ADVANCED PROCESSING
Introduction
Advanced Processing transforms existing media into new assets using the native power of the SToV Pro desktop application. Rather than modifying an asset in place, each method takes what you already have: a video, photo, stack or audio file and produces a new asset from it, ready to use elsewhere in your project.
Different methods are available depending on the type of asset selected, from simple operations like reversing or changing the speed of a clip, through to processing methods such as segmentation, head tracking and pose tracking.
This page describes the Advanced Processing interface and walks through how to use it. Use the menu on the right to navigate the sections.
Interface Overview
The asset preview pane (C) lists the assets available to process. Selecting one determines which methods (A) are available to you (and vice-versa), since the available operations depend on the type of asset selected: video, photo, stack or audio.
Once a method is chosen, its parameters appear in the control tools (A). Some methods, such as segmentation, are guided interactively directly in the preview (C), for example, clicking a subject in the preview to mask it out.
Pressing Process (F) begins an operation. A progress bar (D) tracks the run, and once complete, a new asset becomes available to download and/or use directly within your project.
Methods
Each method below covers what it does, which asset types it applies to, and its available parameters.
Reverse/Change Speed
The stage (A) shows the video the operation will be applied to, with the progress bar (B) tracking how far processing has run, alongside buttons for choosing an output folder and exporting a screenshot of the current frame. The preview timeline control (C) is used to navigate the clip: it sets the film strip's zoom level, lets you jump the play point by entering a time code directly, defines a From/To region either by typing time codes or selecting a range on the film strip below, and includes a play button for previewing.
The speed control (D) sets the playback rate applied to the output, ranging from -100x to +100x. The default value of 1 leaves the video unchanged, values between 0 and 1 slow it down, and any negative value reverses it, running backwards at up to 100 times the original speed. The film strip (E) can be double-clicked to scrub through the video, or a region can be highlighted directly on it, matching the From/To selection made in the preview timeline control. Pressing 'Process Video', next to the speed control (D), then exports only that highlighted region, making it possible to crop the output to just the part of the video you want reversed or sped up.
Video to Photo
Keyframes are used to mark the specific frames you want exported as images. The keyframe toolbar (E) lets you position keyframes along the film strip (F): clicking directly on the film strip creates a new keyframe at that point, and an existing keyframe (G) can be dragged by its circular marker to reposition it. Any number of keyframes can be set between a starting and ending point, giving precise control over exactly which frames are captured. The timeline preview controls (D) provide the same scale and play point navigation used elsewhere, though the region selection is not needed for this method since export is driven entirely by keyframe positions.
Once your keyframes are set, pressing 'Process Video' (B) exports the corresponding frames as individual images, with the icon beside it opening further export settings. Progress is shown in the progress bar (C), alongside the 'Select Output Folder' and 'Export Screenshot' buttons for managing where results are saved.
Segmentation
Markers are placed directly on the stage (A and zoom B): left-clicking marks the foreground subject you want segmented, and right-clicking marks background (optional) areas to exclude. C) Is a progress bar which shows the progression of segmentation processing. The threshold (D) and edge type (E) control how tightly the mask is applied and whether its edge is soft or hard. The output type (F) determines how the result is rendered. Overlay highlights the segmented subject in red over the original image, Chroma places it against a green background, Alpha renders the mask as a grayscale image, and Alpha RGB keeps the foreground as full colour pixels with a transparent background. This last option requires exporting to ProRes, since that is the only format here that supports transparency.
Rather than marking every frame by hand, the auto key point menu (G) can extract key points automatically using pose or face detection. Choosing a method and pressing 'Apply' highlights the detected points on the current image, or toggling Update Each Frame places points automatically frame-by-frame, allowing the segmentation to track a subject through the video without manual annotation. If markers are placed incorrectly at any point, Clear Markers (H) removes everything from the screen so you can start again, and further output settings, including the destination folder and screenshot export, are available alongside (I). Once set up, 'Segment Image' processes a single frame and 'Process Video' runs the whole clip (J), with a icon exposing further output format options where available.
