Floop Hunter Framework is a REAPER vocal cleanup script that detects sibilance, plosives, and breaths on vocal items and writes non-destructive volume automation directly on the timeline.
The script analyzes selected items using three independent detection modules and generates precise automation envelopes. A companion JSFX plugin applies gain reduction without modifying the item, gain, or source audio.
The script combines de-essing, plosive control, and breath reduction into a single REAPER vocal cleanup workflow.
How REAPER Vocal Cleanup Works (Three Detection Modules)
Floop Hunter Framework includes three dedicated detection modules:
Ess Hunter (de-esser module) analyzes high-frequency content to detect sibilant events such as “S” and “Sh”.
Plosive Hunter (plosive detection module) targets low-frequency transient bursts generated by “P” and “B” consonants.
Breath Hunter (breath detection module) identifies inhale and exhale regions using an adaptive sliding-context envelope algorithm.
Each module operates independently, and you can enable or disable it before analysis.
REAPER Vocal Cleanup Profiles (Female, Male, Spoken, Rap)
The script provides four detection profiles:
Female
Male
Spoken
Rap
Each profile adjusts internal thresholds based on vocal characteristics. Selecting the correct profile improves detection consistency.
When multiple modules detect overlapping regions, the script applies a maximum reduction wins rule. This avoids stacking automation and keeps the resulting envelope clean.
REAPER Vocal Cleanup Workflow and Automation
The built-in visualizer displays the waveform of the selected item and highlights detected regions.
After analysis, you can:
Adjust gain reduction per segment
resize segment boundaries
remove false positives
draw custom segments using Ctrl + Right-Click drag
The script supports two automation targets:
- Track FX
- Take FX
The script attaches automation to items and keeps it aligned during editing.
With Live Edit enabled, changing detection parameters rewrites automation in real time.
With Auto Scan, the script automatically analyzes newly selected items.
Requirements
REAPER v7.6x or later
ReaImGui (ReaScript binding for Dear ImGui), installed via ReaPack
Compatibility
Development and testing focused on Windows systems.
The script uses cross-platform logic and should work on macOS and Linux, but I have not fully tested these platforms. Feedback from users on these systems helps validate compatibility.
Installation
Via ReaPack (recommended)
Open Extensions → ReaPack → Browse Packages
Install ReaScript binding for Dear ImGui
Restart REAPER
Open Extensions → ReaPack → Import Repositories…
Add: Repository
Open Extensions → ReaPack → Browse Packages
Search for Floop Hunter Framework
Right-click → Install → Apply
Manual installation is available from the GitHub Releases page.
Limitations and Expected Behavior
Automated detection is never perfect. The script speeds up vocal editing but does not replace careful listening.
False positives — automation written on events that are not actually problematic, are possible, especially on heavily processed sources or unusual vocal performances. The system can also miss real artifacts in subtle or atypical events.
Treat the generated automation as a starting point. Always listen back with the automation active before committing to a mix, and use the visualizer to correct any inaccurate detections manually.
Frequently Asked Questions
What does this REAPER vocal cleanup script actually do?
It detects sibilance, plosives, and breaths and writes volume automation to reduce them.
Is this REAPER vocal cleanup process destructive?
No. The script writes automation envelopes and uses a JSFX plugin. The source audio is not modified.
What is the JSFX plugin used for?
The script installs Floop Hunter.jsfx in the FX chain to apply gain reduction independently from item gain.
Does it work on non-vocal material?
Detection is optimized for vocals. Other sources may require manual adjustment.
Can I use only one module?
Yes. Each detection module can be enabled or disabled independently.
Are detection results always accurate?
No. False positives and missed events can occur. The visualizer allows manual correction.
Is it compatible with macOS and Linux?
The script implements cross-platform support, but I have not fully tested it.
The project is hosted on GitHub, where the source code and manual installation files are maintained.
