╔══════════════════════════════════════════════════════════════╗ ║ SPHINX CHANNEL ║ ║ Signal Verification Report - v1.0.1 ║ ╠══════════════════════════════════════════════════════════════╣ ║ Report: Parameter Verification ║ ║ Generated: 2026-08-29 05:45 UTC ║ ║ Platform: macOS - Apple Silicon - TrueRail ║ ║ Standard: -6 dBFS operating level - 64-bit internal ║ ╚══════════════════════════════════════════════════════════════╝ Sphinx Channel - GUI Parameter-to-DSP Wiring Verification Generated: 2026-08-29 05:45:09 UTC ================================================================= PART 1 - Parameter Inventory ================================================================= Counts the plugin's automatable parameters per module and verifies each module exposes its full documented control set. Total parameters found: 13 Module found expected result MAIN 1 1 OK INPUT 3 3 OK INPUT TX 3 3 OK WIDTH 2 2 OK OUTPUT 4 4 OK MAIN (module count) PASS Per-module parameter count found=1 expected=1 INPUT (module count) PASS Per-module parameter count found=3 expected=3 INPUT TX (module count) PASS Per-module parameter count found=3 expected=3 WIDTH (module count) PASS Per-module parameter count found=2 expected=2 OUTPUT (module count) PASS Per-module parameter count found=4 expected=4 -> 5 pass / 0 warn / 0 fail ================================================================= PART 2 - Poke-and-Measure (per-parameter wiring) ================================================================= Each control is moved from its default to a test value, the same calibrated signal is processed before and after, and the measured effect on the audio is compared. Verdicts: PASS - the control changed the audio as expected WARN - the audio moved in an unexpected direction FAIL - dead control or movement below the noise floor MAIN main-enable PASS Main Enable OFF disengages the chain (THD collapses to dry) H@100 idx: baseline=5.6342 test=0.0000 Δ=-5.6342 expect ≤-0.5000 INPUT input-gain PASS Input Gain +6 dB raises level Level dB: baseline=-9.1860 test=-3.1836 Δ=6.0024 expect ≥+5.0000 INPUT input-circuit-selector PASS Circuit selector - SLL→Amok produces a dramatically different fingerprint H@400 idx: baseline=0.1505 test=5.5364 Δ=5.3859 expect |Δ|≥0.5000 INPUT input-truerail PASS TrueRail B differs from A Level dB: baseline=-6.1775 test=-6.1668 Δ=0.0107 expect |Δ|≥0.0001 INPUT TX inputtransformer-model PASS Input TX model: C9049 vs M1166 - different harmonic profile H@100 idx: baseline=0.1520 test=0.1994 Δ=0.0473 expect |Δ|≥0.0030 INPUT TX inputtransformer-drive PASS Input TX drive 200% raises THD H@400 idx: baseline=0.1726 test=0.3903 Δ=0.2177 expect ≥+0.0200 INPUT TX inputtransformer-enable PASS Toggling Input TX changes chain THD H@400 idx: baseline=0.1501 test=0.1420 Δ=-0.0082 expect |Δ|≥0.0050 WIDTH width-width PASS Width -100% collapses correlation toward +1 Corr: baseline=-0.0104 test=1.0000 Δ=1.0104 expect ≥+0.0500 WIDTH width-enable PASS Enabling Width (at 100 %) moves correlation Corr: baseline=-0.0024 test=-0.6010 Δ=-0.5987 expect |Δ|≥0.0050 OUTPUT output-oversampling PASS Oversampling selector advances Level dB: baseline=-6.1913 test=-6.1851 Δ=0.0062 expect |Δ|≥0.0000 (expected inert) OUTPUT output-age PASS Age selector reaches the DSP (position 4) H@100 idx: baseline=0.8238 test=0.8011 Δ=-0.0227 expect |Δ|≥0.0100 -> 11 pass / 0 warn / 0 fail ================================================================= PART 3 - Range Verification (min / default / max) ================================================================= For representative continuous parameters, walk the full range and verify monotonic motion of the chosen metric. INPUT TX inputtransformer-drive PASS Input TX Drive H@400 idx min=0.142 def=0.232 max=0.390 Δends=0.248 WIDTH width-width PASS Width Corr min=1.000 def=-0.060 max=-0.601 Δends=1.601 -> 2 pass / 0 warn / 0 fail ================================================================= PART 4 - Discrete selector cycling ================================================================= Cycle every position of each discrete selector and verify that all positions produce DIFFERENT outputs (no dead positions, no duplicate positions). The minimum pairwise Δ requirement is set per-selector