Types

Types

MediaProbe

interface MediaProbe {
  durationMs: number
  hasVideo: boolean
  hasAudio: boolean
  width?: number
  height?: number
  mimeType?: string
}

The result of probeMedia, duration, which tracks exist, optional dimensions, and MIME type.

MediaOrigin

type MediaOrigin =
  | { kind: 'blob'; blob: Blob; name: string }

Where the bytes of a media file come from. Today the only member is a blob with a display name.

MediaProberId

type MediaProberId = 'mediabunny'

The closed set of prober ids. Today the only member is mediabunny.

MediaProber

interface MediaProber {
  id: MediaProberId
  probe(origin: MediaOrigin, signal?: AbortSignal): Promise<MediaProbe>
}

An engine that turns a MediaOrigin into a MediaProbe. createAsset accepts one so another engine can replace mediabunnyProber without changing its callers.

MediaProbeErrorCode

type MediaProbeErrorCode = 'unreadable' | 'no-tracks' | 'no-duration'

Why a probe was rejected. unreadable when no demuxer or the browser can parse the file, no-tracks when it parses but has no audio or video track, no-duration when it has tracks but no measurable duration.

MediaSourceLike

type MediaSourceLike = Blob | string

A media source given as a blob or a URL string.

ThumbnailOptions

interface ThumbnailOptions {
  width?: number
  timeMs?: number
}

Options for thumbnails, the output width and the source time to capture.

ExtractAudioOptions

interface ExtractAudioOptions {
  sampleRate?: number
  numberOfChannels?: number
  onProgress?: (progress: number) => void
  signal?: AbortSignal
}

Options for extractAudioToWav, the output sample rate and channel count, a progress callback, and a signal that cancels the conversion and rejects with the signal's reason.

ActiveMediaItem

interface ActiveMediaItem {
  assetId: AssetId
  sourceTimeMs: number
  rate: number
  reversed?: boolean
}

One video feed that is showing at a moment, with its source time, playback rate, and reversal.

PreviewSyncOptions

interface PreviewSyncOptions {
  isPlaying: boolean
  playbackRate: number
}

The playback state passed to PreviewMediaPool.sync, the playing flag and the transport rate.

ContainerFormatId

type ContainerFormatId = keyof typeof containerFormats

A key of the containerFormats table, mp4, webm, or mkv.

ExportFontFaceInit

interface ExportFontFaceInit {
  family: string
  weight?: string
  style?: string
  unicodeRange?: string
  source: ArrayBuffer | string
}

A font face the export worker registers before rendering, given as binary data or a URL.

ExportProgress

interface ExportProgress {
  progress: number
  phase: 'audio' | 'video' | 'finalize'
}

Export progress as a 0 to 1 fraction within the audio, video, or finalize phase.

ExportProjectOptions

interface ExportProjectOptions {
  format?: ContainerFormatId
  videoBitrate?: number | Quality
  fonts?: ExportFontFaceInit[]
  audioSources?: ReadonlyMap<ElementId, string>
  onProgress?: (progress: ExportProgress) => void
  signal?: AbortSignal
}

Options for exportProject, the container format, video bitrate or quality, fonts, per-clip replacement audio URLs, progress callback, and abort signal.

ExportResult

interface ExportResult {
  blob: Blob
  extension: string
}

The exported file as a blob together with its file extension.

ContainerFormat

interface ContainerFormat {
  label: string
  extension: string
  mimeType: string
  createOutputFormat: () => Promise<OutputFormat>
}

A container format's label, file extension, MIME type, and factory for the Mediabunny output format that muxes it.

ContainerFormatEntry

interface ContainerFormatEntry extends ContainerFormat {
  id: ContainerFormatId
}

A ContainerFormat together with its id, as returned by listContainerFormats.

Filmstrip

interface Filmstrip {
  canvas: HTMLCanvasElement | OffscreenCanvas
  frameWidth: number
  frameHeight: number
  frameCount: number
  timestampsMs: number[]
}

The filmstrip canvas with its frame size, frame count, and the source timestamp of each frame.

FilmstripOptions

interface FilmstripOptions {
  frameCount: number
  frameWidth?: number
  startMs?: number
  endMs?: number
}

Options for getFilmstrip, the frame count, frame width, and source range.

