MschA2V Shot Planner:
MschA2V_ShotPlanner turns a scheduled video concept into a block-by-block plan for audio-driven video generation. It organizes the schedule’s timing and prompts, prepares conditioning for each block, pairs each block with the relevant portion of the timeline audio, and records information needed for later generation and transitions. This lets you work from one coordinated timeline while maintaining consistent reference media across the video. The node prepares a plan; it does not render or assemble the finished video.
MschA2V Shot Planner Input Parameters:
prompt_schedule
A required MSCHA2V_SCHEDULE connection containing the scheduled material to plan. The node compiles this schedule into shots and blocks, then uses each block’s timing, prompt, negative prompt, seed, and transition information. No default, minimum, or maximum is specified for this connection.
clip
A required CLIP connection used to prepare text conditioning. It is passed into the conditioning step for each block and encodes that block’s negative prompt. No default, minimum, or maximum is specified.
vae
A required VAE connection used as the video VAE when preparing each block’s audio-video conditioning. No default, minimum, or maximum is specified.
audio_vae
A required VAE connection passed to the audio-video conditioning step for each block. It is a separate input from vae. No default, minimum, or maximum is specified.
timeline_audio
A required AUDIO connection supplying the timeline’s audio. The node reads its waveform and sample rate to determine the full audio duration, then prepares a window of audio for each scheduled block. No default, minimum, or maximum is specified.
master_seed
An integer seed that replaces the schedule’s master seed when their values differ. It affects the schedule used to compile the shot plan; individual blocks also carry their own compiled seed values in the result. Default: 0. Minimum: 0. Maximum: 4294967295.
width
The requested width passed to conditioning for every block. It must be divisible by 32 or the node raises an error. Default: 768. Minimum: 32. Maximum: 1920. Step: 32.
height
The requested height passed to conditioning for every block. It must be divisible by 32 or the node raises an error. Default: 768. Minimum: 32. Maximum: 1920. Step: 32.
fps_override
An optional frame-rate override for the schedule used to compile the plan. Set it to 0 to leave the schedule’s frame rate unchanged; a nonzero value replaces the schedule’s value when they differ. The resulting plan’s frame rate determines how block frame positions are converted into seconds. Default: 0. Minimum: 0. Maximum: 60.
audio_mode
Selects the audio-handling mode passed to conditioning for every block. Available options: lock_source, remix_source, and native. Default: lock_source. This node passes the selected mode onward; the provided source does not define the detailed behavior of each mode.
task_type
Selects the task type passed to conditioning for every block. Available options: auto, t2va, i2va, fl2va, l2va, ref2va, and hybrid. Default: auto. The provided source does not define the requirements or behavior of the individual task types.
add_source_as_reference
A Boolean setting passed to conditioning for every block to control whether the source is added as a reference. Default: True. The provided source does not specify any further effect beyond forwarding this setting.
ref_image_size
Selects the reference-image sizing option passed to conditioning for every block. Available options: match and max. Default: match. The provided source does not define the precise sizing rules for either option.
negative_fallback
An optional multiline string used when a block’s conditioning keyframe has an empty negative prompt. A block’s own nonempty negative prompt always takes precedence. Default: an empty string. No minimum or maximum length is specified.
ref_image_0
An optional IMAGE connection for a global reference image. When connected, it joins the reference-image list passed to conditioning for every block, supporting consistency across the video rather than applying to one block only. No default, minimum, or maximum is specified.
ref_image_1
An optional IMAGE connection for a second global reference image. Connected reference images are collected in socket order and passed to conditioning for every block. No default, minimum, or maximum is specified.
ref_image_2
An optional IMAGE connection for a third global reference image. Like the other reference-image inputs, it applies to every block when connected. No default, minimum, or maximum is specified.
ref_video_0
An optional IMAGE connection passed as the global reference-video input for every block. Its declared socket type is IMAGE; the provided source does not specify how that image data must represent a video. No default, minimum, or maximum is specified.
ref_video_audio_0
An optional AUDIO connection passed as global reference-video audio for every block. It is distinct from the required timeline audio used to determine duration and prepare block audio windows. No default, minimum, or maximum is specified.
ref_audio_0
An optional AUDIO connection passed as global reference audio for every block. It does not replace the required timeline_audio input in this node’s duration and audio-window calculations. No default, minimum, or maximum is specified.
MschA2V Shot Planner Output Parameters:
shot_plan
A single MSCHA2V_SHOT_PLAN value containing the compiled frame rate (fps), total frame count (total_frames), compiled audio path (audio_path), and a blocks list. Each block entry includes its identifiers and whether it is the first or last block in its shot group; prepared positive and negative conditioning; an audio-video latent; a seed override; timing information; block length and frame rate; and fade-in, fade-out, and crossfade frame counts. You can use this structured result in a downstream workflow that accepts a shot plan. It is planning and conditioning data, not a rendered video or audio file.
MschA2V Shot Planner Usage Tips:
- Use
fps_overrideset to0when you want the schedule’s frame rate. Set a nonzero override only when you intend the compiled plan to use a different rate, because frame positions are converted to seconds using the resulting rate. - Choose
widthandheightin multiples of32, such as the default768by768. This satisfies the node’s explicit dimension check. - Connect the full intended timeline audio to
timeline_audio. The node uses its waveform length and sample rate to calculate audio duration before preparing each block’s audio window. - Put a negative prompt on an individual block when it needs specific guidance. Use
negative_fallbackfor blocks whose own negative prompt is empty. - Connect optional reference media when you want it supplied to every block. These sockets are global references, not controls for assigning different media to individual blocks.
- Start with
task_typeset toautoand the other mode defaults unless your downstream conditioning workflow calls for a specific setting. The provided source forwards these choices but does not define their detailed effects.
MschA2V Shot Planner Common Errors and Solutions:
MschA2V Shot Planner: width/height must be divisible by 32, got WIDTHxHEIGHT
- Explanation: This is the node’s explicit dimension error. It occurs when either requested dimension is not a multiple of
32; the message reports the values received. - Solution: Change both
widthandheightto multiples of32within their stated32–1920ranges.
A required input is not connected
- Explanation: The interface requires
prompt_schedule,clip,vae,audio_vae, andtimeline_audio, as well as values for the required settings. A missing connection may prevent the node from running; the exact message depends on the workflow host. - Solution: Check every required socket and connect a compatible source. The reference-media sockets and
negative_fallbackare optional.
Could not read audio waveform or sample rate
- Explanation: The node expects
timeline_audioto providewaveformandsample_rate. An incompatible audio value can fail when it calculates duration; this heading describes a possible failure, not a quoted error string from the node. - Solution: Connect a valid
AUDIOoutput totimeline_audioand check that the upstream audio node completed successfully.
Invalid audio sample rate
- Explanation: Duration is calculated by dividing the waveform’s sample count by
sample_rate. A zero or invalid sample rate can make that calculation fail. The provided source does not specify an exact error message. - Solution: Check the upstream audio source and supply audio with a valid, nonzero sample rate.
Shot timing or conditioning failed
- Explanation: For each block, the node calls timing, audio-windowing, and conditioning functions supplied by other modules. Those functions may report errors, but their specific messages and requirements are not included in the provided source.
- Solution: Note which upstream or downstream component reports the error, confirm that the schedule, audio, model connections, and selected modes are compatible, and consult that component’s error details before changing settings.
