Coalesce capture-engine rebuilds (CompositeCaptureBackend): a swap-triggering
source change now arms a 250ms debounce timer and re-arms on each further
change, so a flap or quick reconfiguration produces ONE rebuild to the final
state instead of a burst (Andre's 16:40 four-rebuilds-in-33s crackle). In-place
updates still apply immediately; a pending rebuild whose target flaps back is
cancelled.
Auto-tune (TickRoute) now keys off the SECOND-highest arrival-gap/render-gap
second in the lookback window instead of the single worst, so a lone ~1s
OS/driver stall no longer balloons the buffer to the 200ms cap (the 16:51
trim burst); sustained jitter still reacts at full speed. Logs both gap-max
(true peak) and gap-used (value acted on).
Mic-privacy detector widened: also catches a per-app Deny aimed at this exe
under ConsentStore\microphone\NonPackaged\<exe>, the HKLM NonPackaged gate,
and the Group-Policy/MDM force-deny (AppPrivacy LetAppsAccessMicrophone=2) —
the block shapes that silence WASAPI capture while ASIO sails past, and that
the old three-value check missed.
Forensic instrumentation so the next log proves what happened: capPeak=
(loudest pre-encode sample, per lane, on the diag line), mic-privacy verdict
logged at startup, ui: capture tick/untick events, and device-event: lines for
Windows endpoint changes.
Docs: mic-privacy + auto-tune sections updated in readme.html, MANUAL.md
regenerated, About-box changelog and RELEASE_NOTES for v3.7.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Replace the generated cmd.exe + robocopy update helper — which silently
failed on some machines — with an in-app C# installer:
* Stage the new version to a per-user temp folder off the install and run
the new RemSound.exe from there, so nothing in the install is locked by
the updater itself.
* Wait for the old process to fully exit (real WaitForExit), then
back-up-and-swap files in C# with retry + rename-aside; roll the install
back to the previous version on any failure, so a failed update can never
leave a half-installed RemSound.
* Log every step to updater.log; clean up old stages and legacy batch
artefacts on launch. Removed the old BuildInstallScript batch generator.
Route the "RemSound is already running" dialog and its follow-up message
through ForegroundDialog so they surface in front from a background-relaunched
copy, matching the earlier post-update fix.
Docs: rewrite the readme update sections for the new mechanism, regenerate
MANUAL.md, refresh the About-box changelog and RELEASE_NOTES.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Recover an unplugged/replugged output sound card automatically (issue #5):
detect the dead WASAPI device and remember the receive-output selection so it
re-ticks and re-opens when the card returns.
- Adaptive per-card WASAPI buffer target sized to each card's pull chunk, held
stable so it never flits about under CPU/network load.
- Consolidate all per-user data (config, profiles, logs, sounds) into one
"user settings and logs" folder; migrate every older layout; exclude it from
the updater so custom cue sounds now survive updates.
- Mic-privacy detector: warn once when a Windows-blocked mic is switched on, or a
profile loads with one already on.
- All warning/notice dialogs now come to the foreground even when minimised.
- Apply volume + mute on profile load (were saved but not restored).
- Crash-safe (atomic) profile/config saves.
- Fix two resource leaks (push-mode capture MMDevice; UPnP DeviceFound handler).
- Remove dead code (baseline-diff machinery, dead ASIO probes, no-op stubs).
- Docs: readme.html, MANUAL.md, About-box changelog and release notes for v3.5.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Builds on the v3.4 freeze (dd70613) with the fixes and tuning from live testing,
plus the v3.4 documentation pass.
Audio (receiver):
- Per-device WASAPI drift correction in MultiOutputPlayout. A pull-side resampler
(mirroring SessionPlayout's proven corrector) holds each output device's buffer at
a fixed low depth, cancelling the slow clock drift that made WASAPI peers "lag
apart" over long sessions. Feed-forward clock-ratio measurement plus a gentle
depth-restoring term; the first measurement window is discarded because WASAPI
start-up priming poisons it. ASIO already self-corrected; this brings WASAPI level.
- Output device buffer requested at 5 ms (WASAPI clamps it up to the device's minimum
period, ~10 ms) instead of 15 ms, since the corrector keeps it fed — a free saving.
UI / accessibility (MainForm, Program):
- Startup notices (what's-new About box, Realtek warning) now run one at a time via a
single sequence instead of separate BeginInvokes, so they no longer stack into
nested modals that couldn't be closed. The loading splash is skipped for a
tray-bound quick switch.
- Quick profile switch keeps RemSound in the tray if it was there, and plays the
switch cue immediately on click.
- Profile-switch cue now plays on click for every switch path (recent menu, quick
switch, File > Open) and no longer on a fresh start into the first profile. It was
also previously dead on the rebuilt form (pendingProfile was nulled first).
- Realtek ASIO toggle's accessible name now reads "Enable"/"Disable" to match the
visible text, instead of "Toggle" (screen reader read the wrong word).
Docs (plain English):
- RELEASE_NOTES.md: v3.4 entry.
- About dialog: v3.4 "what's new".
- readme.html (the canonical bundled manual): quick switch, the hotkey read-outs, the
new profile-menu-open cue, Realtek auto-detect/disable, and the config-folder path.
- MANUAL.md regenerated from readme.html via sync-manual.py so the two stay in sync.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Freezes the v3.4 feature set (everything since the v3.3 public release):
- Fix the receiver handle leak: the 3-second device-refresh timer reopened the
configured ASIO driver every tick, and Realtek's ASIO driver leaks Event+Mutant
handles on every open. Cache the ASIO probe per driver so it is opened once.
- Realtek ASIO block: detect a Realtek ASIO driver, offer once to disable it, and
never touch it again if disabled; Options-menu toggle to reverse. Global config.
- Device hot-plug is event-driven (AudioDeviceChangeNotifier) instead of a 3s poll;
debounced refresh, falls back to polling if registration fails.
- Quick profile switch: new global hotkey opens an NVDA-friendly popup of all
profiles (current marked); Enter/click switches; plays a new "profile menu open"
cue with Preferences mute + custom-sound.
- Announce assigned global hotkeys on the controls/menu items they drive (NVDA reads
"press X anywhere"). File > Open already had Ctrl+O.
- Held-back changes folded in: config-folder migration, codec-column fix, Tailscale
endpoint network-prune, empty-password guard, and the handle-leak diagnostics
(ProcessSelfMeter, HandleTypeProbe).
