How to make an audiogram

Updated 17 August 2026

1. Choose the clip first

This is the step that decides whether the audiogram works, and it is the one most people rush. Scan your episode for a passage that stands on its own in under a minute. A strong claim, a direct answer to a question the listener already has, or a moment where someone is clearly disagreeing. Avoid anything that opens with a pronoun referring to something said five minutes earlier.

Start the cut on the first word of the sentence, not on the breath before it, and end on a landing rather than trailing into the next thought. Dead air at either end reads as sloppy in a feed where the first half-second decides whether someone stays.

2. Trim before you render

Cut the clip in whatever editor you already have — Audacity, your podcast host’s built-in trimmer, or the voice memo app on your phone. Export it as MP3 or WAV. Doing the trim first keeps the render fast and means the exported video is exactly the length you want, with no editing needed afterwards.

3. Pick a style and a shape

A waveform style suits speech: it tracks the rhythm of a voice closely, so the movement reads as connected to the words. Spectrum bars suit music better, where the interest is in frequency rather than cadence. Then choose the aspect ratio for where you are posting — square for a feed, vertical for Reels, Stories, TikTok and Shorts. Getting this wrong is the most common and most visible mistake; see the size guide for the numbers.

4. Export the video

Render an MP4 with H.264 video and AAC audio. That combination is accepted everywhere without conversion. With a browser-based tool the encoding happens locally, so a 30-second clip finishes in seconds and the file lands in your downloads folder without being uploaded anywhere.

5. Add captions at upload

Do not burn captions into the video. Instagram, TikTok, YouTube and LinkedIn all generate them automatically on upload, they are editable if the transcription slips, and platform-native captions are indexed as text — which burned-in pixels are not. It is less work and it performs better.

What actually drives the result

The animation is the least important variable. In practice the ranking is: the quality of the clip, then the captions, then the first frame, and only then the waveform style. A gripping 20 seconds with plain bars will outperform a dull minute with the most elaborate visualisation available. Spend your time on the audio.