VideoPreviewCapability

interface VideoPreviewCapability {
  name?: string
  mimeType?: string
  nativePreview?: boolean
}

The asset fields that decide native preview eligibility, name, MIME type, and an explicit nativePreview override.

AudioPeaks

interface AudioPeaks {
  peaks: Float32Array
  durationMs: number
}

Waveform peak buckets and the duration they cover.

AudioPeaksOptions

interface AudioPeaksOptions {
  buckets?: number
  startMs?: number
  endMs?: number
}

Options for extractAudioPeaks, the bucket count and the source range.

AudioActivity

interface AudioActivity {
  durationMs: number
  soundWindows: AudioActivityWindow[]
  silenceWindows: AudioActivityWindow[]
  summary: AudioActivitySummary
  waveform?: number[]
}

Sound and silence windows over a duration, a summary, and an optional waveform.

AudioActivityOptions

interface AudioActivityOptions {
  startMs?: number
  endMs?: number
  frameMs?: number
  threshold?: number
  minSoundMs?: number
  minSilenceMs?: number
  paddingMs?: number
  waveformBuckets?: number
}

Options for activity analysis, the source range, frame size, threshold, minimum run lengths, padding, and waveform bucket count.

AudioActivitySummary

interface AudioActivitySummary {
  soundMs: number
  silenceMs: number
  soundFraction: number
  silenceFraction: number
  peakRms: number
  peakAmplitude: number
}

Totals and fractions of sound and silence plus the peak RMS and peak amplitude.

AudioActivityWindow

interface AudioActivityWindow {
  startMs: number
  endMs: number
  durationMs: number
  rms: number
  peakRms: number
  peakAmplitude: number
}

One sound or silence window with its bounds, duration, mean RMS, peak RMS, and peak amplitude.

ConstantSpeed

interface ConstantSpeed {
  rate: number
  sourceStartOffsetMs: number
  sourceSpanMs: number
}

A constant playback rate with the source offset where it starts and the source span it consumes.

StereoData

interface StereoData {
  left: Float32Array
  right: Float32Array
  sampleRate: number
}

Left and right PCM channels with their sample rate.

AudioSyncOptions

interface AudioSyncOptions {
  windowS?: number
  maxLagS?: number
  rateHz?: number
  signal?: AbortSignal
}

Options for audio sync, the analysis window in seconds, the maximum lag, the envelope rate, and an abort signal.

SyncResult

interface SyncResult {
  offsetMs: number
  confidence: number
}

A sync offset in milliseconds and the confidence of the match.

ProjectStill

The PNG blob from renderProjectStill, with its pixel width, pixel height, and the ids of the elements drawn.

RenderProjectStillOptions

Optional arguments for renderProjectStill. soloElementId draws only that element, including a multicam composite. width caps the PNG width and keeps the project aspect ratio.

SceneChangeOptions

Optional arguments for findSceneChanges. elementId picks a video or multicam clip and defaults to the first multicam, then the first video clip. source picks a multicam source and defaults to screen, then the first video source. startMs and endMs bound the scan in timeline time and default to the clip span. stepMs defaults to 500. sensitivity runs from 0 to 1 and defaults to 0.5, where a change needs 30% of the picture to change. maxChanges defaults to 40 and keeps the largest changes.

SceneChangeReport

The result of findSceneChanges. It names the element, the source, and the asset it read, the range, stepMs, sensitivity, the threshold fraction it used, and the number of samples. changes lists { timeMs, changed } in timeline order. segments lists the { startMs, endMs } stretches between changes. truncated is true when maxChanges dropped smaller changes.

ContactSheetOptions

Optional arguments for renderContactSheet. It takes the same elementId, source, startMs, and endMs as SceneChangeOptions, plus timesMs, count (default 12), columns (default a square grid), and thumbWidth in pixels (default 320).

ContactSheet

The PNG blob from renderContactSheet, with its pixel width and height, the column count, the element and source it read, and tiles, the timeline time of each tile in grid order.

FaceSample

FaceSample is { sourceMs, box }, one reading of createLocalFaceDetector at an integer source media time. box is the largest face as { x, y, w, h } fractions of the frame, or null when the frame has none. It fits FaceSample from @mcut/editor, so detect output goes straight to planCenterPerson.