Descript

Read the article; extract 5 key takeaways, core insights, and audience. Produce a master summary and bullets.

Choose outputs (blog, podcast, video, social clips, slides) and specify target lengths and channels.

Draft a one-page outline aligning sections with the article’s insights; include intro, main points, conclusion, CTA.

Create per-format script templates mapped from the master outline; note tone and constraints.

Write scripts using prompts with Action, Context, Tone, Format, and Constraints to guide AI drafting.

Prepare prompts for images and video assets (thumbnails, B-roll) consistent with branding.

Combine scripts and assets into a project; arrange timelines, captions, and chapter markers.

Review for branding consistency, tone, readability, and add captions/descriptions for accessibility.

Export deliverables in required formats; organize files with clear naming; set publishing plan.

After recording ends, primary and backup files upload. In the Descript project, open the Project Files panel and navigate to Recordings → Room – [date/time]. Confirm that both the high-quality primary recordings and the browser-based backups exist. If uploads are incomplete, use the Rooms recovery link to finish uploading.

Open the Descript project for the Rooms session in your browser and ensure you can access the recording assets in the Project Files.

Once primary recordings finish uploading, verify that Descript has created a sequence in the script panel and started transcription. Check that speaker labels match participants' Room names.

Use the transcript as the source of edits: cut, trim, and rearrange blocks by selecting text in the script; the corresponding video clips update in the timeline.

If your project uses multiple speakers, enable the Automatic Multicam feature to switch between speaker angles automatically. Verify the timeline reflects the chosen angles.

Apply Studio Sound to the audio tracks to reduce noise and balance levels. Confirm audible improvements and consistent volume across speakers.

Leverage Underlord to adjust layouts, apply templates, or speed up edits. Review the resulting arrangement and ensure it aligns with your intended flow.

Double-check that speaker labels reflect the Room participant names and that the on-screen layout mirrors your desired gallery or single-speaker view.

Save the project, create a new version if needed, and export the final video in your preferred format. Verify the exported file plays correctly and contains the updated transcript.

Clarify branding requirements for captions and text overlays, including preferred fonts, colors, layout, and when to use a captions layer versus a text layer. This establishes the baseline for all subsequent steps and ensures consistency with your WordPress visuals.

In Descript's Elements panel, add Text, Subtitle, and Title layers and rename them to reflect their role (e.g., BodyText, CaptionSubtitle, HeroTitle). This naming helps keep scenes organized as you scale the project.

Select a text layer in the scene editor or Timeline. Use the toolbar to adjust font family, size, alignment, and color. Set positioning and layout across scenes and control layer duration and placement to maintain visual coherence.

Enable live text for dynamic fields: Speaker, Composition name, Marker, and Timer. Configure each live-text target to automatically populate as you edit the script, ensuring captions stay in sync with on-screen context.

Enhance readability by applying a background to text layers. Choose between a bounding box or fixed-width box, adjust border radius, background color, and enable hugging lines to wrap the background to each line.

Insert emojis (e.g., 🎉, 👍, 💡) directly into text layers. Ensure emoji styling matches the surrounding typography by adjusting size, color, and alignment in the Properties panel.

If you need timed captions, switch from a plain text layer to a captions layer. Captions are generated from your script and update with edits, providing synchronized captions for playback.

Review all scenes to confirm consistent font choices, colors, background treatments, and spacing. Apply changes globally where possible to maintain a unified look across the video.

Play back the project to verify timing and visuals. Export captions and assets or prepare WordPress-ready output. Draft a WordPress post describing the workflow and attach the video/caption assets for embedding.

Add your audio into Descript, select the track, and play the unedited clip to document the baseline loudness and noise level. Expected outcome: a reference waveform and identified noisy sections for targeted cleanup. Verification: listen to silent segments and confirm a consistent noise floor.

Scan silent or near-silent portions to understand the noise floor and note where background noise is most evident. Expected outcome: a mapped noise profile to inform cleanup levels. Verification: there is a clear baseline reference before applying denoising.

Turn on Studio Sound to reduce background noise and enhance vocal clarity. Expected outcome: noticeable reduction in background noise and crisper speech. Verification: perform an A/B compare between pre- and post-Studio Sound audio.

Adjust Studio Sound intensity to balance denoising with natural voice texture. Expected outcome: reduced artifacts while preserving vocal character. Verification: listen for any robotic sound and adjust accordingly.

Apply a gentle high-pass filter (around 80–100 Hz) to remove rumble without affecting voice fundamentals. Expected outcome: cleaner low end with preserved vocal presence. Verification: check spectral balance in bass region and confirm voice remains clear.

Make light adjustments to mid/high frequencies to tame hiss and improve intelligibility (focus on reducing harshness while preserving natural brightness). Expected outcome: clearer consonants and reduced sibilance. Verification: listen to speech plosives and overall clarity.

Apply mild compression (e.g., 2:1) to stabilize dynamics, then normalize to a consistent target loudness. Expected outcome: even levels across phrases. Verification: measure peak and LUFS consistency across the track.

Review the processed audio on headphones, laptop speakers, and mobile devices to ensure clarity and lack of artifacts across platforms. Expected outcome: consistent perceived loudness and intelligibility. Verification: side-by-side listening comparisons across devices.

Export the cleaned audio in preferred formats (e.g., WAV/MP3) and save the project version for future reuse. Expected outcome: publish-ready audio; original remains available for reference. Verification: file exports complete without errors and metadata is preserved.