Version bumped to 3.4.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Points users to Aryan Choudhary's community Android receiver (RemSoundAndroid) Releases page for the signed APK. Third-party companion app, TalkBack-tuned, not maintained here.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Headline: all audio is now encrypted (AES-256-GCM), keyed by a per-profile
password. Mandatory — v3.3 only interoperates with v3.3+.
Encryption
- RemSoundCrypto (Core): PBKDF2 key derivation, AES-GCM encrypt/decrypt
(low-alloc, into-span), password fingerprint, light on-disk obfuscation.
- Wire: SenderLane encrypts the audio payload (PCM split across parts when the
+28 overhead crosses MTU); AudioReceiver/StreamSession decrypt via a shared
single-thread AudioDecryptor. Fingerprint piggybacks on the Format packet
(offset 36, backward-compatible) so a peer can detect a password mismatch.
- Profile.Password (scrambled), carried through BuildCurrentProfile; MainForm
derives + pushes the key/fingerprint to sender + receiver (RecomputeAudioCrypto).
- UX: ask-for-password on profile create; File -> Change this profile's password
(ProfilePasswordDialog); Options -> Profile passwords (manager); a gate that
prompts before streaming without a password; and a clear "passwords don't
match" / "peer needs to update" message driven by the fingerprint.
Cue fixes
- CuePlayer (NAudio) replaces System.Media.SoundPlayer, which silently failed
on the 96 kHz/24-bit cue WAVs (and any custom file) — cues now play reliably,
resampled to 48 kHz/16-bit. Also fixes the Preferences preview button.
- Connect/disconnect cues now audio-gated with hysteresis: connected when audio
flows OR heartbeat healthy; lost only when audio stops AND heartbeat
unreachable. Kills false disconnects and the receive-only "no cues" case.
- Honest cue logging (played / muted / not loaded).
Smaller
- Endpoint stickiness: keep the audio target pinned to the heartbeat-proven
address instead of chasing a multi-homed peer's other (unreachable) address.
- "Online/offline" label now audio+heartbeat aware, not discovery-only.
- "Show what's new after each update" preference (on by default).
Docs: About v3.3 block, RELEASE_NOTES, README (encryption as a headline),
manual section 12 "Passwords and encryption" (+ renumber), MANUAL.md regenerated.
Version 3.2.0 -> 3.3.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- New "Update sound" cue (CueId.Update / update.wav). Plays just before an
update starts installing, on every path (manual, background-silent,
startup-silent), so a silent background update gives an audible heads-up
before RemSound closes to restart. Per-profile mute + custom-sound override
in Preferences, same infrastructure as the other cues:
* Profile.EnableUpdateCue + RemSoundSettingsStore Load/Save + round-trip
* MainForm updateSound field, TryLoadCueSound, play in InstallUpdateAsync
gated by LoadEnableUpdateCue
* PreferencesDialog "Update sound" CueRow + ResolveCueFilePath mapping
* update.wav shipped in sounds\ via csproj Content
- Single-instance "switch to the running copy" now actually brings the window
to the front. The second copy grants the running copy foreground rights via
AllowSetForegroundWindow before signalling, and lingers briefly so it can
raise itself before we exit — without this, Windows' foreground lock left
the running window only flashing in the taskbar (Ed's report).
- Docs: About box v3.2 block, RELEASE_NOTES.md, manual cue section (now seven
cues, with the update cue described), MANUAL.md regenerated.
- Version bumped 3.1.3 -> 3.2.0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Fixes the chained-fault runaway Andre hit after an update (multiple copies
stacking, audio climbing to deafening, terminal kill to recover).
- Single-instance lock (SingleInstanceCoordinator + SingleInstanceDialog,
wired in Program.Main). A named mutex makes two copies impossible. A second
launch offers: switch to the running copy (default; surfaces it from the
tray via a named activation event), or force the running copy closed and
start fresh (Process.Kill, retried elevated if the target is elevated).
This is the structural fix that makes the stacking runaway impossible.
- Prompt-free update exit. InstallUpdateAsync sets updatingInProgress before
Application.Exit(); the close path's skipPrompt now honours it, so no
unsaved-changes dialog (whose default button is Cancel) can abort the
update's restart.
- Read-only persistence fix. BuildCurrentProfile now carries
currentProfileReadOnly into the saved snapshot. Previously a deliberate
save of a locked profile wrote ReadOnly=false, silently unlocking it on
disk — which re-armed the save prompt that then blocked the update.
- In-process double-install guard. updateInstallStarted stops the ~4 s
startup check and the background poll both staging an install + helper.
- Docs: About box, RELEASE_NOTES.md, readme.html + MANUAL.md ("Only one copy
of RemSound runs at a time").
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
- Tray icon now re-registers itself (hide+re-show) whenever the meaningful
state changes — peer connect/drop, send/receive toggle, recording
start/stop — so the name a screen reader announces stays in step with
the live state. Fixes the persistent "no peers" then "1 peer" double
announcement that the old "show window then minimise again" trick used
to clear by hand. Same root cause as the v3.1.1 stuck-tooltip fix, just
exposed when the state changes a moment after the icon appears.
- Recording shows as a plain "recording" flag in the tray rather than a
live timer. A ticking timer would have either flickered the icon once a
second or left a screen reader announcing a stale time next to the live
one. Removed the now-dead FormatRecordingElapsed helper.
- Bundles the earlier Win7 updater fix (6cbde0d): a failed secure
connection is no longer mislabelled as "you're up to date".
- About box, RELEASE_NOTES.md, readme.html and MANUAL.md updated to match.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Tech Singer's Windows 7 log from 2026-05-28 had this pattern:
updater: GET https://api.github.com/repos/...
updater: check failed: HttpRequestException: The SSL connection could not be established
...
updater: startup check — up to date (v3.0.1)
i.e. the SSL handshake to GitHub failed (Win7's TLS stack missing
KB3140245 / KB4474419) but the updater logged 'up to date' and the
user-facing message in CheckForUpdatesManually said the same. So a
user with a broken update check has no way to distinguish that from
genuinely having the latest version.
Fix:
* CheckForUpdateAsync now returns a discriminated UpdateCheckResult
(UpdateAvailable / UpToDate / UpdateCheckFailed) instead of the
old UpdateInfo?. Each return-null site is replaced with the
appropriate concrete type.
* The catch-all 'try { ... } catch (Exception ex) { return null; }'
becomes 'return new UpdateCheckFailed(ClassifyFailure(ex), ...)'.