below. INPUT input-circuit-selector PASS Main circuit selector (SLL/Nevy/Amok) H@400 idx across 3 positions: [ 0.151, 0.220, 5.536 ] smallest Δ between (0,1) = 0.0686 threshold=0.0010 INPUT input-truerail PASS TrueRail tier (A/B/C) Level dB across 3 positions: [ -6.177, -6.173, -6.171 ] smallest Δ between (1,2) = 0.0015 threshold=0.0001 INPUT TX inputtransformer-model PASS Input TX model (M1166/C9049/K1166/L1544) H@100 idx across 4 positions: [ 0.213, 0.199, 0.202, 0.152 ] smallest Δ between (1,2) = 0.0029 threshold=0.0001 OUTPUT output-oversampling PASS Oversampling LED selector drives the per-module local oversamplers Level dB across 2x/4x/8x/16x: [ -6.196 -6.196 -6.192 -6.189 ] range=0.007 (active: 0 < Δ < 0.5 dB) -> 4 pass / 0 warn / 0 fail ================================================================= PART 5 - Interactions (TrueRail tier ordering) ================================================================= Cross-parameter behaviour that single-parameter pokes cannot catch. The ancestor's routing/AutoGain/dual-TX/bleed scenarios left with their modules; the tier-ordering guard remains. INPUT truerail-tier-ordering PASS TrueRail A/B/C produce distinct L/R RMS (catches dead-code tolerance scaling) L/R RMS - A=0.00009648, B=0.00002764, C=0.00011180 (all-pair deltas ≥ 1e-7) -> 1 pass / 0 warn / 0 fail ================================================================= PART 6 - Direction sanity (no inverted parameters) ================================================================= Targeted A/B tests for the most-easily inverted controls. DIRECTION width-width PASS Width -100% > Width +100% in correlation (no inversion) metric a=-0.601 b=1.000 Δ=1.601 expected b > a -> 1 pass / 0 warn / 0 fail ================================================================= PART 7 - Typed-value round-trip (double-click value entry) ================================================================= Each knob's typed-value entry is verified: typed text is parsed, limited to the control's range, applied, and read back. Invalid entries (empty, lone '-', lone '.') must leave the control unchanged. PASS = the applied value reads back within one step interval and invalid entries change nothing. TYPED reject:empty PASS Structural reject leaves parameter unchanged input '' (parsed=rejected); before=3.000 after=3.000 (expect rejected, unchanged) TYPED reject:dash PASS Structural reject leaves parameter unchanged input '-' (parsed=rejected); before=3.000 after=3.000 (expect rejected, unchanged) TYPED reject:dot PASS Structural reject leaves parameter unchanged input '.' (parsed=rejected); before=3.000 after=3.000 (expect rejected, unchanged) TYPED reject:dashdot PASS Structural reject leaves parameter unchanged input '-.' (parsed=rejected); before=3.000 after=3.000 (expect rejected, unchanged) TYPED reject:whitespace PASS Structural reject leaves parameter unchanged input ' ' (parsed=rejected); before=3.000 after=3.000 (expect rejected, unchanged) -> 5 pass / 0 warn / 0 fail [MusicMaterial] Source: reference programme material [real track] @ 44100 Hz stereo Input: peak -6.00 dBFS RMS -14.30 dBFS (crest 8.3 dB) OK Full chain SLL : peak -3.93 (d +2.07) RMS -14.67 (d -0.38) [RMS unity |d|<=1.5] OK Full chain Nevy: peak -3.36 (d +2.64) RMS -14.43 (d -0.14) [RMS unity |d|<=1.5] OK Full chain Amok: peak -4.07 (d +1.93) RMS -15.31 (d -1.02) [RMS unity |d|<=1.5] OK Input-only SLL : peak -4.47 (d +1.53) [saturation gentle, d<=+2.5] OK Input-only Nevy : peak -4.00 (d +2.00) [saturation gentle, d<=+2.5] OK Input-only Amok : peak -4.72 (d +1.28) [saturation gentle, d<=+2.5] -- Peak crest build-up SLL: +2.07 dB (phase rotation + OS reconstruction; input true-peak ~-4.85; not saturation) 6 pass / 0 fail ================================================================= TOTAL ================================================================= 35 pass / 0 warn / 0 fail ══════════════════════════════════════════════════════════════ Sphinx Channel v1.0.1 Component-Level Analog Modeling with TrueRail Technology 35 measurements - 0 failures - 0 warnings All measurements passed - signal integrity verified. analogrealism.com - support@analogrealism.com Made in Finland ══════════════════════════════════════════════════════════════