Apply TTS presets with vendor tags and both hives

CURRENT.md

@@ -40,6 +40,14 @@ mobile draft at 390px.
   - [x] C11 preview caching: Discord-hosted image/GIF previews (WebP 320px, hash-addressed, 5 MiB
     cap, one attempt) run after every walk in both modes. Live run `...-20260916090722`:
     `previewed: 2`; thumbs served over `/api/media/<scrape>/<item>/thumb` (200).
+  - [x] C64 Correct preset application. The design's bug-fix section landed in the wrapper:
+    Flextalk rate/pitch now ride as sticky `\!R` / `\!%` vendor tags (attributes only for MSTTS),
+    neutral sliders emit neither, word mode tags every word, and presets write the audible profile
+    to both registry hives with Gender/Line Mode plus a Default Speaker pointer. `--voices`
+    reflects the active preset in the Keith Bell `defaults=` flags so the panel snap points match.
+    New `microslopsam/setup/test-wrapper.sh` verifies the whole contract with stubbed wine and
+    engine binaries (tags vs attributes, registry shape, preset extras, implied VOX, metadata);
+    all checks pass.
   - [x] C63 Half-Life presets. The wrapper now implements `VESKSLOP_TTS_PRESET=vox|hev` (Keith
     Bell profile: VOX pitch 3/rate 17/prominence 22/head 52/volume 70, HEV pitch 15/rate 24,
     expression 59; other voices untouched; preset < profile < expression; the `vox` preset implies

docs/voice.md

@@ -164,7 +164,13 @@ from the TTS panel:
   A **preset** select (`tts.preset`: factory, VOX, HEV) shifts the Keith Bell speaker profile to
   the Half-Life approximation; `vox` also implies VOX prosody unless VOX is explicitly off, and
   overrides travel as `VESKSLOP_TTS_PRESET` / `{preset}`. Precedence is preset < profile <
-  expression, so the sliders still win over the preset.
+  expression, so the sliders still win over the preset. The wrapper writes the audible profile to
+  both registry hives (plus the preset's Gender/Line Mode and a Default Speaker pointer), and for
+  Flextalk it sends rate/pitch as the sticky `\!R` / `\!%` vendor tags rather than
+  `ITTSAttributes`; attributes only apply to MSTTS voices. Neutral sliders emit neither, so the
+  preset is reproduced exactly. `sh microslopsam/setup/test-wrapper.sh` checks this contract with
+  stubbed wine and engine binaries, no Wine install needed. The preset values also show up in the
+  voice metadata defaults, so the panel's "default" snap points match what the engine gets.
 - **Test lines** - Attention, Laboratory, Critical, Health, and Radiation, plus two custom lines
   kept in `localStorage`. Each button speaks the line with the current profile, so a change can be
   auditioned without touching the real queue.

microslopsam/README.md

@@ -22,6 +22,8 @@ pure-native build, and keep a path to a fully portable engine.
   - AT&T Flextalk: Keith Bell, Kerry Watson
 - Test output: `build/test_sam.wav` (22.05 kHz), `build/test_Keith_Bell.wav`,
   `build/test_Kerry_Watson.wav` (11.025 kHz).
+- Wrapper contract test (no Wine needed, stubs wine and the engine host):
+  `sh setup/test-wrapper.sh`.
 
