Learn the official Descript Studio Sound workflow, then use careful listening and restrained intensity adjustments to keep spoken-word audio usable.
What Descript Studio Sound changes
Descript Studio Sound is an AI-powered audio effect that enhances recorded speech by reducing background noise, echo, and other distractions. Descript recommends enabling it when a recording sounds noisy, echoey, or similar to audio captured with a laptop microphone.
The effect requires an internet connection. On current plans, it uses AI Credits; Legacy and Sunset plans track usage differently. Studio Sound also works at the file level, so enabling it affects every instance of that file across the project.
That scope should shape your editing decision. As an editorial suggestion, identify every use of the source file before switching on the effect. Consider Studio Sound an adjustable treatment rather than an automatic finishing step: the most useful setting is the one you prefer after listening to the processed voice in context.

Apply Studio Sound to a script or sequence
For a script, use the workflow documented by Descript:
- Select the script layer.
- Open Properties in the right sidebar.
- Under Audio Effects, enable Studio Sound.
For a multitrack sequence, open the sequence in the Sequence Editor, select the track you want to enhance, and enable Studio Sound in Properties. Descript also allows the effect to be enabled while recording in the main editor or Screen Recorder.
As an editorial suggestion, play a short portion before and after enabling the effect so you have a clear comparison. Then review several parts of the recording rather than judging the treatment from one sentence. Include a clear passage, a noisy passage, and a pause if those examples exist in your material.

Adjust intensity by listening to the voice
The Intensity slider controls how much Studio Sound enhancement is applied to the file. Descript’s troubleshooting instructions specifically recommend reducing Intensity or disabling Studio Sound if the processed audio has a flat or silent waveform.
A cautious starting point is an editorial suggestion, not a setting guaranteed to suit every recording. Begin at an intensity that lets you compare the original and processed voice, then raise or lower it according to what you hear. Give speech intelligibility and a result you prefer priority over maximizing the amount of processing.
Listen through complete phrases rather than checking only isolated words. As suggested quality-control checks, compare consonants, sentence endings, breaths, pauses, and transitions between speakers. If you prefer the untreated version or find the processed speech harder to follow, reduce the intensity or turn the effect off.
Treat only part of a recording
Selecting a visible passage does not change Studio Sound’s file-level scope. To process only one portion, follow Descript’s section-specific method:
- Duplicate the desired audio section into a new composition.
- Use Flatten tracks to turn that section into a new project file.
- Apply Studio Sound to the new file.
- Replace the corresponding section in the original project.
This workflow creates a separate file for the selected passage, allowing it to receive different treatment without applying that change to every instance of the original file.
I suggest using this method when a limited passage needs enhancement but the rest of the source already sounds acceptable to you. After replacing the section, listen across its entry and exit points. That boundary check is an editorial recommendation for evaluating continuity; it does not guarantee that a transition will be inaudible.
Troubleshoot processing and suppressed speech
If Studio Sound produces a flat or silent waveform, Descript says very loud background noise may prevent it from separating speech from noise. Its documented remedy is to reduce Intensity in Properties or disable Studio Sound for that file.
If enhancement fails, wait a few minutes and apply the effect again. For a persistent failure, check Descript’s status page; if systems are operational, replace the file and retry. Files lasting at least six hours or measuring several gigabytes may process slowly, and Descript recommends splitting long recordings into shorter segments and processing them individually.
As an editorial suggestion, preserve an easy way to compare the untreated source while troubleshooting. Make one change at a time—such as lowering intensity before replacing a file—so you can judge which action produced the version you prefer.
Conclusion
My practical recommendation is to use Descript Studio Sound selectively. Confirm which source file will be affected, enable the effect through Properties, and adjust Intensity while comparing representative passages. When only one section needs different treatment, create the separate file using Descript’s duplicate-and-flatten workflow.
Do not keep stronger processing merely because it changes the background more noticeably. That is an editorial recommendation based on listening priorities, not a promised result. If speech is suppressed or the waveform becomes silent, follow Descript’s guidance: lower Intensity or disable the effect. The practical goal is the version of the spoken-word recording that you find clearest and most usable, even when that means applying less enhancement.
Sources
Disclosure
This article was prepared with AI assistance.
Improve this guide
Report a problem or suggest the next question
Submissions enter a private editorial queue. They are reviewed and deduplicated before any public update or new page is created.
