let's say I vocalize a baseline, then want to convert that into a synth bass. Is a two-part workflow the best (only) way to go at this time (audio-to-MIDI converter w/in DAW and routing the resulting MIDI to a synth)? Software, not hardware. Is it possible to keep it all as audio and not using MIDI?

 

Last edited by a moderator: Sep 9, 2025

Standard Workflow

The process begins with recording the vocalized idea in your DAW.

Use an audio-to-MIDI converter (either built into your DAW or as a third-party plugin like Samplab, Melodyne, or Basic Pitch) to translate the pitched elements of your voice into MIDI notes.

The generated MIDI can then be edited for timing, pitch, and length to better match the musical intent or correct artifacts.

Finally, route this MIDI output to a synth or bass instrument of your choice within the DAW, allowing for further sound design and mixing.

 

Thanks, PulseWave, that's the standard way to do things . . . and it works, but wondering if there's an easier or better way . . . keeping it all in .WAV maybe without going MIDI?

 

I look into the crystal ball in front of me:

For an easier, all-audio alternative, start with iZotope VocalSynth 2 or a similar plugin; it's streamlined and yields pro-sounding results without leaving the WAV domain.

Spoiler: Source: ChatGPT KI

You're right to question the "default" audio-to-MIDI workflow — it's effective, but it's not the only path, especially with modern software. Whether or not to use MIDI depends on your goals (flexibility vs realism vs speed), but yes, there are audio-only workflows that skip MIDI entirely, and some hybrid ones worth knowing about.

Here’s a breakdown of the best workflows for transforming vocalized basslines into synth bass without necessarily relying on MIDI — or using MIDI only as needed.

️ STANDARD: Audio-to-MIDI to Synth
(What you're already doing)
✅ Pros: Very flexible, can quantize/edit easily
❌ Cons: Can be time-consuming, pitch tracking isn’t always accurate

Workflow:

1. Record vocalized bassline (WAV)
   
   
2. Use DAW’s audio-to-MIDI converter (Logic, Ableton, FL, Cubase, etc.)
   
   
3. Route MIDI to soft synth (Serum, Massive, Operator, etc.)
   
   
4. Edit MIDI as needed
   

️ ALTERNATIVE: Audio as Modulator (No MIDI Needed)
Convert voice directly into a synth-like tone using audio FX chains.

Workflow:

1. Record vocalized bassline (WAV)
   
   
2. Use pitch tracking effects to turn voice into a synth (examples below)
   
   
3. Shape the tone with distortion, EQ, filters, modulation, etc.
   
   
4. Layer/duplicate for thickness or stereo spread
   

Tools & Plugins:

• Melda MUnison + MWobbler – Turns mono pitch into stereo and adds modulation
  
  
• Ableton's Resonators or Corpus – Adds harmonic resonance based on input pitch
  
  
• Vocoders (Ableton Vocoder, TAL-Vocoder) – Use voice to control synth-like carrier
  
  
• Zynaptiq Morph or iZotope VocalSynth – Morph voice into synth tones
  
  
• Bitwig Audio Rate Modulators – Audio controls synth parameters directly
  
  
• Kilohearts Snap Heap – Creative audio modulation chains
  

✅ Pros:

• Keeps everything in the audio domain
  
  
• Sounds more organic/raw
  
  
• Fun to experiment with
  

❌ Cons:

• Less precise note control
  
  
• Can be hard to “clean up” if tracking is messy
  

HYBRID: Audio-Controlled Synth Parameters
Use your vocal as an envelope follower, pitch follower, or mod source.

Example:

• Route your vocal to a synth's filter cutoff or amp envelope
  
  
• Use envelope follower plugins to map dynamics to synth parameters
  
  
• Use Pitch Tracking (Audio-to-CV) to control oscillator pitch in real time
  

Tools:

• Bitwig Grid
  
  
• Ableton CV Tools
  
  
• Reaktor or VCV Rack for modular routing
  
  
• Puremagnetik plugins (weird, audio-reactive FX)
  

Creative Tip: Use Melodyne or Spectral Editors (No MIDI)

• Open your vocal line in Melodyne or RX Music Rebalance
  
  
• Extract harmonics or resynthesize portions of the vocal
  
  
• Copy formant and pitch contours, paste to another layer
  

This lets you "sculpt" your voice into a synth-like source without needing MIDI conversion.

Summary: Which Workflow Is Best?
Workflow When to Use Tools
Audio-to-MIDI You want precise pitch control/editing Any DAW, Melodyne
Audio FX Chain (No MIDI) You want quick results, no pitch correction Ableton FX, iZotope, Bitwig
Audio as Modulator You want organic, expressive control CV Tools, Envelope Followers
Spectral/Resynthesis You want experimental, evolving textures Melodyne, RX, VocalSynth
Let me know your DAW and synth plugins, and I can recommend a workflow that fits your setup.

 

I've found that Jam Origin's MIDI Guitar works pretty well for vocals too though it's designed for realtime guitar/bass to MIDI. I keep one instance always open so I can play stuff or feed a sample and have it instantly output as MIDI with no offline conversion needed.

