bpy.utils.register_class(VisualizeSelectedObjectsOperator)
+
     bpy.utils.register_class(SimpleAudioVisualizerPanel)
 
+
 def unregister():
     bpy.utils.unregister_class(SimpleAudioVisualizerPanel)
-    bpy.utils.unregister_class(VisualizeSelectedObjectsOperator)
     bpy.utils.unregister_class(SimpleAudioVisualizerOperator)
+    bpy.utils.unregister_class(VisualizeSelectedObjectsOperator)
+
     del bpy.types.Scene.sav_props
     bpy.utils.unregister_class(SimpleAudioVisualizerProperties)

[ederac001]

Disclaimer 1: I asked an AI to fix it and it works for me. It is probably not ideal and I don't know about copyright issues here but still.

EDIT: AI says it's ok if it stays GPL which it does, no need to cite it as author, but it was "Grok by xAI" for transparency and I was astounded that it got it right first try.

DISCLAIMER 2: never trust sources pasted into comments, please check it before running, no warranties!

Note: After importing the soundfile into the video sequencer, I had to set the end of the animation to the last frame of the audio BEFORE I run the script. Else any animations created stop at frame 250 (which seems to be the default in my blender).

You need to replace the __init__.py in $HOME/.config/blender/5.0/extensions/blender_org/blender_org/simple_audio_visualizer with this content:

This program is free software; you can redistribute it and/or modify

it under the terms of the GNU General Public License as published by

the Free Software Foundation; either version 3 of the License, or

(at your option) any later version.

This program is distributed in the hope that it will be useful, but

WITHOUT ANY WARRANTY; without even the implied warranty of

MERCHANTIBILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU

General Public License for more details.

You should have received a copy of the GNU General Public License

along with this program. If not, see http://www.gnu.org/licenses/.

import bpy
import time
import os.path
from bpy_extras import anim_utils # Added for Blender 5.0+ channelbag support

def bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier,
anim_loc, anim_rot, anim_scale, anim_x, anim_y,
anim_z, additive_mode=False):
"""
Bakes sound data to specified transform channels of a given object.
This function handles creating animation data, switching to the graph editor context,
baking the sound to f-curves, and applying an intensity multiplier.
"""
# Ensure animation data exists
if not obj.animation_data:
obj.animation_data_create()
action = obj.animation_data.action
if not action:
action = bpy.data.actions.new(name=f"{obj.name}Action")
obj.animation_data.action = action

# Blender 5.0+ migration: Ensure action slot and channelbag
slot = obj.animation_data.action_slot
if slot is None:
    slot = action.slots.new(id_type='OBJECT', name=f"{obj.name}_slot")
    obj.animation_data.action_slot = slot
channelbag = anim_utils.action_ensure_channelbag_for_slot(action, slot)
fcurves = channelbag.fcurves

# Determine which channels to animate
props_to_animate = []
if anim_loc: props_to_animate.append('location')
if anim_rot: props_to_animate.append('rotation_euler')
if anim_scale: props_to_animate.append('scale')
axes_to_animate = []
if anim_x: axes_to_animate.append(0)
if anim_y: axes_to_animate.append(1)
if anim_z: axes_to_animate.append(2)
if not props_to_animate or not axes_to_animate:
    return # Nothing to do

# If not in additive mode, remove any existing f-curves for the channels
# we are about to bake to ensure a clean bake.
if not additive_mode:
    for fc in list(fcurves):
        if fc.data_path in props_to_animate and fc.array_index in axes_to_animate:
            fcurves.remove(fc)

# Find or create the f-curves to be animated
fcs_to_bake = []
for prop in props_to_animate:
    for axis in axes_to_animate:
        fc = fcurves.find(prop, index=axis)
        if not fc:
            # Blender 5.0+ migration: Use fcurves.new instead of keyframe_insert
            fc = fcurves.new(prop, index=axis, group_name=prop)
        if fc:
            fcs_to_bake.append(fc)

if not fcs_to_bake:
    print(f"Warning: Could not create any F-Curves for {obj.name}")
    return

