Røde AI-Micro USB-C phone interface

I am writing this from a field recordist’s point of view, not a podcaster POV. Podcasting seems to be the design intended use case.

There’s nothing more commonplace in the modern urban setting than somebody with earbuds glommed onto the delights within their a smartphone. Pair some binaural in-ear microphones like Soundman OKMIIs with a smartphone and you potentially have an inconspicuous and high performance street field recording rig. You’ll also look like you’re poor and have a cheap phone, because everyone haz bluetooth wireless earbuds, right?

Problem is that smartphones have only mono microphone inputs, or if they are made by Apple none at all. You get cheap USB-C to 3.5mm jack interfaces, but most are mono in, providing the features of the TRRS jack you used to get. I had high hopes for the Røde AI-micro – a little bit dearer than your Chinese noname product, but more functional – offers stereo in. Røde is a well-known manufacturer of microphones, including some very low noise models – the NT2000 is legendary for a very low self-noise of 7dBA.

The Røde AI-Micro worked pretty well as designed, and will be fine for close-up voice, but as a field recorder front end it’s a bust. Compared to an Olympus LS-10¹ from nearly 20 years ago, it’s noisy. Very noisy.

To test this I injected a -67dBu tone at 150 ohms² source impedance into the mic input, set to stereo, max gain. The Røde AI-Micro has more max gain than the LS10, so I had to boost the LS10 recording by 8.8dB to make the tone peak the same (to within 1dB). This ensures I am comparing like with like, with a known reference point. Both recordings were made at full gain, and neither was clipping.

These two tracks are 30 seconds of tone, followed by silence³ – the generator turned off but still terminating⁴ the input with 150Ω.

Røde AI-Micro

Olympus LS-10, boosted by 8.8dB to match tone levels

One listen to these, particularly the second half tells you pretty much all the field recordist needs to know.

I used Reaper’s JS:Audio stats for some quantitative details. It’s not an Audio Precision test set or software, I’d guess error to be ±2dB what with the manual level matching, but confirms what your ears told you. The relevant fields are RMS window max of about -7dBFS when the tone is on and RMS window min of about -33dBFS (Røde) and -55 dBFS (LS10). Reaper kindly does the difference calculation for you in RMS dynamic range of about 25 (Røde) and 48 (LS10)

compared with the Olympus LS-10

The dynamic range is somewhat higher with the Olympus, because the noise floor is lower, as you could hear if you compare the playback after 40s in each case..

Settings:

Røde AI-Micro was used with the companion Reporter app, gain slider max (24dB), standard 48kHz 24 bit WAV, stereo on input 1, high pass filter off. Since this is connected via the digital USB-C interface performance should be determined by the interface and the settings. The phone was battery powered.

Olympus LS10 mic sens high, manual level set, rec dial 10, low cut off, 48kHz 24 bit WAV. Recorder was battery powered.

Tone source was battery powered, and terminates the input at 150Ω.

A smartphone isn’t going to replace my field recorder any time soon

The Olympus LS-10’s mic amps have of the order of 20dB lower noise than the Røde. Podcasters, you really don’t need to care about this. Your subjects are bellowing into the mic at a couple of feet, as opposed to trying to record a blackbird 20 yards away. You’re going to rack the gain back, and everything will be dandy.

Field recordists can’t afford the 20dB noise penalty. End of story.

The Røde AI-Micro had some other foibles to irritate the field recordist. It’ll do plug in power if you use the mic inputs as two separate mono inputs. It’ll do stereo 3.5mm TRS on input 1. But you’re not getting plug in power on that stereo input. Why ever the hell not? Headphone output is mono. I can sort of see why they did that, for the podcaster combining two interviewees’ microphones, but it would be better as an option, or automatically linked to the configuration settings – merged = mono, split and stereo = stereo.

So to use this with binaural mics requiring plug in power all of a sudden you are toting a stereo to mono breakout cable to get plug in power, or use the stereo in and a battery box, and by the time you have loaded your smartphone down with all the extras you start looking like a one-man-band.

Not only that, you have another battery to worry about in the Soundman A3 PiP battery adapter⁵. You have a perfectly good battery in your phone, and Røde sometimes lets you use it to power your microphone. Just not with the stereo input. Bah.

I really wanted to like this product, but it’s no damn good for my use. I will return this to Amazon, and not risk the wrath of His Jeffness by looking for any more small smartphone stereo mic adapters. Perhaps I should have heeded the ominous warning

in fairness compatibility was fine with both the phone and companion app. It’s the performance and utility I have issues with

AI bandwagoning – grr

I see what you did there, Røde. AI = audio interface rather than the much overhyped alternative use of the moniker.

1. I chose the Olympus LS10 because it was a decent low noise handheld field recorder. You can get better performance with Zoom and Sound devices, but the LS10 and its ilk were small handheld recorders that are of a similar size ot a mobile phone. While these handheld recorders commonly have internal microphones, you can use them with external ones too ↩︎
2. Typical of a balanced condenser mic output, and lower than the typical 2k of a plug-in-power input ↩︎
3. I set the recorders in record pause on the tone before turning the generator off, which is why you don’t get a big glitch when the tone stops. ↩︎
4. I constructed the tone generator, the output pad is 82kΩ in series with 150Ω to ground, output is taken across the 150Ω resistor. The output impedance of the generator will change when switched off, but the worst case change is from 150Ω to 149.7Ω if Zo changed from dead short to open circuit. That’s good enough for government work 😉 ↩︎
5. I know what you’re thinking. £60 is far too much to pay for a plastic box with an oddball 6V battery and a couple of resistors and 3.5mm TRS jacks and sockets, particularly as I know perfectly well how to build one. But the one I constructed uses a 9V battery and it’s almost as big as the phone. ↩︎