Savage Impulse Rekolla
Voice-over studio for Windows
A voice-over studio for Windows.
The recording session is the editing session.
For voice actors, podcasters, and narrators recording on Windows.
What it does
Rekolla records voice-over with one-key retakes. When you flub a line, tap spacebar, Rekolla scans the waveform back to the last natural silence and drops you in for the next take. Double-tap to redo the whole sentence. By the time you stop recording, the audio is cleaned, mastered, and ready to ship. No DAW. No timeline. Local-first, your audio stays on your machine.
How it works
You stumble. You don't stop. Tap spacebar and Rekolla scans backward through the waveform, finds the last natural silence (the pause between sentences), and rewinds there. You hear a 2-second lead-in of your last clean phrase, then a beep. You're back in the read.
The splice sits on a real silence, so it disappears in the final mix. Double-tap within 350 milliseconds and Rekolla skips back further, past the sentence, to the silence before it. Stop recording when you're done. The edits already happened.
What's in the box
Spacebar to step back to the last silence. Double-tap to redo a whole sentence.
After the take, Rekolla flags every "um," "uh," and over-long pause. You decide what to cut.
A small offline model runs filler detection without sending audio anywhere. Cloud transcription available if you'd rather.
A full signal chain (high-pass, denoise, gate, compress, normalize, limit) applied per content type. The file out is mixed and loudness-targeted, ready to ship.
Paste a script or load a .txt file. Auto-scrolls at adjustable speed, current line highlighted.
USB, XLR through an interface, built-in laptop. If the OS sees it, Rekolla records from it.
Configurable
Pick a session, get the right defaults.
Podcast and YouTube. Professional narration for audiobooks, e-learning, and corporate. Short-form for ads, promos, and social. Each preset tunes silence-detection threshold, filler sensitivity, default export format, and target loudness. The same one-key retake works in all three.
You stay in control. Or you don't.
Assisted mode steps through every flagged filler and silence one at a time. You preview each cut, accept or skip, and the keyboard handles the whole pass (Enter to accept, Delete to skip, arrows to navigate). Auto-clean mode marks everything for removal in one pass, with full undo if the result isn't right.
Cloud or local. Your call.
Cloud transcription is faster and handles accents and technical vocabulary more reliably, sends the segment being analyzed to a transcription API and uses your key. The local engine is a small model that downloads once and runs entirely offline, nothing leaves the machine. Set the choice per session.
Get Rekolla
$29.99 · One-time · Windows
No subscription. All version 1 updates included.
No account. No cloud. Your audio stays on your machine.
From the shelf
A small house of desktop tools for content work. Local-first. No telemetry. No account.
