Audio to MIDI workflows are becoming a staple of music production, especially in the age of artificial intelligence. From basic AI MIDI transcription of existing songs to creative applications like beatbox-to-MIDI software, there's something here for everyone.
Table of contents
Samplab 2: Audio-to-MIDI Plugin + Standalone app
Samplab is a DAW plugin and standalone application that transcribes audio files into MIDI. It includes stem separation for rhythm, bass and harmony instruments, making it easier to drag them onto separate MIDI tracks in your DAW.
Unlike traditional MIDI, Samplab's note-manipulation can act directly upon the original instrument timbre. So if you'd like to change the a melody, bass line or chord progression without losing the original tone, just drag the notes up and down on their built-in piano roll. You can bounce as audio or MIDI at any time.
The MIDI transcriptions include an accurate chord detection feature, so you can learn quickly get a feel for the progression in any song. The MIDI chord blocks can be modified in app and dragged into the DAW for virtual instrumentation.
In 2024, the company launched a free audio-to-midi browser tool that you can use without an account. This is a great way to get started and see what the app is capable of. Read our full review of Samplab here.
RipX DAW: Audio to MIDI + Audio Manipulation
RipX is a powerful AI DAW that includes a strong audio-to-midi converter. Where other services like Samplab run on cloud services and a subscription model, RipX runs locally on your machine with a one time purchase.
How it works: Drag an audio track into the workstation and within 1-2 minutes, stem separation and MIDI transcription are complete. Like Samplab, you can move notes up and down, retaining the timbre from your original audio.
We found that the app works best with a clean mix. Tracks with lots of noise, distortion, and effects will still be processed and transcribed with a fair amount of accuracy, but you may find some ghost notes attached to the fundamental tones.
Fortunately, RipX includes a number of audio manipulation tools on the right panel, including noise removal and auto-mix. So if you want to transcribe a noisy track to MIDI, try some pre-processing first.
Read our full review of the RipX DAW for a complete overview.
ACE Studio: A popular AI voice-to-MIDI DAW
If you need a tool that can modify and refine vocal tracks, we recommend ACE Studio. The DAW was originally designed for text-to-singing with AI singing voices, but they have continued to evolve and now support vocal audio to MIDI transcription. Samplab and RipX are better for instrumental transcriptions.
In 2024. ACE released an impressive new AI acapella to MIDI feature that not only captures the notes of your vocal take, but even transcribes the lyrics into the DAW. Provided you're feeding it a clean stem, you should be able to edit the melody, change nuances about the vocal performance, and update the lyrics while holding on to the original vocal style.
You've seen vocal pitch correction software like autotune, but re-rendering a vocal stem with new lyrics and expressive controls is groundbreaking.
Check out the ACE Studio website for more info on their audio to midi features.
Vochlea’s Dubler 2: A realtime voice-to-MIDI plugin
Whether you’re importing a vocal melody or recording a track in realtime, there’s little debate that Vochlea has the best voice-to-MIDI plugin on the market. For those who don’t feel like watching the full 30 minute tutorial, we’ve put together a summary of what to expect.
Dubler 2’s interface includes several tabs, effectively saving you space in the DAW. At the top of the app, you’ll see a project title that allows you to save presets for each song. So you don’t have to recreate the environment as you bounce between projects.
Play - The play screen is your primary interface when you start recording. Melodic vocals will show up on the pitch wheel and MIDI keyboard. On the left, you can monitor the volume of your vocals to ensure they’re not clipping. Next to that meter, you’ll find a set of one-shot pads where you can drag samples to be triggered by your voice. When you click a pad, it toggles over to the Train tab automatically.
To the right side of the screen, Dubler presents four knobs representing your envelope and filter automations. These help you to visualize your vocal articulations, mapped to a third-party plugin via the Assign tab. We’ll get to that in a minute.
Train - This is where you’ll train the application to recognize important vocal triggers. If you’re looking to transfer your beatboxing to a MIDI Instrument, this is where the magic happens. Select a MIDI note as your output and you’ll be able to trigger the corresponding sample in your instrument rack. For example, if you make a “pf” sound with your lips you can map it to a kick drum, while a “ts” sound might me assigned to a hi-hat. You get the idea.
Key - Define the key signature of your vocal melody to auto-tune your vocals and display the right notes on your Play screen. As you change the key signature, the pitch-wheel will update and remove the notes that don’t belong.