# Switch area to Graph Editor for the sound-to-fcurve operator
original_area_type = context.area.type
context.area.type = 'GRAPH_EDITOR'
try:
    # Deselect all f-curves in the action
    for fc in fcurves:
        fc.select = False
    # Select only the curves we want to bake
    for fc in fcs_to_bake:
        fc.select = True
    # Bake sound to all selected F-Curves
    bpy.ops.graph.sound_to_samples(filepath=path, low=lo, high=hi, use_additive=additive_mode)
    bpy.ops.graph.samples_to_keys()
    # Apply intensity multiplier and deselect
    for fc in fcs_to_bake:
        if intensity_multiplier != 1.0:
            for kf in fc.keyframe_points:
                kf.co.y *= intensity_multiplier
                kf.handle_left.y *= intensity_multiplier
                kf.handle_right.y *= intensity_multiplier
        fc.select = False
finally:
    # Restore the original area type to not disrupt the user's layout
    context.area.type = original_area_type

class SimpleAudioVisualizerProperties(bpy.types.PropertyGroup):
columns: bpy.props.IntProperty(
name="Number of cubes",
default=20,
min=1,
description="Number of created objects. Higher values will increase calculation time."
)
path: bpy.props.StringProperty(
name="Audio File",
subtype='FILE_PATH',
description="Set the path to the audio file."
)
mirror_mode: bpy.props.BoolProperty(
name="Mirror Mode",
default=False,
description="Offsets the origins of the cubes to create a mirror effect."
)
intensity: bpy.props.FloatProperty(
name="Audio Visualizer Multiplier",
default=1.0,
min=0.0,
soft_max=20.0,
description="Multiplier for the animation's strength"
)
animate_location: bpy.props.BoolProperty(
name="Location",
default=False,
description="Animate Location"
)
animate_rotation: bpy.props.BoolProperty(
name="Rotation",
default=False,
description="Animate Rotation"
)
animate_scale: bpy.props.BoolProperty(
name="Scale",
default=True,
description="Animate Scale"
)
animate_axis_x: bpy.props.BoolProperty(
name="X",
default=False,
description="Animate X Axis"
)
animate_axis_y: bpy.props.BoolProperty(
name="Y",
default=False,
description="Animate Y Axis"
)
animate_axis_z: bpy.props.BoolProperty(
name="Z",
default=True,
description="Animate Z Axis"
)
additive_mode: bpy.props.BoolProperty(
name="Additive",
default=False,
description="Add animation from the previous frame to the next, instead of replacing it. Recommended for use with the Rotation channel."
)
# Dummy properties for UI tooltips
info_generate: bpy.props.BoolProperty(
name="Generate audio visualizer objects in your scene that will be animated based on the selected Audio File. Uses a logarithmic frequency distribution.",
description="",
default=False
)
info_animate: bpy.props.BoolProperty(
name="Apply animation data to selected objects in your scene by specifying which transform channels and axes will receive the data. The frequency bands are distributed between the number of selected objects and sorted in alphabetical order.",
description="",
default=False
)

