VNCCS Pose Studio:
VNCCS_PoseStudio turns poses arranged in its browser-based 3D editor into rendered images for a ComfyUI workflow. You can use it to prepare multiple poses, combine them into a grid, or render an animation. The node asks the open browser widget to capture the scene during execution, so the result reflects the editor’s current state rather than relying on an older preview. It can also use a reference image to guide pose import or, in Pose Manager mode, body proportions. Rendering takes place in the browser; this node does not provide a separate backend 3D renderer. You therefore need the workflow open in a browser with WebGL available while it runs.
VNCCS Pose Studio Input Parameters:
pose_data
Required STRING containing the Pose Studio widget’s settings and scene data as JSON. Its default is {}; the field accepts multiple lines and does not expand dynamic prompts. The widget normally supplies this value, including settings that determine image versus animation behavior, output dimensions, list versus grid layout, and captured results. If you edit it manually, keep it valid JSON and use settings produced by the widget: the empty default does not itself contain browser-rendered images. No minimum or maximum text length is specified in the node interface.
pose_image
Optional IMAGE used as a reference for browser-side analysis through the SAM 3D Body import path. In ordinary pose editing, the imported result is applied to the pose; in Pose Manager mode, it can be used to analyze body proportions. In Pose Manager mode, the widget’s manager_auto_analyze_proportions setting can disable this analysis, in which case the supplied image is ignored. When the capture_image_size export setting is enabled, the image’s dimensions are also passed to the browser for capture sizing. No default image or general size range is specified for this input. For Capture Image Size, the source requires valid positive dimensions, with width and height each no greater than 4096 pixels, plus an additional total-pixel limit whose value is not shown in the provided code. If an image batch is connected, the import uses its first image.
camera_prompt
Optional STRING input for camera direction, intended to connect to VNCCS Visual Camera Control. Its default is an empty string; no length or choice limits are specified. The node passes the resolved text to the browser during capture and appends a formatted version to each lighting prompt when applicable. It clears the value unless the widget’s directional_skydome_enabled export setting is explicitly true, so connecting a camera prompt alone does not guarantee that it will affect the result.
animation_image_batch
Optional BOOLEAN that determines how animation frames leave the first output. Its default is False; its options are True and False. In animation mode, True returns all captured frames together as one IMAGE batch, which is useful when a downstream node processes images frame by frame. False returns a VIDEO value instead, provided the installed ComfyUI supports its official VIDEO datatype. This switch has no effect outside animation mode. The widget’s animation_image_batch export setting can also enable batch output.
VNCCS Pose Studio Output Parameters:
images
The first output has the interface type IMAGE,VIDEO; its actual value depends on the widget settings. In image mode with LIST output, you receive the captured poses as separate IMAGE values. In image mode with grid output, you receive one IMAGE containing the captures arranged in columns. In animation mode, you receive either one VIDEO assembled from the captured frames at the animation frame rate or, when image-batch output is enabled, one IMAGE batch containing those frames. Connect this socket to a downstream node that accepts the form you selected. Despite its name, images can therefore carry a video.
lighting_prompt
The second output is STRING text associated with the captures. In list and animation results, it provides a prompt for each captured image or frame; missing entries are filled with empty strings, and an applicable camera prompt is appended. For a grid result, it provides only the first capture’s prompt as the grid’s main prompt, not a combined description of every cell. Use it as accompanying text for your rendered result, and check whether it is empty before relying on it for downstream prompting.
VNCCS Pose Studio Usage Tips:
- Keep the ComfyUI workflow open in a browser with WebGL available until execution finishes. The node requests a fresh browser capture on each run and cannot render the scene on the backend if that capture is unavailable.
- Choose LIST output in the Pose Studio widget when you need to process poses individually. Choose grid output when you want a single contact sheet; set the widget’s grid columns to control its arrangement.
- In animation mode, leave
animation_image_batchatFalsefor a VIDEO-compatible workflow, or set it toTruewhen the next node expects an IMAGE batch of frames. Check downstream socket compatibility before connecting the result. - Enable the widget’s directional skydome setting when using
camera_prompt. Otherwise, the node clears that prompt during execution. - Use
pose_imagewhen you want image-guided pose import. For proportion analysis in Pose Manager mode, leave automatic proportion analysis enabled. Import also requires the active browser widget to synchronize its result. - For exact capture sizing, enable the widget’s Capture Image Size setting and supply a valid
pose_imagewithin its dimension limits. The returned browser captures must match the requested width and height. - Keep output dimensions and animation frame counts practical. Larger captures, grids, and frame sequences consume more resources and can exceed the node’s size or capture limits.
VNCCS Pose Studio Common Errors and Solutions:
Pose Studio did not receive images rendered by the browser widget. Backend 3D rendering has been removed; keep the ComfyUI browser tab open and ensure WebGL is available while the workflow executes.