ClassifyFailure walks the exception chain and maps to a coarse
FailureKind enum: SecureConnection (TLS/auth), Timeout, HttpError,
NetworkUnreachable. SecureConnection is broken out separately so
the manual-check UI can point Win7 users at the specific Microsoft
KBs that fix the issue.
* MainForm.CheckForUpdatesManually pattern-matches on the result:
UpToDate -> existing 'you're running the latest' message;
UpdateAvailable -> existing install confirmation;
UpdateCheckFailed -> NEW dialog (ShowUpdateCheckFailedDialog)
whose wording is tailored to the FailureKind. The
SecureConnection branch explicitly names KB3140245 and KB4474419
and offers the manual zip-install URL as a fallback. All branches
keep the technical detail out of the dialog and route it to the
log instead.
* Background and startup polls stay silent on UpToDate and
UpdateCheckFailed (no point nagging the user about something
they can't act on from a timer tick), but the startup-poll log
now records the failure kind and detail instead of mislabelling
the outcome as 'up to date'.
No version bump - this rides along with the next feature release
(planned v3.2 with the Reaper ReaStream integration). The bug is
silent on the affected users today, and shipping a v3.1.2 just for
the error-message improvement would mean another update cycle for
everyone for marginal benefit.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bug: if v3.1 installed via auto-update on a profile with StartMinimised
on, the tray icon's hover tooltip got stuck at the initial "RemSound -
starting up" string. The snapshot tick was running and SetTooltip was
being called every second with the live state, but Windows shell kept
showing the original cached text on hover. The fix-by-workaround was a
hide-then-re-show cycle which forced the shell to rebuild the icon
registration with the latest NotifyIcon.Text.
Root cause: NotifyIcon.Text values set BEFORE the icon's first
NIM_ADD (i.e. while Visible=false) become the shell's "initial"
tooltip when the icon eventually appears. Subsequent NIM_MODIFY calls
from text changes DO propagate, but the shell tends to keep showing
the original text on hover - presumably a tooltip-cache eviction
quirk. In the resume-after-update-with-StartMinimised flow, the
window briefly shows then BeginInvokes a Minimize that flips
Visible=true before the snapshot timer has had a chance to fire, so
the shell registers with the stale "starting up" string.
Fix: drop the hard-coded "starting up" initial text from the
controller ctor entirely. The controller now takes a Func<string>
buildTooltip callback from MainForm and calls it in Minimize() right
before flipping Visible=true, so the shell's NIM_ADD sees current
live state instead of a stale string. The 1 Hz snapshot tick keeps
working for ongoing live updates while the icon is visible.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Two Windows-specific gotchas the previous code didn't handle:
1) The 'python' / 'python3' commands on most Windows installs are
Microsoft Store execution aliases. They appear on PATH, accept
any invocation, exit with code 9009, and print a "go install
from the Store" message instead of running the script.
2) The 'py' launcher accepts --version and returns the right thing
(it knows about registered Pythons via the registry), but on
some setups it refuses to run scripts and falls through to the
Store alias too. Seen here: 'py --version' prints 3.11.9 but
'py sync-manual.py' prints the Store message and exits 9009.
The new approach runs an actual sentinel script via -c with each
candidate and only accepts the candidate when stdout matches the
expected string. User-local install paths come first because they
skip the Store-alias issue entirely. Falls through to 'py' and the
PATH commands as backstops.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Two new cues join the existing connect / disconnect / record-start /
record-stop set:
* profile-save cue (sounds\save.wav by default) fires in
SaveProfileTo after a successful Save or Save As. Honours the
Profile.EnableSaveCue per-profile flag.
* profile-switch cue (sounds\profile.wav by default) fires in the
MainForm Shown handler after a profile finishes loading - covers
both startup-with-profile and mid-session profile switches.
Honours the Profile.EnableProfileSwitchCue flag. Because cue
loading runs AFTER settings.ApplyProfile in the ctor (line 542),
the profile being entered determines which sound plays - not the
one being left, exactly as Ed asked.
The audio cue UI in PreferencesDialog stays as a CheckedListBox (up /
down navigates, Space toggles) with TWO action buttons below that
operate on whichever cue is selected:
* Play [cue name] (Alt+P) - previews via SoundPlayer.Play on the
resolved path (custom override if set, default in sounds\
otherwise). Independent of the tick state.
* Browse for [cue name]... (Alt+B) - opens a WAV picker. If the
user picks a file inside RemSound's own sounds\ folder, it's
treated as "use default" and the override is cleared - avoids
pinning the user to a specific shipped default that a future
release might replace. Right-click "Use default sound" reverts.
Custom cue paths AND enable flags are per-profile. Lives on
Profile.CustomCuePaths (Dictionary<string,string>) and the per-cue
EnableXxxCue bool? properties. Cache mirror in
RemSoundSettingsStore.Settings.
All default WAVs moved from the install-root flat layout into a
sounds\ subfolder (csproj Content rules updated). save.wav and
profile.wav are bundled defaults.
Tray menu rewritten (MainFormTrayController) per Ed's spec:
* Show RemSound (W) - now uses Win32 SetForegroundWindow after the
standard Activate() because WinForms Activate is blocked by the
foreground-lock when invoked from a tray-menu click, which left
screen-reader users having to Alt+Tab to reach the restored
window.
* Enable sending (S) / Enable receiving (R) - tickable, reflect
current state, TOGGLE rather than always-on.
* Profiles (P) - submenu populated from AppConfig.RecentProfiles
with the same &1..&5 mnemonics the File menu uses. Pre-populated
once at construction so WinForms recognises it as a submenu and
fires DropDownOpening - originally I relied entirely on the
open event, which the framework skipped for items with no
DropDownItems, producing the "Profiles does nothing" bug.
* Exit (X).
Tray tooltip now built dynamically from snapshot tick (1 Hz):
"RemSound - [recording for MM:SS,] N peer(s), sending (lane),
receiving (lane)". Recording timer only included while
RecordingController.IsRecording is true (added
RecordingStartedUtc accessor for the elapsed calculation). Lane is
derived from which device-list ticks are active, not just the audio-
mode setting, so a BothIndependent user with only WASAPI inputs ticked
honestly reads as "sending (WASAPI)".
Fixes:
* Initial tooltip "RemSound" produced a "RemSound RemSound" read on
NVDA because the process name and tooltip matched. Set to
"RemSound - starting up" so the duplicate disappears.
