Why no mics on cables?

So, we get asked a LOT about why we do not put mics and triple buttons on our cables.  The simple fact is - because it degrades the sound quality!  Now, we've made this claim often, but decided to actually document the results formally for people to check out.

We started with a pair of cables from a single vendor: Blue Microphones.  In particular, we're using the straight wire (3 meter) and mic cable (1.5 meter) from their Ella headphone.  Why this model and cables?  Because we were actually involved in the design of all that (yeah, we're not kidding when we say we designed headphones for other people).  So we know what we're getting, and we know the cables are about the best that could be done.

The cables are pretty straight-forward: a 3.5mm jack on the phone side (TRS for straight wire, TRRS for the mic cable), and a 2.5mm TRRS on the headphone side.  On the headphone side, the tip is left, upper ring is right, lower ring is right ground, sleeve is left ground:

The measurement configuration is pretty simple: use a Samsung Galaxy Note 8 running Keuwlsoft's Function Generator to create a 0 dBFS 1 kHz tone.  We set the volume control of the Note 8 to 6 steps below full output; this represents the "warning volume" level of 179 mV RMS output (1 mW into 32 Ohms).  

We plug a test cable's 3.5mm jack into the phone, and then connect the other end of the cable to our trusty AudioPrecision APx515.  We set the AP to 192 kHz sampling rate, 192 kHz FFT size, and let it rip! 

This is the spectrum captured (20 consecutive measurements averaged together) for the straight wire (in blue; red is the other channel with the same connection style, but shorted together, this shows the noise floor of our AP):

That actually looks pretty good!  We see we're getting about 90 dB between the signal and the peaks of THD+N.  And we've got a general noise floor that is more than 100 dB down across the entire bandwidth.

ASIDE: we also see the drop in output above ~22 kHz; this is because the phone is generating a 44.1 kHz audio signal, so there is no audio information above half that frequency.  Nyquist Frequency visualized!  It also shows the aliasing filters in the Note 8 are pretty darn good as there is nothing in the spectrum above the Nyquist frequency.

Now, what happens when we do the same thing with the cable that has the mic on it?  Well, this is what we get:

WHOA!  What the HECK happened?  Well, it actually shouldn't be a surprise!  And I'll show you why.  You know that little mic and buttons on your cable?  How do you think they're connected to the rest of the cable?  Here's a quick diagram showing how they are connected:

WTF?  Seriously?  Yeah.  See the ground for all the buttons AND the microphone?  They return back to THE SAME AUDIO GROUND AS THE SPEAKERS!  So you're going to get all the current/noise from those components coupled on to the negative half of your speaker connections.  No joke!

Now, if your audio cable was really low impedance AND the amplifier had a very low output impedance, then it wouldn't be that big of a deal.  However, low impedance cables are BIG; it's common to see several Ohms between the buttons/mic and the jack in the phone/tablet.

AND, the output impedance of most headphone amps in phones is pretty high too - usually with DC blocking caps in the bottom end (which means the impedance rises in lower frequencies - just like we see the noise floor do in the measurements).

So, IF you had an ideal situation, this wouldn't be a problem.  But guess what - we don't live in an ideal world, so we have problems.  Let's look at the THD+N as a singular number, and see what we get.  Here's the THD+N (Total Harmonic Distortion plus Noise) for the straight cable:

Nice and low - 0.017%!  Yeah, we can live with that.  What about the mic cable?  It's gonna be bad, but how much worse?

Yikes!  0.082%!  Almost 5 times higher!  Not good.

Now, we know a lot of people will say "but it's still less than 0.1% - big deal!" and for them, that may be right.  But when you're at the bleeding edge of performance, and worrying about the sound differences of one DAC as compared to another, well - this should be considered.

So there you have it: empirical proof that mics and buttons on cables degrade the sound quality.  Every single time.  And since we're concerned about audio performance, well, we're gonna just not going to put a mic and buttons on the cable.  Because performance.

PRO TIP: Many people don't realize that when you have your headphones plugged into your phone, you can STILL take a call if you don't have a mic on your cord!  The mics on your phone are STILL LIVE and can be used for capturing your voice.  And the audio will play back through your headphones.  In fact, you're going to have a VASTLY improved audio experience this way (well, the person you're talking to will) because your phone inevitably has two (or more) mics for audio capture, and does not only beamsteering but ANC and AEC - resulting in a cleaner capture.  Hard to do that with a single mic on the cable!