Channel Strip TX/RX: BYPASS suppresses RN2; RN2 captured in presets (#3054). Principle VI.

[aethersdr-agent]

Summary

Fixes #3054

What was changed

Channel Strip TX/RX: BYPASS suppresses RN2; RN2 captured in presets (#3054). Principle VI.

Files modified

• src/core/AudioEngine.cpp
• src/core/AudioEngine.h
• src/core/ChannelStripPresets.cpp
• src/gui/AetherialAudioStrip.cpp
• src/gui/ClientChainApplet.cpp
• src/gui/StripTubePanel.cpp

 src/core/ChannelStripPresets.cpp | 32 ++++++++++++++++++++++++++++++++
 src/gui/AetherialAudioStrip.cpp  |  8 ++++++--
 src/gui/ClientChainApplet.cpp    |  6 ++++--
 src/gui/StripTubePanel.cpp       |  4 +++-
 6 files changed, 74 insertions(+), 5 deletions(-)

---

Generated by AetherClaude (automated agent for AetherSDR)

---

_{🤖 aethersdr-agent · cost: $21.6045 · model: claude-opus-4-7}

[aethersdr-agent]

Nice fix — the BYPASS-leaks-RN2 behavior was a real footgun and the comments tying RN2's special-case to its out-of-enum location (#2813) make the intent very clear to a future reader. TX/RX mirror each other cleanly, the contains("rn2") guard in applyPresetJson preserves backward compatibility with old preset files, and the parallel-bool snapshot pattern is the right shape given RxChainStage/TxChainStage can't hold RN2.

A couple of observations rather than blockers:

1. BYPASS now writes RN2 state to AppSettings on every toggle (TX side).
setRn2TxEnabled() calls saveAetherialTubePreampTxSettings() unconditionally. So engaging BYPASS persists "RN2 TX = off", and releasing it persists "RN2 TX = on". This matches the precedent in setStageEnabled (which calls saveClient*Settings() per stage), so it's consistent — but worth being aware that a crash mid-BYPASS leaves AppSettings out of sync with the user's intended RN2 state. The RX side doesn't have this issue because setRn2Enabled doesn't persist (a pre-existing asymmetry, not introduced here).

2. Manual RN2 toggle during BYPASS is silently overwritten on un-bypass.
If a user clicks the RN2 button in StripTubePanel while BYPASS is engaged, the next un-bypass will restore the snapshotted value and clobber their change. This is identical to the existing stage-bypass behavior, so not a regression — just worth noting that the new tooltip language ("the button stays pressed across Channel Strip profile switches") might lead users to expect RN2 toggles to "stick" too. Probably fine to leave as-is.

3. Preset apply order interaction is safe today, but fragile.
setRn2Enabled(true) disables NR2/NR4/BNR/DFNR/MNR via mutual exclusion. The preset format doesn't currently capture those NR-cluster siblings, so there's no conflict — but if someone later adds nr2/bnr/etc. to the preset schema and applies them after rn2, the last-write-wins ordering will silently disable RN2. Worth a brief comment near the RX rn2 apply block flagging that any future NR-cluster keys must apply before rn2, or after considering the mutex.

None of these block merge — the PR delivers what #3054 asks for and the code is clear. Thanks for the careful comments and the consistent tooltip updates across AetherialAudioStrip, ClientChainApplet, and StripTubePanel.

---

_{🤖 aethersdr-agent · cost: $4.1706 · model: claude-opus-4-7}