How to make a podcast audiogram with captions

The short answer

Upload the episode as MP3, WAV, M4A or AAC. Subtitle AI transcribes it and opens it as an audiogram with a waveform and captions. In the Style tab, pick a template and a video format such as 9:16. In the Highlights tab, click AI Generate to find the best moments. Tick the ones you want, click Export, open the Video tab and click Clips.

By Subtitle AI · Updated

Time
About 15 minutes, plus processing and render time
Before you start
  • A Subtitle AI account. You can start free.
  • The episode as an MP3, WAV, M4A or AAC file. The free plan takes up to 30 minutes and 200 MB per file.
  • Your cover art as an image file, if you want it in the video.
  1. Step 1 of 6

    Upload the episode

    Open the projects page and click Upload.

    Drop the audio file in the window.

    Under Spoken language, pick the language of the episode. If you are not sure, leave it on Autodetect.

    Click Upload.

    An audio-only MP3, WAV, M4A or AAC file opens with the Studio template, so the project already has a picture. Other audio formats may open without it; convert them to MP3 first.

    Drop the file, pick the language, upload.

  2. Step 2 of 6

    Check the transcript

    Open the project when it shows Ready.

    Read through the subtitles and correct names, jargon and misheard words.

    The captions in the audiogram come from these subtitles, so a fix here shows up in every clip and every export.

  3. Step 3 of 6

    Pick a template and a format

    Click the Style tab.

    Click a template card, such as Studio, Vertical Clip or Spotlight. The cards preview your own episode; hover one to see a few seconds play, muted.

    To change the shape, open the Video format menu and choose 9:16 for Reels, TikTok and Shorts, 1:1 for feeds, 16:9 for YouTube, or 3:4. For 4:5, pick the Editorial template.

    The preview shows exactly what will be rendered.

  4. Step 4 of 6

    Add your brand

    In the Style tab, find the Brand section.

    Under Logo or cover art, add your image.

    Set the five colours, the Heading font and the Caption font.

    Choose a Visualiser style: Bars, Wave, Line, Dots or Ring.

    Only what you change is saved, and it carries over when you switch to another of the new templates. Click Reset to go back to the template's own colours and fonts. Without an image, the template shows placeholder art.

  5. Step 5 of 6

    Let AI find the moments

    Click the Highlights tab.

    Click AI Generate, read the prompt, and click Generate Highlights.

    To add a moment yourself, click New Highlight and set its start and end.

    Each highlight gets a title and a description you can paste into your post. Generating highlights uses credits.

    Click AI Generate, then review the list.

  6. Step 6 of 6

    Render the clips

    In the Highlights tab, tick the highlights you want.

    Click Export and open the Video tab.

    Click Clips to render each ticked highlight, or Export full video for the whole episode.

    Renders run in the background; download them from the Renders page. The export window shows what a render costs before it starts.

    Open the Video tab, then click Clips. Only ticked highlights are rendered.

Which template to pick

Studio (16:9) is the default: dark, with cover art, a bold title, live bars, word-by-word captions and a progress bar. Night Wave (16:9) is deep navy with a glowing wave. Broadcast (16:9) is a light newsroom look with a calm line.

For phones: Vertical Clip (9:16) has big word-by-word captions for Reels, Shorts and TikTok. Spotlight (1:1) puts round cover art inside a ring that pulses with the voice. Editorial (4:5) suits Instagram and LinkedIn feeds.

The older Audiogram, Gradient Audiogram, Modern Podcast, Minimalist Audio and Clear Parallax templates are still in the list.

Each template opens in its own format. You can change the format of any of them in the Video format menu.

Good to know before you render

The visualiser reacts to the episode audio and is evened out per file, so a quiet recording still moves. If it moves too little or too much, change the Sensitivity in the waveform settings.

Podcast videos animate in during the first half second, so the preview at the very start of the timeline looks mostly empty. Play it to see the layout.

There are no automatic speaker names on screen. Add a text layer if you want one.

Free accounts get a small subtitle.ai mark in the corner of rendered videos until the first purchase. Subtitle and transcript files never carry it.

Podcast audiograms in Subtitle AI

Common questions

What is an audiogram?

A video made from audio: a still image, a waveform that moves with the voice, and captions. Podcasters use them to share episode clips on social platforms, which play video, not audio.

Can I make an audiogram without a video recording?

Yes. Upload the audio file. Subtitle AI supplies the picture with a podcast template.

What format should an audiogram for Instagram be?

Use 9:16 (1080×1920) for Reels and Stories, and 1:1 (1080×1080) for the feed.

How long should an audiogram clip be?

Short enough to stand on its own: one idea or one answer. AI highlights pick moments that make sense without the rest of the episode. You can shorten or lengthen any of them before you render.

Can I add captions in another language?

You can translate the subtitles into 60+ languages in the editor and export them as SRT or WebVTT files.