For Windows musicians, the practical distinction is straightforward: Vocal Transformer changes the sound of your voice, while Functional Harmony builds an editable arrangement from a recording. They are not a single mode that transforms live singing, generates four-part accompaniment and sends MIDI simultaneously. Algoriffix documents these as separate workflows with separate constraints.
Vocal Transformer: change the voice, not the MIDI
Vocal Transformer provides scale-based pitch correction, octave-up or octave-down transposition, independent formant shifting, growl and added vocal harmonies. Formant controls change vocal character without necessarily changing the note: lower settings produce a larger, darker sound; higher settings make it smaller and brighter.
Two controls shape pitch correction:
- Catch governs transitions between notes, from an immediate jump at 0 ms to smoother changes up to 200 ms.
- Retune governs correction during sustained notes, with longer settings preserving more vibrato and pitch variation. Its range extends to 500 ms.
The important operational catches are less glamorous but more useful than another effects adjective. Passthrough must be enabled to hear the result, and MIDI output stops while Vocal Transformer is active. It supports Chromatic and Diatonic scales, not Automatic; its added vocal harmonies require Diatonic. Output is mono.
Algoriffix also offers a Low delay mode, quoting approximately 20 ms rather than 40 ms, with potentially rougher transformations. Those are developer-stated figures, not measurements from a WindowsForum test.
Functional Harmony: a recorded melody becomes an arrangement
Functional Harmony takes a different approach. Rather than attaching a chord independently to each detected scale degree, it considers the recorded melody as a whole and generates four-part harmony, using the key of each section and cadences to shape phrase endings. It works only in major and minor keys.
That differs from Pitch Perfekt’s existing Parallel harmony, which can send scale-degree chords as live MIDI. The transformer has its own audio version of parallel harmony, adding sung voices above or below the melody. Similar names, different destinations: MIDI events, processed audio or a converted MIDI file.
How to generate a harmony file on Windows
Algoriffix’s conversion guide documents this procedure:
- With Functional Harmony activated or its trial running, use a major or minor mode—Ionian or Aeolian.
- Right-click the robot button and select Functional harmony before conversion.
- Drop a WAVE, AIFF or MP3 recording onto that button.
- Wait for conversion to finish.
- On Windows, look for the resulting MIDI file beside the original audio file.
The Standard MIDI File includes note data and an optional harmony track, together with tempo, time-signature and key-signature tracks. Algoriffix suggests the standalone application for this task because it avoids involving a DAW and its processing chain. If no harmony is generated, check the selected mode first: Functional Harmony has no effect outside Ionian and Aeolian.
The practical appeal is an editable starting point rather than a finished recording. A sung sketch can become material for further arrangement instead of remaining stranded in a voice memo.
Setup details that can prevent a silent session
Algoriffix’s getting-started guide recommends selecting the microphone’s channel in Line input, then adjusting Gain until the meter reads around 0 VU. Enable Passthrough to check for clicks and pops, taking care to avoid audio feedback. For mains hum in North America, the documented filter setting is 60 Hz.
For live MIDI use, switch Vocal Transformer off. If default MIDI routing produces no result, Algoriffix advises selecting an explicit virtual MIDI output device from those available. Scale and parallel-harmony settings also require transcription to be stopped before adjustment. These checks separate routing problems from feature limitations—a useful distinction before blaming the microphone.
Existing users should also expect reactivation: versions before 2.0 used a different licensing system. Algoriffix says the earlier activation continues occupying a seat until revoked.
Compatibility, licensing and the rest of the update
Algoriffix currently lists Pitch Perfekt 2.0.1 for both platforms. Windows requires Windows 10 or later and an x86-64 CPU, with VST3, AAX and standalone versions. macOS requires version 12 or later and additionally offers AU.
The core application remains free. The developer lists the combined Premium package at US$69 plus tax, paid once rather than by subscription. Each premium feature has a seven-day trial with no credit card required; trials start through the crown button and require an internet connection. First launch also requires free license activation.
Beyond the headline additions, the v2 release notes list all seven diatonic modes, a mains-hum filter, key-aware note names, new meters, faster keyboard response, interface revisions and stability fixes. Improved pitch tracking is also listed, although the announcement supplies no comparative accuracy measurements.
The useful takeaway is not simply “more vocal effects.” Pitch Perfekt now offers distinct paths for performing with a transformed voice and developing a recording into harmonized MIDI. Choose the path first; the right buttons—and the missing MIDI—make considerably more sense afterward.