class SimpleAudioVisualizerOperator(bpy.types.Operator):
bl_idname = "object.simple_audio_visualizer"
bl_label = "Create Objects"
bl_description = "Creates new cube objects and animates them to the audio"
bl_options = {'REGISTER', 'UNDO'}
def execute(self, context):
sav_props = context.scene.sav_props
if not sav_props.path:
self.report({'ERROR'}, "Please specify an audio file path first.")
return {'CANCELLED'}
path = os.path.abspath(bpy.path.abspath(sav_props.path))
if not os.path.exists(path):
self.report({'ERROR'}, f"Audio file not found: {path}")
return {'CANCELLED'}
start_time = time.time()
columns = sav_props.columns
intensity = sav_props.intensity
total_objects = columns
spectrum_Start = 10 # Frequency in Hz
spectrum_End = 21000 # Frequency in Hz
mirror_mode = sav_props.mirror_mode
maxheight = 6.0 # Default height value
# For iterative frequency ranges
l = 1
h = spectrum_Start
base = (spectrum_End / spectrum_Start) ** (1 / columns)
print(f"STARTED (Cubes_Spectrum={columns})")
bpy.ops.screen.frame_jump(end=False)
# ---------------------------------------------------------------------
# Create Visual Elements
# ---------------------------------------------------------------------
def create_and_bake_cube(x, y, zmax, lo, hi, intensity_multiplier):
bpy.ops.mesh.primitive_cube_add(location=(x, y, 0))
obj = bpy.context.active_object
if mirror_mode:
# In mirror mode, the object is centered on the ground plane (Z=0)
# and scales symmetrically. The default cube creation already
# places the origin at the center, so no change is needed.
pass
else:
# Default origin at base of the cube
bpy.context.scene.cursor.location = obj.location
bpy.context.scene.cursor.location.z -= 1
bpy.ops.object.origin_set(type='ORIGIN_CURSOR')
# Move object up to place its base at z=0
obj.location.z += 1
# Scale and apply transform
obj.scale.x = 0.4
obj.scale.y = 0.4
obj.scale.z = zmax
bpy.ops.object.transform_apply(location=False, rotation=False, scale=True)
bake_sound_to_channels(obj, context, path, lo, hi, intensity_multiplier,
anim_loc=False, anim_rot=False, anim_scale=True,
anim_x=False, anim_y=False, anim_z=True)
# ---------------------------------------------------------------------
# Generate the Cubes Along the Spectrum
# ---------------------------------------------------------------------
for i in range(columns):
l = h
h = round(spectrum_Start * (base ** (i + 1)), 2)
print(f" c: {i} l: {l} Hz h: {h} Hz")
height = maxheight
create_and_bake_cube(0, i, height, l, h, intensity)
# Optional progress display
done_percent = (i + 1) / columns * 100
elapsed_minutes = (time.time() - start_time) / 60
print(f"{done_percent:.1f}% done in {elapsed_minutes:.2f} minutes")
print(f"CREATED {total_objects} OBJECTS")
# Reset 3D cursor to the world origin
bpy.context.scene.cursor.location = (0, 0, 0)
# Deselect the final object to leave a clean state. This is done
# without operators to be context-agnostic.
if bpy.context.active_object:
bpy.context.active_object.select_set(False)
bpy.context.view_layer.objects.active = None
return {'FINISHED'}

class VisualizeSelectedObjectsOperator(bpy.types.Operator):
bl_idname = "object.visualize_selected_objects"
bl_label = "Visualize Selected Objects"
bl_description = "Animate the selected objects based on the audio."
bl_options = {'REGISTER', 'UNDO'}
@classmethod
def poll(cls, context):
return context.selected_objects
def execute(self, context):
sav_props = context.scene.sav_props
selected_objects = context.selected_objects
if not sav_props.path:
self.report({'ERROR'}, "Please specify an audio file path first.")
return {'CANCELLED'}
path = os.path.abspath(bpy.path.abspath(sav_props.path))
if not os.path.exists(path):
self.report({'ERROR'}, f"Audio file not found: {path}")
return {'CANCELLED'}
has_channel = sav_props.animate_location or sav_props.animate_rotation or sav_props.animate_scale
has_axis = sav_props.animate_axis_x or sav_props.animate_axis_y or sav_props.animate_axis_z
if not has_channel or not has_axis:
self.report({'ERROR'}, "Please select at least one transform (Loc/Rot/Scale) and one axis (X/Y/Z).")
return {'CANCELLED'}
start_time = time.time()
columns = len(selected_objects)
intensity = sav_props.intensity
spectrum_Start = 10 # Frequency in Hz
spectrum_End = 21000 # Frequency in Hz
# For iterative frequency ranges
l = 1
h = spectrum_Start
base = (spectrum_End / spectrum_Start) ** (1 / columns)
print(f"STARTED (Visualizing {columns} selected objects)")
# Sort objects by name for consistent frequency distribution
# Jump to the first frame to ensure sound baking starts from the beginning
bpy.ops.screen.frame_jump(end=False)
sorted_objects = sorted(selected_objects, key=lambda o: o.name)
original_active = context.view_layer.objects.active
for i, obj in enumerate(sorted_objects):
context.view_layer.objects.active = obj
l = h
h = round(spectrum_Start * (base ** (i + 1)), 2)
print(f" Animating: {obj.name} l: {l} Hz h: {h} Hz")
bake_sound_to_channels(
obj, context, path, l, h, intensity,
sav_props.animate_location,
sav_props.animate_rotation,
sav_props.animate_scale,
sav_props.animate_axis_x,
sav_props.animate_axis_y,
sav_props.animate_axis_z,
sav_props.additive_mode)
done_percent = (i + 1) / columns * 100
elapsed_minutes = (time.time() - start_time) / 60
print(f"{done_percent:.1f}% done in {elapsed_minutes:.2f} minutes")
context.view_layer.objects.active = original_active
print(f"FINISHED Animating {columns} objects")
return {'FINISHED'}