* Recent profile menu items no longer carry a "Recent profile N:"
AccessibleName prefix in either the tray submenu or the File
menu's Recent profiles - now just the profile name. Number-key
mnemonics (&1..&5) untouched.
Manual (readme.html) updated: new section 17 "Audio cue sounds"
documents all six cues, the Play/Browse buttons, the right-click
"Use default sound", and the per-profile semantics. Sections 17-21
renumbered to 18-22. New "System tray icon and its menu" subsection
inside section 4 documents the redesigned right-click menu and the
hover tooltip. MANUAL.md regenerated via sync-manual.py. About box
gets a v3.1 block at the top. RELEASE_NOTES.md fully rewritten for
v3.1.
No wire format change - v3.1 talks to other v3.0.x machines exactly
as before.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
People landing on the repo page were having to install RemSound just to read what
it does and how to use it. Two doc changes fix that:
1) README.md rewritten as a plain-English landing page. Drops the developer-focused
highlights / build-from-source / project-layout sections in favour of what
RemSound is, who it's for, how to install it, and a prominent link to the manual.
No jargon, no command lines, no NuGet / SDK / ASIO-protocol talk. The dev-side
information that used to live here (build commands, source layout, relay setup)
is still discoverable for anyone who wants it — the source itself is on the same
page, and the relay docs are under server/README.md.
2) MANUAL.md added at the repo root as the GitHub-rendered version of the F1 help.
Markdown derived directly from readme.html via sync-manual.py (new), so visitors
can read the manual inline on the repo page with no download. readme.html stays
exactly where it was (bundled inside RemSound, opened by F1) — it remains the
canonical source of the manual content; MANUAL.md is auto-generated from it.
The sync-manual.py script is invoked automatically from build-release.ps1 as step 0,
before any other release work. It regenerates MANUAL.md from readme.html and then
checks `git diff` on MANUAL.md — if the file changed, the release is paused with a
message asking the user to commit the updated MANUAL.md alongside the release commit.
That makes it structurally impossible to ship a release with a stale GitHub-facing
manual: forgetting to commit MANUAL.md after editing the bundled help triggers a
deliberate release-time stop.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bug: Andre reported audio latency feeling laggier after long sessions
on his Win10 desktop receiving Opus from his laptop. His 23-hour log
showed working set climbing from 83 MB at startup to 3.5 GB at the
end, with the managed heap staying tiny (~5-7 MB) the whole time.
CPU climbed alongside (from steady-state ~7% mid-session to peaks of
~60% by the end) and audio threads ended up doing ~4x the work they
did at the start. Andre's perception of latency drift was the CPU
pressure showing up in audio scheduling, not the buffer itself
growing (bufAvg stayed roughly stable at 25-28 ms).
Root cause: Concentus.Native (introduced in v2.2 / shipped in v3.0)
returns concrete NativeOpusDecoder / NativeOpusEncoder objects that
implement IDisposable and own native libopus state. Three call sites
were taking the IOpusDecoder / IOpusEncoder interface reference and
never calling Dispose:
* StreamSession.Dispose — comment literally said "IOpusDecoder has
no Dispose; nothing else to free", which was correct for the
pure-managed Concentus.OpusDecoder pre-v2.2 but stopped being
correct the moment we added the native binding
* OpusEncoderState.Dispose — same misleading comment, same bug
* SenderLane.OnCodecChanged — overwrote the existing encoder field
without disposing the old instance on codec change
Compounding factor: Program.Main sets GCSettings.LatencyMode =
GCLatencyMode.SustainedLowLatency to keep audio scheduling smooth
(it suppresses gen2 collections). That's correct for the hot path
but it ALSO suppresses the finalizer pass that would have released
the leaked native handles as a backstop. Because the managed heap
stayed tiny, the GC never saw enough pressure to force a gen2 pass
on its own, and the native state piled up indefinitely. Multi-output
receive multiplied the per-output growth.
Fix is in two parts:
1. Call (... as IDisposable)?.Dispose() at every release point —
StreamSession.Dispose, OpusEncoderState.Dispose,
SenderLane.OnCodecChanged before overwrite, AudioRecorder's
Concentus.Oggfile-backed OpusOggFileWriter.Dispose. The
as-IDisposable cast handles both the native and the pure-managed
path transparently (managed-only IOpusDecoder isn't IDisposable;
the as-cast yields null and the null-conditional is a no-op).
2. Periodic native-memory reaper in MainForm.SnapshotLogIfDue — once
every 300 snapshot ticks (~5 min), run
GC.Collect(2, Optimized, blocking, !compacting) +
WaitForPendingFinalizers on a background Task.Run so the gen2
work doesn't hitch the UI thread. Audio threads are separate and
unaffected. Serves as belt-and-braces for any future code path we
forget to wire and for cleaning up any per-call native scratch
the underlying library might accumulate that isn't owned by a
single .NET wrapper.
Expected behaviour after fix: working set settles around 100-200 MB
on a typical receive session and holds roughly flat for as long as
the app stays running. CPU stays at its early-session baseline
across multi-hour sessions. Andre's "latency drift" symptom should
disappear.
Wire format unchanged; same codec list, same UI, same defaults.
v3.0.2 talks to other v3.0.x peers exactly as v3.0 / v3.0.1 do.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The script's ValidatePattern was '^v[0-9]+\.[0-9]+$', which rejected v3.0.1
because the v3.0 line predated any hot-fix releases. Widened to accept an
optional third component so v3.0.1 (and future patch releases) build via
the same path as v3.0 / v2.2 / etc.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bug: ticking the "Automatically open my router for incoming connections
(UPnP)" box in Preferences could freeze the WinForms message pump until
Mono.Nat's NatUtility.StartDiscovery() returned. On Andre's setup it never
did — audio kept flowing (audio threads are independent of the UI thread)
but the window stopped repainting, the system-tray hotkey stopped
responding, and the only way out was Task Manager.
Pre-existing latent bug in the v2.1 UPnP code; we just shipped without
anyone exercising the path on a problematic network (multiple adapters /
VPN / SSDP-swallowing router).
Fix: three call sites moved off the UI thread via Task.Run -
* MainForm OnShown (startup re-enable from saved AppConfig.UpnpEnabled)
* MainForm Preferences applyUpnpEnabled callback (user ticks the box)
* MainForm power-resume handler (Refresh() after sleep/wake)
RouterPortMapper.Start() returns "immediately" only when StartDiscovery
returns quickly; on a slow network it can block synchronously for many
seconds. Same is true of Stop()'s socket teardown and Refresh()'s
teardown-then-restart sequence. All three are now safely backgrounded.