For simple melodies with less than seven notes, you can take it a step further and de-select notes from the pitch-wheel so auto-tune doesn’t accidentally hit them. Once you hit record on the DAW, your vocal melody will be recorded as MIDI notes in real-time.
Chords - Tab over to the chord settings to setup polyphonic triggers. So if you sing a single note, the app will add additional MIDI notes on top to fatten it up. Best of all, it stays within the key signature of your choice. Choose from presets like Triads, Pop Simple, and Pop Advanced. At the bottom of the screen, you can decide whether chords are played in close harmony (within an octave) or spread out over multiple octaves in an open voicing.
Assign - If you prefer to use Dubler’s built-in synth, this screen is where you’ll find a variety of MIDI instrument presets for both single-note and chord voicings. Below that section, the Pitch Bend feature lets you define how far you can slide vocally in either direction. Most pitch wheels on a MIDI controller only support a whole note in either direction. Dubler extends that range dramatically, so you can configure whatever your vocal melody calls for.
Technical Specs: The Dubler 2 VST is compatible with most digital audio workstations on Windows or MacOS. The standalone app has been confirmed to work with Logic Pro, Cubase, Garageband, Ableton Live, and FL Studio. So just about any producer can use it.
Vochlea’s Dubler Studio Kit 2 includes a signature USB microphone that’s optimized for pitch recognition and noise-reduction. If you prefer to use your own mic, you’ll calibrate when you first boot up the software and then you can jump right in. The app is compatible with all Mac Audio Interfaces. If you're on Windows, your audio interface will need dedicated ASIO drivers.
Ableton’s Audio to MIDI Converter
If you’re an Ableton user, save yourself some money and try out their native audio to midi feature in the sampler. You'll find options like Convert drums to new MIDI track and convert melody to new MIDI track. Users right click on an audio file and generate a separate MIDI track from the sample’s transients. You can use any file type, but wav and lossless formats are best to avoid unwanted artifacts.
One impressive part of converting audio to MIDI in Ableton is that you don’t have to train the software. The algorithm will automatically detect low impact sounds as a kick, snappy “t” sounds as a snare, and hissy “s” sounds as a hi hat. The MIDI file will be generated on a generic drum rack that you can update after the fact in your piano roll.
Ableton’s melody converter works in a similar way. Choose an existing monophonic melody and transpose it to a MIDI track. Then you can assign the synth or musical instrument that you want to use. If you establish your project BPM ahead of time, it’s easy to quantize the MIDI output.
Popular Audio to MIDI Conversion Tools
Here are a few more audio-to-midi software options that we liked. They're not quite as powerful as the tools mentioned above but they deserve a mention.
Wealthy Sound has a popular page that shares tutorials for native audio-to-midi features in Pro Tools, Cubase, Logic Pro X, and FL Studio. Save yourself the hassle and use what’s right in front of you!
Waves OVox is a competitor to Vochlea’s Dubler 2. It includes many of the same features, including chord harmonizing and arpeggios effects. With hundreds of presets to choose from, named after the famous artists that it’s imitating, you can jump right and in get the sound you’re looking from. Includes 9 customizable LFOs/Sequencer, ADSR and ORS modulators, plus effects like AutoPan, Chorus, Comp, Delay, Distortion, and Reverb.
Jam Origin offers a tracking processor specifically for guitar players, so you can play riffs and turn them into accurate MIDI transcriptions. Like Vochlea, Jam Origin includes complex chord recognition and key mapping, built-in synths, and a variety of output controls.
Imitone deserves recognition for its fun, colorful interface. It would be a great tool for younger audiences that want a simpler tool with the same core functionality.
Spotify’s Basic Pitch is a free browser application that will convert audio recordings to MIDI using open-source, machine learning algorithms. Unlike the other applications listed here, Basic Pitch can transcribe full chord arrangements. It works best when the audio features a single musical instrument and doesn’t have excessive delay or reverb in the mix.
Runners up
A2M is a free iOS and MacOS application. It’s one of the more reputable Audio-to-MIDI tools for mobile devices, with an advanced feature set to help control latency during real-time performance.
Guitar Tabs X released a Voice-to-MIDI app on the Google Play store that apparently introduces a lot of extra noise and artifacts. Reviews indicate that it can be fun to play with but not a good fit for professional musicians.
Dodo MIDI 2 is a free, lightweight plugin with only a couple of parameters. It could be a good starting point if you’re intimidated by feature-rich tools like Dubler 2.
MIDIMorphosis was an early player in the Audio-to-MIDI space, but the app has not been updated in almost a decade, rendering it somewhat obsolete.