Timing comes from the transcript
Whisper records when each word starts and ends. HookWise uses those times to fill or light up the word being spoken, and splits the text into lines by the words-per-line number you set.
A plain .srt or .vtt file only has a time for each line, so HookWise spreads the words across the line by their length. YouTube auto-captions carry their own word timing, and HookWise keeps it. When you fix a typo in the transcript and the word count stays the same, every word keeps its own timestamp. If the word count changes, the new words are spread across that line.
Styling captions for a phone screen
Start from one of the 10 presets, then adjust font, color, outline, shadow and words per line. Karaoke fills each word gradually as it is spoken. Active word highlights only the current word, and active word + pop adds a pop animation to it. With a fast talker, fewer words per line are easier to follow on a phone.
Type keywords separated by commas and they are highlighted every time they appear. Save the highlight for numbers, names and the one claim the clip rests on. CAPS works for short punchy lines, and the 5 text animations, such as pop word by word, give the hook line extra weight.
What it needs and where it can slip
Captions run offline on every plan, including Starter. Their accuracy follows the transcript. Strong accents, music under the voice or two people talking over each other can produce wrong words, so read the transcript and fix names before you export. The Small and Medium models usually make fewer mistakes than Base and take longer.
AI keyword suggestions are the one part that goes online. They need the Pro or Studio plan and your Gemini key, and Gemini proposes up to 6 single words from the text. Captions are burned into the video, and the .srt file written next to it holds the same text.