StatusChanged is unaffected - it already fires on the mapper's own thread
and the PreferencesDialog handler BeginInvokes back to the UI thread.
Live status label updates correctly during the new background discovery.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
* New "Opus, live latency" codec mode: 2.5 ms frames (120 samples/ch at 48 kHz)
via the float-input encode path. End-to-end codec delay drops to ~5 ms (vs
~12.5 ms at standard 10 ms Opus). Test on LAN: 400 pps/lane, zero missed /
reordered / duplicate packets, ~15 ms one-way saved end-to-end.
* Wire-format change: AudioFormatInfo.FrameDurationMilliseconds renamed to
FrameSamplesPerChannel (int sample-count at announced sample rate). Removes
the lossy 48000*ms/1000 conversion that couldn't represent 2.5 ms. v3 <-> v3
exact; v3 <-> v2 still passes audio (Opus decoder is self-describing from
packet TOC) but v2 side over-sizes its buffer wildly. v2.x profiles auto-
migrate via <120 sentinel rule in RemSoundSettingsStore (anything below 120
is treated as legacy ms and multiplied by 48). Profile JSON key kept as
OpusFrameMilliseconds via [JsonPropertyName] so old profile files still load.
* Codec dropdown rebuilt with use-case names: "PCM 48K 24 bit - uncompressed",
"Opus, broadcast quality - loss tolerant", "Opus, live latency - for jamming
and monitoring". Middle 10 ms option retired; saved 480-sample profiles
collapse to broadcast quality (safer-side default).
* Profile auto-resume after self-update: RemSoundUpdater writes a one-shot
_resume-after-update.txt sentinel containing the active profile title before
exit; Program.Main reads + deletes it on next start and silently loads that
profile, skipping the picker. Helper batch's robocopy /XF excludes the
sentinel and the failure-branch cleans it up if the install aborts. Falls
through to normal startup behaviour (StartWithProfileTitle or picker) if the
sentinel is missing, empty, or names a profile that no longer exists.
* Read-only profile saves now go through on explicit Ctrl+S / File -> Save
with a one-time TaskDialog warning ("Save anyway" / "Cancel" + Do-not-show-
again). Lock continues to suppress the automatic unsaved-changes prompt on
close / profile switch (its main job). AppConfig.SaveOnReadOnlyMessageSuppressed
renamed to SaveOnReadOnlyWarningSuppressed; v2.x suppression flag is silently
discarded since the behaviour changed and the user needs to see the warning
once on each machine.
* Manual (readme.html) updated: codec table rewritten with the three new
choices and corrected bandwidth figures, send-rate description updated, new
sections "The same profile picks up automatically after an update" and
"Saving on purpose while a profile is locked".
* Subsumes the never-separately-released v2.2 work: native Opus encoder
(~97% less per-second memory churn on Opus send path via Concentus.Native),
efficiency tidy-ups (item 4 ASIO probe rate, item 6 WaitHandle, item 7
snapshot cache, items 14/16 heartbeat + discovery), legacy cleanup
(items 30/34/35/36: KeepAlive infrastructure, drift drop/repeat/accumulator
fields, fan-out cache stat). New diagnostic columns cpu/memMB/wsMB/
allocKBps/captureMs/sendMs/recvMs/renderMs gated on Enable-logs.
About dialog updated with v3.0 block at top; v2.2 block retained for the
subsumed work.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Single biggest change: added the Concentus.Native NuGet package. Concentus
2.0+ auto-detects native libopus at runtime and routes encode calls
through it; encoder state lives on the C side and is reused across calls
rather than `new`ing ~15 working buffers per call (Concentus issue #22,
open since 2018). Measured on the desktop test at 15:36:55 — Opus 10 ms
allocation rate dropped from 4,625 KB/s to 108 KB/s, a 97.7% reduction.
Process CPU dropped from 4.7% to 1.6% in the same config. Audio is bit-
for-bit identical (it's literally the same encoder, just better
packaged). `OpusEncoderState.cs` itself unchanged on the call site.
Diagnostic / measurement layer (gated on Enable-logs, zero cost when off):
* ProcessSelfMeter: CPU%, managed heap MB, working set MB, allocation
rate per second, GC counts per generation
* Per-thread work-time counters: captureMs / sendMs / recvMs / renderMs
expressed as milliseconds of CPU consumed by each audio thread per
second
* Inter-packet arrival gap measured at the user-space UDP socket
(rxNetGapMs) — pinpoints whether arrival jitter is in the network or
our own dispatch path
Small efficiency wins (each one was small but cumulative):
* deviceRefreshTimer interval 1s -> 3s (item 4)
* WaitHandle array allocations eliminated in MixingEngine.MixLoop and
MultiOutputPlayout.ProduceLoop (item 6)
* MultiOutputPlayout caches its output-buffer snapshot and only rebuilds
on SetOutputDevices, instead of rebuilding every 10 ms (item 7)
* HeartbeatService reuses an outbound ping byte[] instead of allocating
per send (item 14)
* PeerDiscoveryService caches broadcast addresses and invalidates on
Windows' NetworkChange event instead of walking all NICs every 1.5 s
(item 16)
Legacy / dead-code removal:
* KeepAlive packet's implementation (struct, enums, writer, reader, size
constant) — all dead since HeartbeatService landed 2026-05-06. Kept
the RemPacketType.KeepAlive enum value and silent-drop dispatch for
wire compat with any pre-2026-05-06 build still in the wild (item 30)
* driftDropFramesTotal / driftRepeatFramesTotal fields and accessors —
Phase-2 splice corrector relics, never incremented since Phase-4
resampler design landed; backed five always-zero diag log columns
(items 34 + 35)
* DriftAccumulator (always returned 0) — same shape, removed alongside
the driftAcc= column (item 35)
* TakeMaxFanOutCacheBytes / Ms + fanCacheMs column — FanOutSource was
retired in May (item 36)
Project documentation:
* RemSoundefficiency.md added as the canonical record of the efficiency
analysis, every item's status, and the measured wins from this round
* Honest item-by-item review of the original 50-item list — several
items I had sized optimistically in the original analysis turned out
to be already-done (item 20), already-optimal (item 22), or below
the meter floor (items 9, 15, 17, 25). Recorded so future passes
don't re-investigate.
Wire format and audio pipeline unchanged from v1.5 onward — v1.5 through
v2.2 peers interoperate.
