By Descript
Transform your audio podcast into social-ready audiogram clips with waveforms, captions, and illustrations.
# Create audiograms from audio podcast recording
# SET-UP
Ask these questions one at a time as succinctly as possible. Wait for an answer before asking another question.
## 1. Set-up: Aspect ratio and clip preferences
- I may want clips in a different aspect ratio, optimized for different places: Landscape 16:9, Portrait 9:16, or square 1:1. If I say I want portrait clips, only use portrait layouts. This is important.
- I may have a specific number of clips in mind, or special moments I want included. Please ask.
## 2. Set-up: Brand styles
In some cases, I want to represent my brand colors (by name, by hex color, or upload a screenshot to extract colors) in the video. I can upload a logo or a brand color palette.
# WORKFLOW
Do this all automatically, one after another. Don't stop to ask questions along the way.
- Create audiograms in the specified aspect ratio
- Switch to the "Luxembourg celery - light" layout pack. Be sure to set the project to the preferred aspect ratio.
- Never use the camera layout or any camera placeholders - this is audio only so the camera layout won't make sense
- Apply the "Audiogram" layout to every scene EXCEPT for "interesting clause scenes"
- If I uploaded a logo, add it to the upper right corner of the video.
- Add interesting Clause Scenes: About 2-3 times per clip, find the interesting / non-filler clause in the sentence. Create a media scene based on that clause and generate an image of that clause in a flat illustration style using the brand colors. It should not include text, just an illustration of the interesting clause.
- Add a subtle Ken Burns effect to still photos for movement.
- Complete all WORKFLOW steps for all audiograms
Remember: If I specify brand colors, update all relevant color properties—including active word background color for captions, element layer fills, and canvas background color—to reflect my brand. Don't change the active word fill color.
- Be sure to update the waveform border color and captions active word background color to match the primary color of my brand.
Complete all WORKFLOW steps for each audiogram
---
After completing the workflow, offer a few suggestions for additional ways to repurpose the content.Click Use template to open Descript with a preloaded AI video template. Upload your video or audio file, paste a link, or start recording — Descript’s AI co-editor automatically generates a ready-to-edit project in seconds.
Yes! Every project created with a Descript AI template is fully editable. You can rewrite your script, trim audio or video clips, add captions, or change layouts and aspect ratios for YouTube, TikTok, or Reels — all inside Descript’s text-based video editor.
Descript’s free AI templates help you quickly create videos, podcasts, and social media clips. Whether you’re a marketer, educator, or content creator, you can transform raw recordings into professional, share-ready media in minutes using Descript’s all-in-one AI video editor.
Import your audio, pick your aspect ratio, then let Descript's AI co-editor auto-cut clips and add waveform captions. It also inserts 2–3 illustrated images for emphasis per clip and exports social-ready files.
Use Descript’s AI to generate audiograms for free: set your aspect ratio, auto-apply waveform captions, and export clips. Underlord also adds illustrations for key moments and subtle Ken Burns motion so your posts are ready for IG Reels, TikTok, or YouTube.
Yes. Upload a logo and brand colors; Descript updates canvas fills, waveform borders, and caption active-word backgrounds to match your primary color. Your logo is placed prominently in the audiogram.
Choose portrait 9:16 for Reels and TikTok, square 1:1 for Instagram feed, and landscape 16:9 for YouTube. Set your preferred ratio first; Descript keep visuals optimized for audio-only posts.
Descript's AI video agent detects 2–3 key moments per clip and generates on-brand illustrations with a subtle Ken Burns effect. These appear as media scenes while the rest use the audiogram layout with waveform and word-by-word captions.