Measurement of Headphone Outputs
Introduction
For various reasons, I own a couple of very different audio / headphone outputs (“ADCs”). After one evening, where I dove too deep into the rabbit hole of audio quality, I decided I should just go and measure them.
To be absolutely clear: I have no idea what I’m doing. I have successfully ruined my ears with a combination of Rock, Techno and video games (and I would do it again). I do, however, technically own a related degree, and although I admit I kind of wriggled my way through signal theory, this should not be too hard to pull off.
To make it easy to expand this section, each device I test will get an article of its own.
Test Methods
Simple for now: 12 kHz sine wave, 0.9 amplitude (generated with Audacity). 48 kHz sample rate everywhere, so no resampling.
Labtool logic analyzer/oscilloscope with a no-name 1x test probe (not that it should matter, we’re talking audio frequencies here), 200kHz sampling, adjusted to about 50% amplitude.
No load, measured with a standard 3.5 mm “aux cable” simply clipped to the probe / ground. All devices adjusted to loud, but still comfortable headphone level using a run-of-the-mill pair of headphones.
FFT over one whole data set (around 300 ms), Hamming window, converted to decibel. Reference value is always the highest peak in the fft; for the null measurement, it’s the peak from the 12 kHz test.
Generally crowded desk, laptop charging via USB-C dock, two monitors and lots of cheap cables around. This is aimed at real world performance, not some kind of test bench values.