Headline features:
* Automatic router port opening (UPnP / NAT-PMP / PCP). Opt-in via
Preferences; surfaces external address + carrier-grade NAT detection.
* Lock profile (read-only). New File-menu tick that makes a profile
load-only — session changes don't persist, no save prompt on close.
Unblocks unattended shutdowns (NVDA gone, remote dropped, hibernate)
where the existing save prompt could deadlock.
* Check for updates on startup (default on) + brief countdown notice
before silent updates install, so a launch-time update doesn't make
the app silently vanish.
* "Cue sounds" -> "Audio cue sounds" label clarification.
Bug fixes:
* No sound after the computer wakes from sleep. PowerResumeHandler
rebuilds the audio backend automatically on resume; brief
"Reconnecting to audio driver" splash during the rebuild.
* Receiver audio silent after waking from hibernate. RefreshAudioDeviceLists
now treats a transient ASIO probe failure (returns -1/-1 because the
driver is mid-teardown / mid-reinit) as "retry next tick" instead of
clearing the user's tick selection.
Diagnostic-only changes (gated on the existing Enable-logs checkbox,
zero cost when off):
* AudioStepProbe split into cross-buffer vs within-buffer maxes so log
inspection can tell a real-content sharp transient apart from a
pipeline-boundary glitch. Plumbed through every probe owner.
* New rxNetGapMs + gc0/gc1/gc2 delta columns in the diag log to split
receive-side jitter into network-layer vs managed-runtime causes.
Files touched: RELEASE_NOTES.md + readme.html + 24 source files across
RemSound.Core / RemSound.Sender / RemSound.Receiver / RemSound.App.
Three new app files: PowerResumeHandler, RouterPortMapper,
UpdateInstallNoticeDialog.
Wire format and audio pipeline unchanged from v1.5 onward — v1.5
through v2.1 peers interoperate.
Opening an ASIO driver takes 1-3 seconds, synchronously, during MainForm
construction — confirmed in logs as a ~2.8s dead gap on an ASIO-profile
launch (a WASAPI-only launch is ~131ms for the same stretch). During that
gap the main window is blank / "Not Responding" and looks hung.
New AsioLoadingSplash shows a small "Loading audio driver, please wait..."
window on its OWN dedicated STA thread with its own message loop, so it
stays painted while the main thread is blocked opening the driver.
Program.cs starts it before new MainForm() and dismisses it after.
Deliberately, the ASIO driver open stays on the main UI thread — that
STA/message-pump thread is what ASIO/COM drivers are most compatible with,
and moving the open off it risks breaking drivers that can't be tested.
Only the cosmetic splash moved to a side thread; no ASIO/driver code is
touched. Splash shows only for profiles with an ASIO driver selected;
WASAPI-only profiles build fast and get no splash.
Version bumped to 2.0.0 to mark the milestone.
No wire-format or audio-pipeline changes — v1.5 through v2.0 interoperate.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The install helper's robocopy line was:
robocopy "{stagingRoot}" "{installDir}" ...
installDir is AppContext.BaseDirectory, which always ends in a directory
separator, so the destination argument was a quoted path ending in a
backslash: "D:\...\publish\". Windows command-line parsing reads the \"
as an escaped quote, so robocopy never received a valid destination,
rejected the command line, and exited 16 (usage error, nothing copied)
instantly. The auto-updater has never worked in any release because of
this — v1.0-v1.2 failed silently, v1.3+ detected the failure and wrote
update-failed.txt but never fixed the robocopy line.
Fix: BuildInstallScript now strips trailing separators —
stagingArg = stagingRoot.TrimEnd('\','/'), installArg likewise — and
the robocopy line uses the trimmed forms. Verified by running the
corrected robocopy against real staged files: exit 3 (success), files
copied.
The broken helper is baked into every shipped build including v1.8, and
the helper is generated by the running version — so v1.8 and earlier
cannot auto-install v1.9. v1.9 must be installed by hand once; from v1.9
onward the updater works.
No wire-format or audio-pipeline changes — v1.5 through v1.9 interoperate.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Updater (RemSoundUpdater.cs), all from Andre's feedback:
* CheckForUpdateAsync now requests /releases?per_page=100 instead of the
default 30-item page, so a burst of server-vX.Y relay releases can't
push the newest client release off page 1.
* The install helper's robocopy now excludes remsound.config.json and the
logs / profiles / recordings folders — an update replaces app files
only and can never overwrite the user's own config or data.
* On a successful update the helper now also deletes _update-helper.log
and any stale update-failed.txt (the _update folder was already
removed), leaving a tidy install folder. The failure branch still keeps
them all for diagnosis.
* update-failed.txt rewritten as plain user-facing instructions: numbered
steps, no brand names, names the real _update folder, no robocopy
jargon. The exit code now goes to _update-helper.log only.
Manual (readme.html): rewritten in plain language — developer jargon
removed or explained in everyday terms — and a second pass removed the
keystroke-navigation choreography and screen-reader narration from the
prose. The Keyboard shortcuts section and all shortcut reference are
kept intact.
No wire-format or audio-pipeline changes — v1.5 through v1.8 interoperate.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The client and the relay server are published from the same GitHub repo;
the server's releases use "server-" prefixed tags. RemSoundUpdater hit
/releases/latest, which is repo-wide — when a server release was newest,
the updater fed "server-v2.3" to ParseTag (-> a bogus 0.0.3) and concluded
"up to date", silently skipping real client updates.
CheckForUpdateAsync now lists /releases and picks the highest-versioned
release whose tag is a RemSound client tag (new IsClientReleaseTag: after
an optional leading "v", first char must be a digit). Drafts and
pre-releases are skipped. The server-side updater already filters to
"server-" tags, so client + server coexist in one repo cleanly.
Also rewrites build-release.ps1 with a data-safety check: it publishes to
a fresh staging folder and aborts the release if any logs/, profiles/,
recordings/ folder, .log file or remsound.config.json is present in the
staged output or the finished zip — preventing a repeat of the v1.5/v1.6
zips that shipped with developer logs and profiles.
No wire-format or audio-pipeline changes — v1.5/v1.6/v1.7 interoperate.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Three reliability fixes. Wire format and audio pipeline unchanged from
v1.4 / v1.5 — all interoperate.
