During development at JDS Labs, we utilize Audio Precision’s APx555 and PrismSound’s dScope Series III audio analyzers to conduct professional sound quality benchmarks and characterize performance. An audio analyzer is capable of generating reference quality audio signals and measuring the resulting output of an audio device under test. This allows us to perform impulse response tests, fast-Fourier transforms (FFTs), and continuous-time analysis in order to observe specifications which directly impact audio quality. In simplest terms, an audio analyzer generates exceptional audio signals and checks to see how closely a tested device compares to its own quality.
Noise: Noise measures the presence of unexpected sound when testing for absolute silence.
Standard: <= -100 dBV
SNR: Signal-to-noise ratio specifies the amount of noise present relative to a particular test signal. We measure referenced to 1V.
Standard: <= -100 dBV
Dynamic Range: Dynamic Range is SNR referenced to the device’s maximum output level.
Standard: <= -110 dB
Output Impedance: Output impedance dictates the steadiness of an amplifier’s output voltage as the headphone load changes with frequency.
Standard: 2 ohms or < 1/8 of headphone impedance
Frequency Response: A device’s frequency response shows its output consistency across a frequency range.
Standard: +/-0.15 dB, 20-20kHz
THD+N: Total harmonic distortion plus noise indicates the presence of undesired harmonics and background noise relative to a test signal.
Standard: <= 0.005% @ 32-600 ohms, 20-20kHz
IMD CCIF: Intermodulation Distortion CCIF checks for non-harmonics under equal intensity tones, typically 19kHz + 20kHz.
Standard: <= 0.005% @ 19k+20kHz
IMD SMPTE: This standard test checks non-harmonics while generating 60Hz and 7kHz signals at a 4:1 ratio.
Standard: <= 0.005% @ -2dBFS
Crosstalk: Crosstalk, also known as channel separation, reveals how much voltage appears on a muted channel while the neighboring channel is actively driven.
Standard: <= -60 dBr
Channel Balance: Channel balance represents the audible volume difference between the Left and Right channels.
Standard: +/- 0.6 dB
Digital Jitter: Jitter reveals the presence of undesired signals due to timing inconsistency (especially relevant to DAC performance). We observe the RMS sum of modulation components using 8x averaging at 11025Hz -1dBFS @ 44.1kHz and 12,000Hz, -1dBFS @ 48kHz.
Standard: <= -100 dBFS
Output Power: Power is the amount of energy an amplifier can drive into a specific load (ohms), at a given frequency (Hz), for a specific duration, while maintaining a desired threshold of noise and distortion.
We conduct Maximum Output Power tests into a resistive load at 1kHz, keeping THD+N <= 0.005% for a continuous duration of 45+ minutes.
Peak Output Power is tested similarly, with test duration limited to 10 seconds. Please see Power is Volume for additional detail.
Standard (both): Refer to SPL Chart
THE IMPACT ON THE LISTENING EXPERIENCE
PERFECT BLACK SOUND
We approach numerical testing in the same way our ears tell our brain how something sounds. The first thing you may notice when listening to a subpar audio player is unpleasant background hiss, which is especially obvious with sensitive IEMs. So, the first test we perform on a new amplifier design is a basic SNR check.
When SNR exceeds our specifications, you hear a silent, perfectly “black” background.
BEAUTIFUL SPATIAL SOUNDS
The human brain is designed to react to spatial sounds. You can instantly turn your head towards the sound of a snapped branch or someone saying your name. The same area of the brain notices subtle changes in stereo audio imaging when listening to familiar audio tracks.
Excellent crosstalk and channel balance together produce the stereo image your mind anticipates.
We also expect our music to sound natural. Instruments and voices should sound like instruments and voices, rather than like recordings of instruments and voices. Our strict frequency response standard mandates exceptionally transparent audio.
Output impedance plays an even more important role in this regard. When output impedance is non-ideal and a headphone’s impedance fluctuates (common with dynamic headphones and balanced armature IEMs), the amplifier’s output changes with frequency.
In other words, high output impedance can lead to unnatural audio. Keeping output impedance below the recommended level is key to achieving natural sound.