Target Audio Description POC

Note: AI output should be reviewed by a team member before high-stakes use.

Generate audio descriptions with AI

Upload a video or add a public YouTube link. TadPOC will create reviewable audio description text, captions, narration audio, and downloadable assets.

Video source

Choose a video source.
How will you provide the video?

Pick one. Only the option you choose is used.

Choose a video file from this device.

A link to a video on YouTube.

A public direct link to a video file, like https://example.com/clip.mp4. This is not a YouTube link.

Choose a video file or drag and drop one here.

No video file selected.

Source captions (optional)

Add captions for the source video

Optional. Upload the source video's own captions. These are the captions for the original video, not the audio description. Accepted formats: .vtt, .srt, or a timestamped .txt file. Maximum size 1 MB.

No captions file selected.

Video categories

Choose any categories that fit the video (optional).

Generate reviewable audio description text, captions, narration audio, and described video downloads from one video.

Downloads to generate

Choose downloadable assets

The audio description transcript always appears on the results page. Select any additional files you want TadPOC to create.

Transcripts

Visual description for blind and low-vision guests and team members.

Dialogue, meaningful sounds, and on screen text from the source video.

A shorter readable transcript that removes low value detail.

Source transcript plus visual description.

Captions

Audio and video files

Creates a WAV file that combines the source audio with the audio description narration. This does not render a video.

Mixed audio WAV is available for uploaded files only. It is not available for YouTube links.

Video with audio description is available for uploaded files only. It is not available for YouTube links.

Burned-in captions require an uploaded file because TadPOC does not download YouTube videos. You review and edit the captions before this video renders.

Audio description timing files

These are audio description cue and timing files, not source captions. They work for uploaded files, direct video URLs, and YouTube.

Structured timing plan and QA warnings for downstream tooling.

Audio description narration cues, timed on the source timeline. Different from source captions.

Audio description narration cues in SRT. Different from source captions.

A plain text TadPOC timing handoff with insertion points, estimated durations, and pause needs.

Extended audio description (pauses the video)

Extended AD pauses the video so descriptions can play without covering dialogue. The exported video will be longer than the source video. These downloads require Extended AD planning (set Audio description timing to Extended AD planning above).

An MP4 that pauses at planned points so the description can play. Uploaded files and direct video URLs only.

The audio description on the Extended output timeline, with pause notes.

Audio description cues on the Extended output timeline. Not source captions.

A plain text editorial handoff for the Extended output timeline.

Analysis settings

Analysis model

Use Best quality for the most complete audio description. Choose Gemini Flash or Flash-Lite when you want to compare speed, cost, and output quality.

Highest quality baseline for audio description.

Audio description timing

Choose how the audio description should relate to the video length.

Keeps the video length the same and tries to fit descriptions into natural gaps.

Marks where the video would need to pause so important visual details can be described. This phase marks where pauses are needed but does not create a paused video yet.

Narration settings

Choose a voice category

Choose a voice category, then select a specific voice below.

No voice preference saved yet.

Review and export options

Use this when you want to check timing, wording, brand names, or on-screen text before audio is created. Analysis runs first and stops before narration so you can edit the description, then narration is generated.

Create Standard, Concise, Detailed, Marketing, and Training versions from one analysis. Switch between styles on the results page without reanalyzing the video.

Start analysis