Peer address recovery:
* When a tracked peer goes Unreachable (its resolved address — often a
stale DNS / Pi-hole record, or a peer that rebooted onto a new IP) but
the same peer is still heartbeat-pinging us from a different address,
RemSound now adopts the live address instead of transmitting to a dead
one. HeartbeatService records untracked ping sources; MainForm's
TryAdoptLiveHeartbeatAddress (1 Hz) re-points the sender, heartbeat
tracking and receiver allow-list. Conservative: fires only on the
unambiguous one-unreachable-and-one-source case, private-range (RFC1918)
addresses only so a relay can't hijack the sender, 10 s cooldown.
Reconnect crash:
* Fixed IndexOutOfRangeException in MainForm.SyncConnectedList. A churny
peer-list rebuild (peer reboot) left SelectedIndex pointing past the
rebuilt item array; the 1 Hz status timer read SelectedItem and crashed
the app. New SafeSelectedItem bounds-checks the index; applied to all
three timer-driven sync methods. The status tick is also wrapped in
try/catch so a transient WinForms hiccup logs instead of crashing.
Long-run memory / CPU leak:
* A receiver left running for hours grew to gigabytes and climbing CPU.
Decoder sessions orphaned by peer reconnects were not reaped — every
reconnect mints a fresh (endpoint, streamId) key, and PruneIdleSessions
silently skipped sessions whose PlayoutEngine lookup missed. Rewrote it
to reap on each session's own LastWriteUtc (no cross-dictionary lookup),
added a hard MaxLiveSessions cap as a backstop, and a "stream sessions
live: N" diagnostic line. Bounds both memory and render-thread CPU.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bug fixes:
* BothIndependent recording was double-tapped — when both WASAPI and
ASIO outputs were ticked, recordings came out garbled and ~2x the
expected duration. AudioRecorder now uses four per-lane rings
(sent-wasapi, sent-asio, recv-wasapi, recv-asio) plus a writer-
thread mix step that drains min(wasapi, asio) frames and sum-mixes
with the soft-tanh limiter. Tap signatures gained a RenderRoute
parameter; Mixed maps to the wasapi slot.
* A peer announcing on the WASAPI lane was inaudible when the
receiver only had an ASIO output ticked (and vice versa). The
session opened, samples flowed into the SessionPlayout ring, but
ReadForRoute(AsioLane) skipped any session whose Route was
WasapiLane so nothing drained the ring. PlayoutEngine now tracks
per-lane "is this lane backed by a device" via volatile bools
settable through SetLaneActive(RenderRoute, bool), called from
CompositeRenderBackend.SetOutputDevices whenever the device split
changes. ReadForRoute admits orphan sessions when the other lane
is inactive.
Menu reorganisation:
* New Options menu (Alt+O) holds Recording settings (Alt+S), Keyboard
shortcuts (Alt+K, Ctrl+K), Startup behaviour (Alt+T), Preferences
(Alt+P, Ctrl+P). Pre-v1.5 these were scattered across File menu,
Record menu, and inside the Preferences dialog itself.
* Record menu mnemonic moved from Alt+O to Alt+K, rendered as
"Record (Alt+K)" so the chord is visible despite K not being a
letter in "Record". Alt+R is taken by Receive audio.
* File menu — new Recent profiles submenu (Alt+F, R) listing the
five most-recently-opened profiles. Press 1..5 inside the submenu
to jump to a slot. Missing files are skipped from the menu but
kept in storage.
* Rename current profile moves to Alt+M (was R), Minimise to tray
moves to Alt+N (was M).
* Lock to audio clock was Alt+K, now Alt+D.
UX additions:
* Ctrl+O = Open profile (matches the menu chord).
* New global hotkey: Start / Stop recording. Pickable from Options
-> Keyboard shortcuts. Unbound by default. Works system-wide.
Wire format and audio pipeline unchanged from v1.4 — v1.4 and v1.5
peers interoperate.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Ships the dialog reorganisation and mnemonic adjustments from
commit 45b709d. No wire-format or audio-pipeline changes — v1.3 and
v1.4 peers interoperate.
About dialog gets a v1.4 release-notes block; csproj Version
property bumped 1.3.0 → 1.4.0; RELEASE_NOTES.md rewritten with the
v1.4 highlights and a one-paragraph upgrading-from-v1.3 note (clean
auto-update via Help → Check for updates is now expected to work on
Dropbox-installed copies since v1.3's helper hardening took effect).
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Recording settings dialog
=========================
Promote what were row attributes inside the format-attributes table
into their own dedicated listboxes. Pre-rework each format's
attribute list contained a row per (quality knob × channel mode)
combination — doubling the row count for no real benefit. WAV had
6 rows, MP3 / OGG-Opus had 8, FLAC had 4. Channel mode is now its
own listbox; per-format attribute lists shrink to 3 (WAV), 4
(MP3 / OGG-Opus), 2 (FLAC).
FLAC compression level was previously a hidden RecordingSettings
field locked at the libFLAC default of 5. Now exposed as its own
conditional listbox with all 9 levels (0..8). Visible only when the
selected file format is FLAC; the column collapses to 0% width
otherwise so the dialog reads cleanly as 4 columns for non-FLAC
formats and 5 for FLAC.
Mnemonics:
* Source Alt+S (unchanged)
* File format Alt+F (unchanged)
* Attributes Alt+A (unchanged)
* Compression Alt+L (new, FLAC-only visible)
* Channels Alt+C (new)
* OK Alt+O (unchanged)
* Cancel Alt+N (was Alt+C — moved to free C for Channels)
Record menu mnemonics
=====================
Item mnemonics inside the Record menu reorganised so the keystroke
inside the menu matches the same letter used for Ctrl+R outside it,
and Settings reads naturally:
* Start recording / Stop recording Alt+R (was Alt+S)
* Recording settings Alt+S (was Alt+T)
* Open recordings folder Alt+O (unchanged)
* Change recordings folder Alt+C (unchanged)
UpdateStartStopRecordingMenuLabel keeps the underline on a literal
R in both states ("Sta&rt recording" → "Stop &recording") so the
keystroke does the same job regardless of which label is active.
No version bump — the running version stays at 1.3.0. These changes
will be bundled with the next public release alongside whatever else
is in flight.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Anyone on v1.0 / v1.1 / v1.2 in a Dropbox-synced folder will hit the
auto-updater bug v1.3 fixes — but the fix lives in the running
version's helper, so their installed binary can't use it. Adds a
prominent "Already on an older version and inside a Dropbox folder?