- Explanation: The node has no usable browser captures, and it cannot create replacement renders on the server.
- Solution: Keep the executing workflow open in the browser, confirm that WebGL works, and run it again. Check that the Pose Studio editor has finished loading and can display the scene.
Pose Studio frontend sync failed: Live capture timed out; old captures cannot be used. Keep the executing workflow open in the browser and hard-reload ComfyUI before retrying.
- Explanation: The browser did not return a fresh capture before the synchronization timeout. The node deliberately does not substitute an old capture.
- Solution: Keep the workflow tab open, hard-reload ComfyUI, wait for the editor to become ready, and retry. For a long animation, reduce the frame count if browser capture remains slow.
Pose Studio frontend sync failed: <sync error>
- Explanation: The browser widget reported a capture or synchronization failure; the text after the colon gives the specific reason.
- Solution: Read that reason first, then check that the editor scene is ready and the browser tab remains open during execution. Retry after correcting the reported issue.
Pose image SAM import failed: <reason>
- Explanation: Image-guided pose analysis, model loading, or the transfer of the analyzed result to the browser failed. The final part of the message identifies the immediate cause.
- Solution: Check the supplied image, leave the browser widget open, and retry. If the reason identifies a missing model or dependency, address that installation issue before using image-guided import.
Pose image analysis requires an active browser widget.
- Explanation: A
pose_imagewas supplied, but the node could not identify an active widget to receive the analyzed pose. - Solution: Open the workflow containing this Pose Studio node in the browser and execute it there. Disconnect
pose_imagewhen you do not need image-guided analysis.
Pose image frontend sync timed out.
- Explanation: The image analysis did not synchronize back from the browser within its allotted time.
- Solution: Keep the workflow tab open and wait for the editor to finish loading before retrying. Check the browser and ComfyUI logs if it repeatedly stalls.
SAM import returned empty pose data.
- Explanation: The image-analysis path did not produce usable pose information.
- Solution: Try a clear image in which the body is visible. Check any accompanying model or processing errors in the ComfyUI log.
Capture Image Size requires a valid IMAGE input
- Explanation: Capture Image Size was enabled, but
pose_imagedid not have the dimensions expected of a ComfyUI IMAGE value. - Solution: Connect a valid IMAGE to
pose_image, or disable Capture Image Size if you do not need the reference image to determine capture dimensions.
Capture Image Size supports images up to 4096 x 4096 pixels
- Explanation: The supplied image exceeds a width, height, or total-pixel limit for Capture Image Size.
- Solution: Resize the image before connecting it. Keep each dimension at or below 4096 pixels; a smaller image may still be necessary to meet the separate total-pixel limit.
Capture Image Size expected browser captures at <width> x <height>, received <width> x <height>
- Explanation: The browser returned a capture whose dimensions differ from the size requested by Capture Image Size.
- Solution: Let the browser editor finish updating to the requested size, then capture again. Check the widget’s export dimensions and the connected reference image.
Pose Studio Animation output requires a ComfyUI version with the official VIDEO datatype (CreateVideo).
- Explanation: Animation mode attempted to produce VIDEO, but the installed ComfyUI does not provide the required official video support.
- Solution: Use a ComfyUI version that supports the official VIDEO datatype, or enable
animation_image_batchand connect the resulting IMAGE batch to a compatible downstream node.
view_width must be a positive integer / view_height must be a positive integer / grid_columns must be a positive integer
- Explanation: One of these widget export settings is missing a valid whole number greater than zero, or contains a non-finite value.
- Solution: Correct the setting in the Pose Studio widget. If editing
pose_datamanually, use positive whole numbers, such as512for width or height and2for grid columns.
Pose Studio output dimensions are too large
- Explanation: The requested export width multiplied by height exceeds the node’s total-pixel limit.
- Solution: Reduce the output width or height in the widget and retry. The exact total-pixel limit is not specified in the provided source.
Pose Studio grid dimensions are too large
- Explanation: Combining the captures into one grid would exceed the node’s total-pixel limit.
- Solution: Reduce capture dimensions, use fewer captures, or select LIST output instead of a grid.
captured_images must be a list / captured_images entries must be non-empty strings
- Explanation: The captured-image data has an unexpected structure, so the node cannot decode it.
- Solution: Generate a fresh capture through the Pose Studio browser widget rather than manually altering the capture data in
pose_data.
captured_images limit is <limit> / captured_images payload is too large / captured image is too large / captured image dimensions are too large
- Explanation: The capture exceeds a count, encoded-data, individual-image, or image-dimension safeguard. The provided context does not state the numeric values of these limits.
- Solution: Reduce the number of poses or animation frames and lower the capture resolution, then render again.
Captured sequence is too large. Reduce resolution or frame count.
- Explanation: The combined pixel count of an animation or other capture sequence exceeds the permitted limit.
- Solution: Reduce the frame count, lower the capture resolution, or both, then retry.