It's worth a shot certainly.

 

This ^
Also, I haven't tried it, but some skilled pal here mentioned that applying some pitch correction plugin before "MIDI Guitar" helps to get more "steady" notes. I mean to avoid quick jumps 1 or 2 semitones up and down. Makes sense. An Autotune or similar

 

Ohh, for vocal input that might be just the trick. My guitar playing is in tune enough not to suffer from this, but from humming bass lines this might just be the ticket.

Thank you for this idea, gotta try this.

 

Put Antares Autotune one insert slot before (above in Logic) your Melodyne and tune it there before it ever hits Melodyne; when you transfer the audio into Melodyne. You can get rid of almost all the pitchiness first. You can also do a similar thing with Nuro Xpitch or other tuner if you have a good one you like already. You can put it into Autotune Graph Mode even, if you have >v11.

I do the autotune trick almost every time I use Melodyne, because I try to do as little as possible in Melodyne. It is tedious.

 

I’m sorry my answer is becoming a broken record on this forum.

but that’s one of the beauties of Suno. You can dry beatbox, sing, hum, whistle audio into it and it will generate what you want as well as convert that to midi.

you can also do any of those things over the top of your current track and it will stay in key with it.

if Suno isn’t your vibe. My fave audio to midi monophonic tools are

imitone - it’s really good at locking into key and fine tuning what you are trying to convey through your mouth (no diddy)
https://imitone.com/

Vochlea - pretty much the same as above with a larger price point. Also has chords and drums
https://vochlea.com/

 

And I think we found him

You said it. For guitar that plugin alone is incredible. But for vocals/humming (my case and I sing horribly bad) you might want to use that trick

Suno does incredible things, but sometimes you want to stay in your DAW workflow.
Didn't know it can also do audio to MIDI. Is there a decent free plan in Suno?

 

Last edited: Sep 10, 2025

Is Suno online only or is there a Mac version? Or, better yet . . . does the Sister site have it?

 

Spoiler: Source: Grok KI

Suno is primarily a web-based application accessible through browsers on various operating systems, including macOS, Windows, and Linux, without the need for a dedicated desktop application. There is no specific Mac version of Suno, as it runs smoothly on any modern browser like Safari or Chrome on a Mac.

Additionally, Suno offers mobile apps for iOS and Android, but these are designed for mobile devices and not specifically for macOS. However, you can enhance the experience on a Mac by using WebCatalog Desktop, which allows Suno to run in a distraction-free window with additional features like account management.

If you’re looking to download Suno-generated music on a Mac, you can use third-party tools like sunodownloader.io or suno-top.com by pasting the song’s share link, which works seamlessly on macOS browsers.

 

Standard Workflow (Audio → MIDI → Synth)

Yes, this is currently the most precise and flexible method.

Steps:

• Record audio (e.g., your vocalized bassline in your DAW).
• Convert audio to MIDI using tools like:
• Melodyne
• Ableton Live's "Convert Audio to MIDI"
• Samplab or AI-MIDI2
• Route MIDI to a software synth (e.g., Serum, Massive, Arturia).
• Design and automate your sound.

Advantages:

• Precise control over pitch, timing, and velocity.
• Easy layering and editing.
• Synthesizers respond directly to MIDI data.

 

Understandable about staying in daw. Sadly imitone, and Vochela are both out of daw tools too.

Suno has the best audio to midi algo i have tried over the years. Barely any ghost notes. The free plan doesn't give those features. Lots of cloud based algos running to stem separate into 12 stems and convert to midi.

No to both. It will always be subscription based because its multiple stacks of ML algos that need to be glued together for all of the features. Each path you take with a tool needs multiple API calls doing the most on the back end.

There are local versions of single features of one part of suno that arent that good.

I always think of this diagram when it comes to most of what we do in music and this also applies to these new AI algos.



DAW steps:
Record audio + autotune + Melodyne + get midi + spend some time cleaning up midi....... (free-ish with sister site help)

Suno steps:
Record Audio + upload audio + Generate audio + Stem audio + get clean midi...... (monthly sub)

similar destinations pick a path that gives you what you want

 

And also understandable the other way around. Advanced AIs are way too compute-intensive to work offline, let alone real-time.

 

Yep all depends on the task. But like that image I posted shows. Pick a side of the venn. If you want quality, or real time. It might take away from the others.

I try to pick the best tool for the task with the path of least resistance.

 

https://bedroomproducersblog.com/2025/09/10/pitch-perfekt-pitch-correction/

you could also try this. They just updated it

 

Why some people have an aversion to using Midi is still not a normal thing to my way of thinking. If you want to make music it is nearly impossible to completely avoid. At some point, it becomes the only option.

But to avoid Midi completely, AI suggested Orange Vocoder and Izotope Vocalsynth with them put into Pitch Tracking mode, instead of Midi mode. It didn't mention it, but I think you could probably use Arturia's Vocoder the same way. It's more of a big synth than a standard vocoder plugin usually are.