class SimpleAudioVisualizerPanel(bpy.types.Panel):
bl_label = "Simple Audio Visualizer"
bl_idname = "GRAPH_EDITOR_PT_simple_audio_visualizer"
bl_space_type = 'GRAPH_EDITOR'
bl_region_type = 'UI'
bl_category = "Simple Audio Visualizer"
def draw(self, context):
layout = self.layout
sav_props = context.scene.sav_props
# Global settings
layout.prop(sav_props, "path")
layout.prop(sav_props, "intensity")
layout.separator()
# Box 1: Create new objects
box_create = layout.box()
row = box_create.row()
row.label(text="Generate New Objects")
row.alignment = 'RIGHT'
row.prop(sav_props, "info_generate", text="", icon='QUESTION', emboss=False)
box_create.prop(sav_props, "columns")
box_create.prop(sav_props, "mirror_mode")
box_create.operator("object.simple_audio_visualizer", text="Create Visualizer")
# Box 2: Animate selected objects
box_visualize = layout.box()
row = box_visualize.row()
row.label(text="Animate Selected Objects")
row.alignment = 'RIGHT'
row.prop(sav_props, "info_animate", text="", icon='QUESTION', emboss=False)
# Sub-box for toggles
sub_box = box_visualize.box()
col = sub_box.column(align=True)
row = col.row(align=True)
row.prop(sav_props, "animate_location", toggle=True)
row.prop(sav_props, "animate_rotation", toggle=True)
row.prop(sav_props, "animate_scale", toggle=True)
row = col.row(align=True)
row.prop(sav_props, "animate_axis_x", toggle=True)
row.prop(sav_props, "animate_axis_y", toggle=True)
row.prop(sav_props, "animate_axis_z", toggle=True)
box_visualize.prop(sav_props, "additive_mode")
row = box_visualize.row()
row.enabled = bool(context.selected_objects)
row.operator("object.visualize_selected_objects", text="Visualize Selected Objects")

--------------------------------------------------------------------------

Registration

--------------------------------------------------------------------------

def register():
bpy.utils.register_class(SimpleAudioVisualizerProperties)
bpy.types.Scene.sav_props = bpy.props.PointerProperty(type=SimpleAudioVisualizerProperties)
bpy.utils.register_class(SimpleAudioVisualizerOperator)
bpy.utils.register_class(VisualizeSelectedObjectsOperator)
bpy.utils.register_class(SimpleAudioVisualizerPanel)

def unregister():
bpy.utils.unregister_class(SimpleAudioVisualizerPanel)
bpy.utils.unregister_class(VisualizeSelectedObjectsOperator)
bpy.utils.unregister_class(SimpleAudioVisualizerOperator)
del bpy.types.Scene.sav_props
bpy.utils.unregister_class(SimpleAudioVisualizerProperties)

Tanks a lot!