The Master Guide / 17 Glossary
T to Z
Terms from "tag" to "XML".
Terms from "tag" to "XML".
Tag. A reference's handle in a prompt, @image1 to @imageN in media order. Video with Seedance → Attaching references
Take. One generated result of a shot, judged against the treatment, anchors and plan, never a storyboard. Video with Seedance → Take selection and review
Take log (your take log). Your record of every attempt, kept or not, written before the next one runs. Planning, budgets and takes → Your take log
Templates A to F. The fill-in Seedance prompts: A, B and D in Video with Seedance → Templates A, B and D; C, E and F in Speaking shots and lip sync → Templates C, E and F.
Tempo map. The locked guide track's measured BPM, bar grid, section starts and hit points, written into your shot plan in frames. Music → The music-first route
Texture once. One grain per film, applied once after the look and every picture check. The grade and texture → Texture once
Three layers. Every sound comes from the shot's own sound, real recordings or designed effects. Sound design and mix → Three layers and the order of the sound work
Three texts per line. The canonical line, the engine text and, for the Franco quote only, the Franco line. Voice and Arabic dialects → Three texts per line
Timing gate. Route 1's working numbers: span within ±5%; pauses and the emphatic beat within ±3 frames. Speaking shots and lip sync → Route 1, for close-up hero lines
"Too perfect" tells. The signs a correct take still reads as generated, each with its first fix. Making it feel real → The "too perfect" tells
Treatment. The film decided in words and pictures, the third client document: the brief restated, the scenes, every department, the casting and the storyboard. Writing the treatment → The order of work
Trigger. What starts, changes or ends a camera move. Camera and lens → The trigger
True peak. See dBTP. Sound design and mix → Loudness: −23 TV, −16 online, −14 short social
UGC. User-generated content: ads styled as a creator's own phone video. Genres → UGC and selfie
Usual AI defaults. The looks generated work drifts to; each needs a written decision in the treatment. Making it feel real → The usual AI defaults to avoid
VFX (visual effects). Picture added or changed after the shot, such as an insert composited into a live plate. Video with Seedance → The six start-frame cases
VO (voiceover). A voice heard over the picture from off screen, a narrator's or a character's. Voice and Arabic dialects → Voice takes and the delivered file
Walla. Crowd murmur. Arabic murmur comes from real CC0 recordings or recorded voices, never Seedance (decided). Video with Seedance → Sound inside the shot
WCAG (Web Content Accessibility Guidelines). Here, the contrast ratios client pages meet: 4.5:1 for body text, 3:1 for large text. Client presentations → Type, colour and the phone
Wear with a history. Wear written as the result of a life; never on the client's product. Making it feel real → Skin, make-up and wear with a history
"White" Arabic. A pan-Arab read built from casting plus pan-Arab wording; no engine offers one. Voice and Arabic dialects → Cast the dialect, don't prompt it
Window (the window). Time planned in whole frames: a shot's slot, or the part of a take that plays. Planning, budgets and takes → Timing, grouping and trimming
World rules. Three to five rules every department obeys, so the one planned break lands. Writing the treatment → World rules and the planned break
XML. Here, the editable timeline file exported from Premiere (Final Cut Pro XML), with every asset and the earlier takes beneath. Editing → The editable timeline
