Video Post-Production & Subtitle Glossary
A comprehensive, technically rigorous encyclopedia covering speech recognition, video codecs, keyframe geometry, canvas compositing, and audio standards.
Automatic Speech Recognition (ASR)
subtitlesThe capability of an artificial intelligence model to process spoken audio acoustic waves and convert them into machine-readable digital text.
Modern ASR uses deep neural transformer architectures (e.g., OpenAI Whisper) operating on log-mel spectrogram features with beam search decoding and subword tokenization.
Powers Editoria’s zero-cost speech transcription engine, generating timestamps with sub-50ms millisecond precision.
Group of Pictures (GOP) & Keyframes (I-Frames)
videoA GOP defines the structural interval between successive Intra-coded frames (I-frames/Keyframes) in an MPEG/H.264 video compression bitstream.
I-frames encode a complete visual image without reference to past or future frames. P-frames (predicted) and B-frames (bi-directional) only encode difference deltas. Long GOP structures save bandwidth but cause seeking lag.
Editoria enforces closed GOP intervals with fixed keyframe distances (every 30 or 60 frames), completely eliminating macroblock tearing and pixel pixelation when seeking in video players.
Burned-In Subtitles (Open Captions)
subtitlesSubtitles that are permanently rendered and rasterized directly onto the pixel canvas of video frames, rather than stored as an external selectable track.
Open captions cannot be disabled by the user. They bypass device font limitations, ensuring 100% typography, stroke, drop shadow, and animation consistency across Instagram, YouTube, and TikTok.
Editoria composites kinetic typography and boxes frame-by-frame, guaranteeing that what you see in the preview canvas is exactly what exports into the final MP4.
Dual-Channel Reinforcement (McGurk Effect in Captions)
algorithmsThe perceptual phenomenon where visual text stimuli and auditory speech cues fuse together in the human brain to enhance linguistic comprehension.
Presenting dynamic synchronized text simultaneously with spoken voice accelerates recognition by over 30%, compensating for noisy mobile viewing environments where sound may be muted.
Our 30+ Alex Hormozi box and kinetic bounce animation presets leverage this perceptual phenomenon to boost average view percentage.
LUFS (Loudness Units Full Scale)
audioThe international standard (ITU-R BS.1770) for measuring audio perceived loudness as experienced by human ears over time.
Streaming algorithms on YouTube Shorts and Instagram normalize audio to approximately -14 LUFS. Over-compressed or overly loud audio above -14 LUFS is automatically attenuated by the platform.
When audio tracks and viral sound effects (SFX) are mixed in Editoria, audio levels remain balanced to prevent algorithmic clipping.
Audio Ducking
audioAn audio post-production technique that temporarily compresses and lowers the volume of secondary background music whenever a primary vocal frequency is detected.
Implemented using side-chain compression with defined attack, release, and hold times (typically 50ms attack, 250ms release) to ensure vocal clarity without unnatural pumping.
Built into Editoria’s B-roll and sound effect library to ensure spoken vocals always remain crisp and intelligible.
Canvas Compositing (Hardware Acceleration)
videoRendering multi-layered 2D graphic elements (text, boxes, SVG stickers, video feeds) using the browser's GPU-accelerated HTML5 Canvas 2D or WebGL pipeline.
Utilizes requestAnimationFrame loops and offscreen canvas buffers to achieve 60FPS fluid playback without burdening the CPU or causing frame stutter.
Editoria executes all preview and Turbo GPU rendering via browser hardware acceleration without uploading your raw video to cloud buckets.
IndexedDB Local Storage
algorithmsA transactional, low-level client-side key-value database embedded in modern web browsers for storing structured binary and font data.
Unlike localStorage (which is limited to 5MB and synchronous strings), IndexedDB handles hundreds of megabytes of binary font files (TTF/OTF) and serialized project state asynchronously.
Your uploaded custom brand fonts and unsaved project timelines are cached securely in IndexedDB so you never lose work on accidental refresh.
Tanglish / Hinglish Speech Code-Switching
subtitlesThe linguistic practice of alternating between two languages (e.g., Tamil and English, or Hindi and English) within a single conversation or sentence.
Acoustic models must be conditioned on bilingual phoneme dictionaries to avoid misidentifying vernacular loan words as spelling errors.
Editoria features India’s first hybrid multi-language acoustic pipeline designed specifically to translate colloquial Tanglish speech into polished viral English captions.
Fast-Start MP4 (moov Atom Positioning)
videoPlacing the MP4 file index and metadata box (the "moov atom") at the absolute beginning of the file rather than at the end.
When the moov atom is located at the start, web browsers and mobile feeds can begin streaming and playing video immediately without downloading the entire file first.
All videos rendered by Editoria’s MP4 exporter are automatically muxed with the fast-start moov flag enabled.
Safe Zone Coordinates (9:16 Vertical Video)
videoThe central display rectangle within a 1080x1920 vertical video frame that remains unobstructed by platform overlays (usernames, buttons, sidebars).
Instagram Reels and TikTok overlay UI elements on the bottom 250px and right 150px. Essential graphics and subtitles must reside within the safe horizontal and vertical margins.
Editoria provides interactive Safe Zone visual guide overlays to ensure your subtitles never get hidden behind social platform icons.
SubRip Text (.SRT) Format
subtitlesThe most widely supported subtitle interchange format, composed of numeric sequential counters, timecode intervals, and plaintext lines.
Timestamps follow the standard format 00:00:00,000 --> 00:00:00,000 with millisecond precision separated by a comma.
Editoria allows both importing pre-existing .SRT files and generating automatic timed subtitles for immediate visual styling.
Experience These Technical Capabilities Firsthand
Test our Whisper AI transcription, 60FPS closed GOP keyframe rendering, and kinetic subtitle styles for free.