The scrub toolbar (K) is used to navigate the video. It sets the film strip's zoom level, lets you drag the play point to preview any moment, and provides From/To time code fields, set either by typing directly or by selecting a range on the film strip below, to define a precise working region, alongside a play button for previewing. For subjects the automatic detection methods struggle with, the keyframe toolbar (L) allows markers to be set manually at chosen points in time. Keyframes can be placed anywhere, with segmentation points set independently at each one, giving frame accurate control over tricky tracking cases. The film strip (M) itself shows thumbnails from across the video, and double-clicking any thumbnail jumps the play point to that moment.
Head Tracking
The stage (A) shows the video with bounding boxes tracked around each detected head, updated frame by frame as the video plays. The timeline preview controls (E) are used to navigate the clip and, if needed, crop processing to a specific From/To region, with the film strip (F) below providing a visual overview of the whole video.
Pressing 'Process Video' (D) runs head tracking across the selected region, with progress shown in the progress bar (C) alongside the 'Select Output Folder' and 'Export Screenshot' buttons. Alongside the video output, tracking also exports a data file containing the tracked signal for each point, which can be brought into the main interface and used to drive graphics, letting you position or animate elements using the tracking data as a live signal. If only this data is needed, the 'Annotation Only' toggle skips exporting the annotated video and produces just the tracking data, and the icon opens further export settings.
Face Landmark Tracking
The stage (A) shows the video with facial landmark points tracked across each detected face, updated frame by frame as the video plays. The timeline preview controls (E) are used to navigate the clip and, if needed, crop processing to a specific From/To region, with the film strip (F) below providing a visual overview of the whole video.
Pressing 'Process Video' (D) runs face landmark tracking across the selected region, with progress shown in the progress bar (C) alongside the 'Select Output Folder' and 'Export Screenshot' buttons. Alongside the video output, tracking also exports a data file containing the tracked signal for each landmark point, which can be brought into the main interface and used to drive graphics, letting you position or animate elements using the tracking data as a live signal. If only this data is needed, the 'Annotation Only' toggle skips exporting the annotated video and produces just the tracking data, and the icon opens further export settings.
Pose Tracking
The stage (zoom A and B) shows the video with a pose skeleton tracked across each detected figure, updated frame by frame as the video plays. The timeline preview controls (E) are used to navigate the clip and, if needed, crop processing to a specific From/To region, with the film strip (F) below providing a visual overview of the whole video.
Pressing 'Process Video' (D) runs pose tracking across the selected region, with progress shown in the progress bar (C) alongside the 'Select Output Folder' and 'Export Screenshot' buttons. Alongside the video output, tracking also exports a data file containing the tracked signal for each pose point, which can be brought into the main interface and used to drive graphics, letting you position or animate elements using the tracking data as a live signal. If only this data is needed, the 'Annotation Only' toggle skips exporting the annotated video and produces just the tracking data, and the icon opens further export settings.
Audio / Video Sychronisation
This method requires both a video and an audio file to be selected from the asset pool, rather than a single asset as with the other processing methods. The video is shown as a film strip (F), and the audio as a spectrogram (G) below it, with the timeline preview controls (D) providing shared scale, play point, and From/To region selection across both, along with a play button for previewing.
The keyframe toolbar (E) is used to place keyframes marking points that should connect between the two. Any number of keyframes can be created, including between two existing ones, and each keyframe can be repositioned independently on the film strip and on the spectrogram. Between matched keyframes, the video is stretched or contracted to fit the audio: frames are duplicated to slow the video down, or subsampled to speed it up, so that visual events land exactly where they should against the audio. This makes it possible to synchronise a video to a beat, for instance taking a clip of someone clapping or hitting a drum and aligning it precisely to a separate audio recording of the same rhythm, by oversampling or subsampling the video frames between keyframes.
Once the keyframes are set, pressing 'Process Video' (B) exports the video and audio combined and synchronised through those keyframe points, with progress shown in the progress bar (C) alongside the 'Select Output Folder' and 'Export Screenshot' buttons, and the icon opening further export settings.