I’ve admitted before that I am a sucker for strange, and often gimmicky, musical devices. A giant touchpad with swappable silicone playing surfaces? Sign me up. A bisected grapefruit that you tilt and smack? I’ll take two. So, an app that lets you play a synthesizer simply by humming a melody was clearly going to grab my attention.
Dubler first debuted last year as a packaged deal: a custom designed USB microphone and an app that turned your beatboxing, singing or any other audio really into MIDI data. With the launch of Dubler 2, Vochlea is making the special mic optional. The new app allows you to quickly and easily calibrate it to work with any microphone or audio source of your choice. But I want to be frank about something right off the bat: Dubler works best with its own special microphone.
When I tested Dubler using both an SM57 and an SM58 run through a Focusrite 2i4 the results were decidedly mixed. The MIDI data was always a little messier, there was more noticeable latency and the pitch tracking just wasn’t as good. The latency was mostly an issue when using Dubler to trigger drums. With the buffer size in my audio settings at 128 samples it was almost impossible to keep a rhythm as the drum hits came just a little too slow and kept throwing me off.
If you use the dedicated mic sold by Vochlea, though, the story is a bit different. In fact, Dubler 2 is quite impressive. There is basically no noticeable latency. Pitch tracking is surprisingly accurate. And the resulting MIDI data is reasonably clean… or at least as clean as the source. Which I suppose is one of the first things I should lay out: If you can’t carry a tune, Dubler is only going to be so useful. My vocal skills are borderline nonexistent under the best of circumstances. But the onset of fall allergy season has made things significantly worse. (I would describe my current vocal style as chain-smoking donkey.) Could I get Dubler to spit out something resembling a melody? Sure. But it wasn’t always easy and I had to clean up a lot of MIDI after the fact.
So, I did what any smart person would do: Enlist the help of their sister who has spent a good chunk of her life singing and training with vocal coaches.
The difference was immediately obvious. Where Dubler often struggled to figure out what was going on with my flat unsteady humming it was generally able to follow my sister’s vocal histrionics without much issue. You can tell Dubler to automatically snap notes to a particular scale which I highly recommend regardless of your skill level. But even in chromatic mode the app was able to pretty accurately identify notes my sister sang and even follow along the vibrato of her voice fairly well.
To get truly usable results out of Dubler though, you’ll need to learn to use it properly. While it’s fun to belt out a melody or clumsily beatbox your way through a drum part there are a lot of tricks to making the most of the app. For one, you’ll need to tailor your vocal sounds so that the app can recognize them easier. This is especially true of the drum triggers. Simple beatboxing is probably going to trip up the software unless you’re really good. Sounds like “ti” “ts” and “ta” will often get mixed up, so it’s better to use more distinct sounds that you might not reach for if you were actually trying to beatbox, like “ka”. It’s also best if you keep the number of triggers you use to three or four. You can program up to eight, but it’s hard to make varied enough sounds for it to be useful.
The same is true for melodies too. If you just hum with your lips closed, Dubler will struggle more to tell when you’ve changed notes and you’ll end up with messier data. It’s better to go with percussive sounds like “da” to help separate notes from each other.
You can also put Dubler in chord mode which I actually found was a fun and interesting way to come up with new progressions when I was feeling stuck. But perhaps my favorite feature is the ability to turn vowel sounds into MIDI CC information. That means that if you held a note and slowly shifted it from an “ooo” to an “aaa” you could cause the filter to open up. It’s difficult to control these features though, and it was better for creating strange and interesting noises than actual music.
If you’re vocally inclined then Dubler holds a lot of promise as a musical sketch pad. I sometimes struggle to turn what I hear in my head into melodies on a keyboard or guitar and humming it might be the quickest and easiest way to get an idea out.
While the ability to use the app with your own mic is appealing, right now the calibration seems like it could use a little work. That being said, my microphones are old and have seen a lot of abuse. And it’s not like the SM57 is known for its perfect vocal fidelity. I’d probably have better luck if I had SM7B fresh out of the box to test with. But I had to make due with what I had. And, what I had, wasn’t as good as Vochlea’s Dubler Studio mic when it came to capturing clean MIDI data. The software is pre-tuned for that mic and there’s fewer variables to contend with.
Which leads us to our final point: price. Dubler is not cheap. And that’s not surprising. While Vochlea isn’t the first or only company to try turning audio into MIDI, it’s still not a simple task. So you’d have to be pretty committed to the idea of singing to control your synths to drop $329 on the Dubler Studio Kit or even the $249 for the app on its own ($78 to upgrade if you own the original Dubler already). For someone like myself, it’s a tough sell. But for a vocalist who wants to start dabbling in the world of music production and synthesizers (even just for idea generation) without having to learn a whole new instrument I can certainly see the appeal.