A PRACTICAL AUDIO GUIDE
Prompt-Based Podcast Audio Editor for Short Recordings
For an interview excerpt or short podcast recording, a few careful changes can be enough. Upload a clip, describe what distracts from the conversation, and listen to the result.
Upload audio and start editingWork with a focused excerpt
The current limit is 20 minutes and 100 MB, with additional browser memory limits. For a full-length episode, export the section you want to improve first. This workspace edits one recording at a time; it is not a multitrack podcast mixer.
Ask for clarity without an overprocessed voice
A restrained request is usually a better first pass: “reduce low rumble and make the speech more even.” Compression can reduce the difference between quiet and loud phrases, while gentle EQ can improve intelligibility. Neither can separate overlapping speakers or restore a badly clipped microphone.
Treat pauses as part of the conversation
Ask to remove long silent stretches, then listen around every cut. Quiet speech, inhalations and room tone may fall near the detector’s threshold. The assistant does not know which pauses are emotional or editorially important from level data alone.
Export and check in context
Peak normalization is available, but measured LUFS mastering and podcast-platform certification are not. After downloading, listen on the device or publishing workflow you intend to use. If a distributor requires a specific loudness target, measure and master the export in a suitable tool.
A prompt to try
Make the speech clearer, gently even out the volume, and remove only the long silent pauses. Do not make the conversation feel rushed.
Listen for chopped breaths and abrupt room-tone changes. Compare quiet and loud speakers.
Common questions
Does it transcribe the episode?
No. The current editing workflow does not create a word-level transcript.
Can I upload separate guest and host tracks?
This version edits one audio file at a time. Mix separate tracks in your recording tool before uploading an excerpt.