Read this first" section at the top of the release notes with the
manual-install steps, and renames the existing install steps to
"Install (clean machine)" for clarity.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
v1.2's self-updater silently failed when the install folder lived
inside a Dropbox sync path. Dropbox held write locks on the existing
RemSound.exe / DLLs during the brief window between the parent exiting
and the helper script copying the new files in. The helper's robocopy
(/R:5 /W:1) gave up after 5 seconds, and the helper then
unconditionally relaunched the OLD binary — so the user saw the same
version they started with after pressing "Yes" on the install prompt,
with no visible error.
Helper script (BuildInstallScript) changes:
* Robocopy retries bumped to /R:60 /W:1 — up to 60 seconds per file.
Dropbox lock release happens reliably within that window in
practice.
* Robocopy exit code is captured and checked. Codes >= 8 are real
failures. On a failure the helper writes update-failed.txt to the
install folder with the cause + recovery steps, leaves the staging
folder intact, and does NOT relaunch the old binary. Earlier
versions silently relaunched the unmodified old binary, hiding the
failure.
* Helper appends a step-by-step trace to _update-helper.log (in the
install folder), with robocopy's own output included via /LOG+:.
* update-failed.txt, _update-helper.log, and _apply-update.cmd are
added to the /XF exclusion list so the helper's own state files
don't get copied to themselves on a repeat update run.
DownloadAndStageInstallAsync also clears any stale update-failed.txt
at the start of every new attempt, so a successful run leaves the
install folder clean.
readme.html "If install fails" section expanded with the new
update-failed.txt marker file behaviour and the _update-helper.log
location.
Wire format, audio pipeline, and recording feature unchanged from
v1.2 — this is updater-machinery-only.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
New user-facing features:
* Recording. Dedicated Record menu (Alt+O — moved from Alt+R to
avoid clashing with the Receive audio checkbox), Start/Stop on
Ctrl+R, settings dialog, per-profile source / format / bit-depth /
channel-mode / folder. Three source modes (received only, sent
only, both). Files are crash-resilient — a process crash
mid-recording leaves a playable file containing everything up to
the last header refresh (~5 seconds).
* Four output formats, all functional:
- WAV: 16/24-bit PCM or 32-bit float, custom writer with
periodic RIFF re-patching.
- MP3: LAME 128–320 kbps CBR (via NAudio.Lame).
- OGG-Opus: 96–256 kbps VBR (via Concentus.Oggfile, reusing the
Concentus encoder from the wire path).
- FLAC: 16/24-bit lossless (via CUETools.Codecs.FLAKE — pure
managed, no native DLL).
* Recording start/stop sound cues. record start.wav and
record stop.wav play around the recording transition. Played via
System.Media.SoundPlayer to the default Windows output, separate
from the recording pipeline so a normal recording does not contain
the cue.
* Per-cue Preferences. The old single "Mute connect/disconnect
sounds" checkbox is replaced by a CheckedListBox: Connect /
Disconnect / Recording start / Recording stop. Old profiles with
the legacy MuteConnectionCues=true are honoured on first load via
a migration path in the new Load* helpers.
* Receiver-side drift compensation switched from discrete
single-frame splices to a continuous WdlResampler at a smoothed
rate ratio. SessionPlayout.cs rewrite.
Diagnostics (only active with Enable logs ticked):
* Per-stage discontinuity probes — sender raw capture (per backend,
PushModeWasapi + Asio both wired), sender pre-encode (now per
lane in BothIndependent, fixing a cross-stream artefact), receiver
post-decode, post-ring, post-resampler.
* Wire-level packet sequence tracking on each PCM stream — in-order
/ missed / reordered / duplicated counts in the diag log.
* Clipped-sample delta in the diag log.
* New AudioStepProbe in RemSound.Core with per-channel scan helper.
UI changes:
* Record menu uses Alt+O (Rec&ord). Inside the menu, item mnemonics
unchanged (S / T / O / C).
* Auto-tune interval combo label is mode-aware: "Auto-tune latency
interval" in classic modes, "Auto-tune interval — WASAPI and ASIO"
in BothIndependent. The combo's Enabled state now follows EITHER
lane's auto-tune checkbox (was only the WASAPI one — bug).
Wire format and audio pipeline unchanged from v1.1 — v1.1 and v1.2
peers interoperate.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Adds a v1.1 block above the existing v1.0 block in the About dialog,
covering the priority-mode feature, always-on network packet priority,
larger UDP kernel buffers, MMCSS Critical bump, and the logging-toggle
bug fix. Version property in the .csproj bumped from 1.0.0 to 1.1.0 so
the About dialog heading and the self-updater's CurrentVersion both
report 1.1.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Priority mode (per-profile toggle, Audio profile tab, Alt+U):
- Bundled PROCESS_POWER_THROTTLING_STATE with EXECUTION_SPEED and
IGNORE_TIMER_RESOLUTION flags (keeps CPU at full clock + 1 ms quantum
pinned regardless of OS idle decisions)
- PowerSetRequest(EXECUTION_REQUIRED) + SetThreadExecutionState so the
system can't decide we're idle and downclock
- timeBeginPeriod(1) for 1 ms scheduler quantum
- ProcessPriorityClass.High, ProcessMemoryPriority normal, working-set
minimum locked via SetProcessWorkingSetSizeEx(HARDWS_MIN_ENABLE)
- Fully reversed on toggle-off and on form close
- Off by default (battery cost on laptops); explicit opt-in per profile
MMCSS thread priority on hot threads bumped from High to Critical
(network listener, mix loop, render thread) — always on.
Network packet priority (always on, no toggle):
- qWAVE flow on the outbound UDP socket at QOS_TRAFFIC_TYPE_VOICE with
QOS_NON_ADAPTIVE_FLOW (Voice priority, DSCP 46/EF, WMM Voice on Wi-Fi)
- Silent fallback if qwave.dll missing or QoS service disabled
- Sender + receiver kernel UDP buffers bumped to 1 MB each (was
256 KB send / 512 KB receive) for resilience against ~30 ms stalls
Bug fix: PreferencesDialog logging toggle no longer flags the profile
as dirty. Logging is a machine-local AppConfig setting; the spurious
ChangedAnyProfileSetting flag was triggering the unsaved-changes prompt
on exit after toggling logs alone.
Manual updated: new "Use CPU and Windows performance settings in high
priority mode (Alt+U)" subsection on the Audio profile tab, new
"Network packet priority (qWAVE / Voice / WMM)" subsection in the
network chapter, Alt+U row added to keyboard-shortcuts table.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
EdnunpandClaude Fable 5
Ednunp