FILE 00 · Familiarization recording
download · 13 MB · 176.4 kHz · 24-bit · FLAC · 1 minute
What it is used forTo become familiar with the piece before comparing it with its quiet version.
How to play itPlay this first, starting at a low volume. If needed, carefully choose a comfortable level and note it for comparison.
What it testsNothing. It is a reference, not a measurement.
What you will hearA grand piano in a hall, for one minute. Ordinary clean sound.
What you will learnYou will become familiar with the piece. This is a listening reference, not proof that the signal path makes no changes.
FILE 01 · Quiet recording
download · 8 MB · the same piece at a very low level
What it is used forTo compare how the audio path handles a quiet musical signal: the audibility of decays, background noise and possible unwanted sounds.
How to play itAfter file 00, with processing off and a low, comfortable volume. Do not raise the level to maximum to hear the quiet signal. Check volume again before returning to normal music.
What it testsHelps reveal audible background noise and the behavior of quiet decays under specific conditions. It does not by itself determine effective bit depth or confirm bit-perfect playback.
What you will hearYou may hear the quiet piece, background hiss, or nothing distinguishable at the chosen volume. Note the result and conditions; this is not a hearing assessment.
What you will learnNo distinguishable noise — you cannot hear noise at this level under these conditions. That does not prove delivery of all 24 bits or the absence of processing. Audible noise — possible causes include the file signal, processing, bit-depth conversion, electronic noise or the surroundings. Keep the same level and test one change at a time. This result cannot identify a faulty component or determine which recordings are worth buying.
FILE 02 · Ultrasonic signal for measurement
download · 44 MB · quiet piece plus ultrasound · measurement setup only
What it is used forFor measurement-based investigation of filtering and possible distortion in the presence of ultrasound. This is not a listening test or a way to assess headphone quality.
How to play itOnly on a properly prepared setup, with headphones and speakers disconnected. Use compatible measurement equipment with known level and bandwidth limits, and set suitable attenuation in advance. If you do not have such a setup, skip this file.
What it testsHelps examine the measured output spectrum and compare modes. This file alone does not establish whether resampling occurred or where it occurred. How it works: the file includes noise above 50 kHz. Proper downsampling filters out content above the new supported band, so there may be no audible change. Audible-band components can arise from inadequate filtering or nonlinear distortion; measurements are needed to distinguish the causes.
What to observeNo listening is required. On the measurement setup, compare the spectra of 01 and 02 under identical conditions and account for the measurement input's own limitations.
What you will learnNo additional signal in the audible band — this is compatible with both an unchanged sample rate and proper resampling. Check the actual output sample rate separately. Additional components appear in the measurement — investigate filtering, levels, nonlinearity in the audio path and the measurement input itself. Software or hardware may be responsible. The Android path depends on the device, output and mode; there is no universal rule that all audio is always 48 kHz.
FILE 03 · Near-silence
download · 3 MB · no music at all · 40 seconds
What it is used forTo observe background noise while a very low-level signal is playing and compare selected modes. The file does not automatically separate hardware noise from software processing.
How to play itSet a low, comfortable volume before starting. Do not raise it abruptly to search for noise. Compare at the same level and stop playback if the sound is uncomfortable.
What it testsShows what background noise is distinguishable in your system with a nearly silent digital signal. The file is not empty: its least significant bit changes. Any observed noise needs a separate investigation to locate its source.
What you will hearEither complete silence or a steady hiss.
What you will learnQuiet — background noise is not distinguishable at these settings. This does not rule out noise or distortion with another signal or prove that every component is fault-free. Audible noise — note the level and settings. Compare one output mode or one connection while keeping other conditions unchanged. A single file cannot rule out the player, operating system or electronics.
FILE 04 · Steady tone
download · 3 MB · one steady A note · 30 seconds
What it is used forTo observe how the audio path handles a single tone and compare possible unwanted sounds or instability.
How to play itLower the volume again before starting. A steady tone may be perceived differently from quiet music. Choose a low, comfortable level and repeat comparisons at that level without abrupt changes.
What it testsProvides a simple signal for observation and measurement. Audible or measured additional components can arise at different stages, including the DAC, amplifier, transducer and measurement input.
What you will hearEither one steady note, or a note surrounded by hum, rattling or other unwanted sounds.
What you will learnJust the note — at this level, you cannot distinguish additional sounds. This is not a measurement of all distortion in the audio path. Extra sounds — note the observation and compare modes at the same level. Silence with file 03 does not rule out hardware distortion under signal. Locating the cause requires systematic checks and, where needed, measurements.
How the set came about. A forum member brought us a quiet recording that exposed a fault of our own: since August, the player had been requesting a reduced-precision DAC mode instead of the full one. It never showed up in ordinary songs. We fixed it in an evening, then built this set so such problems could be found in a minute rather than a month. Thank you, vinitsa.
