These steps are a physical verification procedure for publicly released v0.2.0 Stable. The owner reports having tested the product; the template remains useful for collecting detailed host/driver compatibility evidence.
The exact commit is in BUILD-INFO.txt; SHA256SUMS.txt lists binary hashes.
The compressor adds zero algorithmic sample latency, not zero hardware
round-trip/host/driver latency. Do not assume Maono PS22 stability from CI.
Standalone: launch Standalone/ArSonComp.exe. In Audio Settings choose
the real interface and its native ASIO driver if available. If no ASIO
driver appears, verify the vendor installer/driver first; WASAPI can be used
for comparison. Start with a headphone-safe volume.
VST3: copy the entire Plugin/ArSonComp.vst3 bundle to
C:\Program Files\Common Files\VST3\, then rescan plugins.
Test in ArSonMiX and one independent DAW that supports VST3.
Full size is 612×580, minimum 580×540, MINI height 320. Confirm the tall black-glass dynamics display, transfer curve, IN/GR/OUT meters, seven rotary knobs, search/preset browser, FILE/Save/Open, A/B and Bypass. The bottom tabs switch Dynamics / Sidechain without changing audio.
The six horizontal controls (Range, Hold, Mix, Link, HP, LP) retain the user-approved flat silver brushed-metal track and round silver thumb. M7 makes the visible thumb smaller without reducing its mouse drag area; caption and fader are slightly lower. The body is matte brushed graphite, with flat horizontal grain rather than wave-like sheen.
Hover/drag any control to see its exact value in the LCD bottom readout. Double-click any slider/knob to restore the specified factory default. Test MINI/FULL, minimum/default width, Sidechain and preset browser for readability and control collision.
- Start at 48 kHz, 128 samples, one plugin instance. Keep LISTEN SC disabled unless you are intentionally auditioning the detector.
- Select Clean mode, approximately 2:1–4:1, set Threshold for 3–6 dB gain reduction on a steady voice. Adjust Makeup conservatively to avoid downstream clipping. Compare Active/Bypass at similar loudness.
- Repeat with vocals, bass, snare/drum transients, speech and a stereo mix. Compare Clean / Feedback / Smooth at matched loudness, checking attack, recovery, pumping, sibilance, stereo coherence and unwanted distortion.
- Only after stable 128-sample operation, test 64 then 32 sample buffers. Log audio glitches and the actual host xrun count/callback spikes, not only Task Manager CPU. Repeat with 1 then 8 active instances.
A: Audio interface direct monitoring, no application processing. B: Host routing with no ArSonComp inserted. C: Plugin inserted and host-bypassed (immediate bit-transparent bypass). D: Plugin active with fixed parameters.
If A/B/C crackles, investigate driver, host routing, DPC, USB and interface before blaming compressor DSP. If only D crackles, record the digital input and processed output, current parameters, buffer, host xrun log and automation. The plugin's own Active/Bypass control is smoothed, unlike immediate host bypass.
Save/reopen a real DAW project and verify state, factory/user preset, A/B, all automation, MINI/FULL and external mono/stereo sidechain routing. Disable SC LISTEN again after checking. Run a 60-minute live audio test, open/close the editor, observe private-memory growth, actual xrun counters and any UI stalls. No physical result is declared until recorded.
Optional: right-click QA/Collect-ArSonComp-Hardware.ps1 and run with
PowerShell (or run directly from a PowerShell terminal). It writes a
read-only device inventory and a manual checklist into QA/Reports/.
It does NOT measure xruns, audio timing or hardware latency.
The included Screenshots/ are actual JUCE CI renders;
benchmark.txt and realtime-stress.txt are machine-dependent synthetic
engine timings, not real DAW latency or glitch measurements.
ASIO SDK distribution licensing must be reviewed before any public installer. The 46 factory presets, audio DSP, parameter IDs and latency are unchanged by M7.