The headset mic input on all current iOS devices uses the same filter as the built-in mic, (depending on iOS version) and so offers no improvement to performance over the built-in mic.
Because of this, we can’t recommend any mic that plugs into the headset connector, such as micW. Although micW may perhaps be using a higher quality capsule than the iPhone mic capsule, since it is using the headset mic input, it is subject to the same filters and Apple analog electronics as the built-in mic.
Again, even in iOS 6, we see no improvement in frequency response, or flatness over the Apple built-in mic, with micW.
micW does have lower sensitivity, so it will clip at a higher SPL point, although it’s noise floor is also therefore higher, so it can’t measure as low of noise levels. The calibration we normally see for this mic is around +13.
Connecting an External Mic
The headphone / input jack on iOS devices has a 4th pin that has a small polarizing voltage on it with very low drive current. The Apple device monitors this line, and can tell when a typically headset mic is plugged in. If its not, they don’t even turn on the mic input. Most people don’t have very much success connecting a non-Apple mic to this port.
Some of the third-party mics, like MicW and IMM6, do use this input, but really they are just swapping one small mic capsule for another. No great improvement here.
Also, there are no analog inputs on either the 30-pin or Lightning jack, so its not possible to connect something there.
Really the only way to get an external mic into an iOS device is with one of the third-party adapters, like for example our iAudioInterface2 or iTestMic, which convert the mic signal to digital audio before routing it to the iOS device.
Another Flaw in the Headset Jack
We have recently discovered another major flaw in the Apple headset connector. It seems that the way that Apple has implemented the headset jack as an input and output connector, when the output is active, it bleeds back into the mic input. This bleed can easily overpower the mic input signal, rendering the input completely useless.
The bleed gets worse as the output load impedance goes down, and as the output level increases. Using a typical high-impedance input, such as the line input on a consumer receiver, the bleed may be low enough that if the acoustic level in the room is fairly high, it won’t make much of a difference to the measurement. The bleed is much worse when the generator is in mono mode, since the left and right channels are both sending out the same signal. In balanced mode, the problem nearly disappears. However, in balanced mode, the signal to each channel gets distorted, apparently due to the crosstalk between the channels.
The summary is that if you do plug a mic into the headset jack, don’t use the output. Even turning on the output with no load attached can cause phantom readings rising up to +20 dB at 20 kHz.
The amount of this bleed, or crosstalk, varies widely from model to model. In every case, there is at least 20 dB of boost at 20kHz. In some devices there is as much as 50 dB wide band — across the entire audio spectrum.
We have tested a number of iOS devices, using Sony over-ear headphones as the load, and verified this flaw in the following devices. The dB level shown is the wide band boost.
- iPod touch 5 – 50 dB
- iPhone 4S – 10 dB
- iPhone 5 – 20 dB
- iPhone 5S – 10 dB
- iPad 3 – 50 dB
- iPad 4 – 50 dB
- iPad Air – 10 dB
