MAutoPitch

MAutoPitch is a real-time pitch correction plugin that snaps incoming vocal audio to a chosen scale, smoothing out sung notes that drift away from the intended pitch. It runs as a VST, VST3, AU, or AAX plugin inside any modern DAW, sits on a track in the signal chain, and processes audio with low CPU overhead. The plugin is free for personal and commercial use, which puts it in a small group of legitimately capable pitch tools that don’t require a license fee.

The category it competes in is dominated by paid solutions: Antares Auto-Tune defined the genre, Celemony Melodyne handles offline pitch editing with surgical precision, and Waves Tune sits between them at a lower price point.

MAutoPitch is built for the real-time end of that spectrum, the side concerned with live correction during tracking and mixing rather than note-by-note offline editing. Within that scope it’s surprisingly competent, especially considering it costs nothing.

How the pitch detection and correction works

The plugin runs a pitch tracker on the incoming signal, identifies the fundamental frequency of each note, compares that against the closest note in the selected scale, and bends the audio to match. The speed of that bending is controlled by a Depth and Speed parameter pair. High speed produces the aggressive snap-to-pitch sound associated with the T-Pain effect (and a thousand pop records since). Lower speed produces subtler, more transparent correction that lets natural vibrato and pitch movement survive while still pulling drifting notes back into key.

Scale selection accepts the standard set: chromatic, major, minor, plus several specialized scales. The chromatic option corrects to whatever note is nearest, which is the safest setting for vocals that may include accidentals or modulate between keys. The diatonic options force notes into a specific key, which catches more pitch errors but can mangle intentional bends and chromatic passing notes. The right choice depends on the song and the singer.

Formant correction is the feature that separates usable pitch correction from the chipmunk effect. When you pitch-shift audio with a naive algorithm, vowel formants shift with the pitch, producing the cartoon-character sound that early pitch shifters were famous for.

MAutoPitch preserves formants by default, keeping the timbral character of the voice intact even when the pitch gets moved significantly. The formant shift slider lets you intentionally alter that character if you want, useful for creative effects or gender-bending vocal experiments.

The width and depth controls

Two parameters less common in the category are the depth and stereo width controls. Depth determines how strongly the correction pulls toward the target note. At zero, the plugin passes audio through unchanged. At maximum, every note locks to the nearest scale degree. The natural setting for transparent correction sits around 30-50%, with higher values reserved for stylistic effect.

The stereo width control adds a chorusing effect by detuning the pitch-corrected signal slightly between left and right channels. It’s an unusual addition to a pitch corrector and one of the more creative features the plugin offers. Used subtly, it thickens vocals in a mix without sounding obviously processed. Pushed harder, it produces an artificial doubling effect that can fill out a thin solo vocal without needing actual stacked takes.

The dry/wet mix is the safety net. Blending the corrected signal with the original reduces the perceptibility of the processing while still catching the worst pitch errors. For a vocal that’s mostly in tune but needs gentle help on a few notes, a 30-50% wet mix often produces more natural results than 100% correction at low depth.

Workflow inside a DAW

Setup is the same as any other plugin. Drop it on a vocal track, select the scale, pick a depth and speed, and adjust to taste. The interface is small, the controls are immediate, and there’s no project file or external editor to manage. This is a deliberate design choice: MAutoPitch is for the case where you want a knob to turn, not a piano roll to edit.

For more granular editing, the workflow pairs naturally with a DAW like Audacity for cleanup before correction (removing breaths, normalizing levels, trimming silence) or with a dedicated offline pitch editor for note-by-note work.

The plugin handles the real-time tracking and mixing case where global correction settings make sense across an entire performance. It’s less useful for hand-tuning individual notes in a chorus or fixing single bad words in an otherwise clean take.

CPU usage stays low across a full session. A typical mix with the plugin on every vocal track, doubles, and ad-libs adds only minor processing load, which matters for users running large arrangements on modest hardware.

Latency considerations during tracking

Pitch correction inherently involves analyzing audio before processing it, which introduces latency. MAutoPitch keeps that latency low enough for monitoring during recording in most cases, though singers sensitive to even small delays may prefer to track dry and apply correction afterward.

The plugin reports its latency to the host, so DAW delay compensation handles playback correctly. During tracking with input monitoring through the plugin, the latency adds to the round-trip delay of the audio interface, which is where total system latency becomes audible.

Lower-latency audio drivers like ASIO4ALL reduce that round-trip time enough that the combined delay stays comfortable for most singers.

Limitations that matter

The plugin is good but not perfect, and the limitations are worth being honest about. It tracks pitch reliably on clean solo vocals. It struggles with overlapping voices on the same track (which is why backing vocals usually need their own instances).

It can mistrack on heavily breathy or whispered passages where the pitched component of the signal is weak. And it has no manual override for individual notes, meaning a one-off correction error has to be fixed by automating the plugin’s bypass on that specific section.

For polished commercial vocal production, dedicated paid alternatives still have an edge in tracking accuracy on difficult material and in the depth of available controls. For demos, podcasts, livestreams, hobbyist recording, and the substantial portion of music production where transparent gentle correction is the goal, MAutoPitch does enough to make the paid options feel optional rather than required.

Conclusion

MAutoPitch sits at a specific intersection: free, real-time, capable enough for serious work, and limited in ways that only matter for surgical editing. The target audience covers home recording musicians, podcasters fixing the occasional off-key intro, livestreamers wanting transparent voice polish, beginning producers who need pitch correction without a license fee, and demo producers laying down rough vocal tracks.

It’s the wrong choice for users doing high-end vocal production where every nuance of tracking accuracy matters, for offline note-by-note editing of complex multi-track vocal arrangements, or for anyone who needs the deep parameter set of dedicated commercial solutions.

For everyone else who needs pitch correction to work, sound transparent enough not to draw attention, and not cost anything, the plugin delivers more than its price suggests it should.