 ## Build
 

microslopsam/setup/test-wrapper.sh

@@ -0,0 +1,110 @@
+#!/bin/sh
+# Regression test for the Veskslop TTS wrapper contract. No Wine needed: the wine
+# and engine-host binaries are stubbed, so this checks the arguments the wrapper
+# builds, the registry import it generates, and the voice metadata it publishes.
+#
+# The bug this guards against: rate/pitch sent as ITTSAttributes (-s/-p) override
+# the Flextalk speaker profile, so the VOX/HEV presets sounded too fast and not
+# deep enough. Flextalk must use the sticky vendor tags (\!R / \!%), and neutral
+# sliders must emit neither.
+set -eu
+
+HERE=$(CDPATH= cd -- "$(dirname -- "$0")" && pwd)
+WRAP="$HERE/veskslop-tts.sh"
+WORK=$(mktemp -d)
+trap 'rm -rf "$WORK"' EXIT
+
+mkdir -p "$WORK/build" "$WORK/bin" "$WORK/out"
+"${MICROSLOP_FFMPEG:-ffmpeg}" -hide_banner -loglevel error -f lavfi \
+    -i "sine=frequency=200:duration=0.3" -ar 11025 -ac 1 -c:a pcm_s16le -y "$WORK/sine.wav"
+
+cat > "$WORK/build/sapi4host.exe" <<'STUB'
+#!/bin/sh
+{ printf 'ARGV:'; for a in "$@"; do printf ' [%s]' "$a"; done; printf '\n'; } >> "$STUB_LOG"
+OUT=""
+while [ "$#" -ge 2 ]; do
+    [ "$1" = "-o" ] && OUT=$2
+    shift
+done
+cp "$STUB_SINE" "$OUT"
+STUB
+
+cat > "$WORK/bin/wine" <<'STUB'
+#!/bin/sh
+[ "${1:-}" = "reg" ] && [ "${2:-}" = "import" ] && cp "$3" "$STUB_REG"
+exit 0
+STUB
+chmod +x "$WORK/build/sapi4host.exe" "$WORK/bin/wine"
+
+printf 'Attention all personnel.\n' > "$WORK/text.txt"
+
+export MICROSLOP_BUILD="$WORK/build"
+export MICROSLOP_WINE="$WORK/bin/wine"
+export MICROSLOP_FFMPEG="${MICROSLOP_FFMPEG:-ffmpeg}"
+export DISPLAY=:99
+export STUB_LOG="$WORK/host.log"
+export STUB_REG="$WORK/profile.reg"
+export STUB_SINE="$WORK/sine.wav"
+
+FAIL=0
+check() {
+    if [ "$2" = "$3" ]; then
+        echo "ok   - $1"
+    else
+        echo "FAIL - $1: expected [$2], got [$3]"
+        FAIL=1
+    fi
+}
+
+run_wrapper() {
+    : > "$STUB_LOG"
+    sh "$WRAP" "$WORK/out.wav" "$WORK/text.txt" >/dev/null
+}
+
+# The spoken text is the -t argument; the engine host writes to the wrapper's temp file.
+text_for() {
+    sed -n '1s/.*\[-t\] \[\(.*\)\] \[-o\].*/\1/p' "$STUB_LOG"
+}
+
+attr_count() {
+    grep -oE '\[-s\]|\[-p\]' "$STUB_LOG" | wc -l | tr -d ' '
+}
+
+# Neutral Flextalk sliders emit neither attributes nor tags.
+VESKSLOP_TTS_VOICE=keith-bell run_wrapper
+check "neutral flextalk text" "Attention all personnel." "$(text_for)"
+check "neutral flextalk has no attributes" "0" "$(attr_count)"
+
+# Adjustments ride as tags; attributes stay out even at non-neutral values.
+VESKSLOP_TTS_VOICE=keith-bell VESKSLOP_TTS_VOX=0 \
+    VESKSLOP_TTS_SPEED=1.3 VESKSLOP_TTS_PITCH=3 run_wrapper
+check "flextalk rate tag" "\!R0.769 \!%1.189 Attention all personnel." "$(text_for)"
+check "flextalk has no attributes" "0" "$(attr_count)"
+
+# MSTTS has no vendor tags and keeps the attributes.
+VESKSLOP_TTS_VOICE=sam VESKSLOP_TTS_SPEED=1.3 VESKSLOP_TTS_PITCH=3 run_wrapper
+check "mstts text has no tags" "Attention all personnel." "$(text_for)"
+check "mstts keeps attributes" "2" "$(attr_count)"
+
+# The vox preset implies VOX prosody when VESKSLOP_TTS_VOX is unset.
+VESKSLOP_TTS_VOICE=keith-bell VESKSLOP_TTS_PRESET=vox run_wrapper
+check "vox preset implies VOX parser" "1" "$(grep -c 'wH1' "$STUB_LOG")"
+
+# The preset writes the full audible profile to both hives.
+VESKSLOP_TTS_VOICE=keith-bell VESKSLOP_TTS_PRESET=vox run_wrapper
+check "preset writes both hives" "2" "$(grep -c 'Speakers\\{11260610' "$STUB_REG")"
+check "preset pitch (both hives)" "2" "$(grep -c '"Pitch"=dword:00000003' "$STUB_REG")"
+check "preset rate (both hives)" "2" "$(grep -c '"Speaking Rate"=dword:00000011' "$STUB_REG")"
+check "preset gender and line mode" "4" "$(grep -cE '"Gender"=dword:00000002|"Line Mode"=dword:00000001' "$STUB_REG")"
+check "preset default speaker" "2" "$(grep -c '"Default Speaker"' "$STUB_REG")"
+
+# A plain Flextalk run keeps the defaults and writes no preset extras.
+VESKSLOP_TTS_VOICE=keith-bell run_wrapper
+check "plain run writes both hives" "2" "$(grep -c 'Speakers\\{11260610' "$STUB_REG")"
+check "plain run has no preset extras" "0" "$(grep -cE 'Gender|Line Mode|Default Speaker' "$STUB_REG")"
+
+# Voice metadata follows the active preset.
+check "voices reflect hev" "1" \
+    "$(VESKSLOP_TTS_PRESET=hev sh "$WRAP" --voices | tr ',' '\n' | grep -c 'keith-bell|Keith Bell|male|en|expression=59+defaults=pitch:15:rate:24')"
+
+[ "$FAIL" = "0" ] && echo "wrapper tests passed" || { echo "wrapper tests failed"; exit 1; }

microslopsam/setup/veskslop-tts.sh

@@ -33,7 +33,14 @@ ROOT=$MICROSLOP_ROOT
 # `expression=NN` marks a Flextalk voice with an Expression profile parameter and
 # its registry default. Expression is applied at voice load, so the wrapper writes
 # it into the Wine registry before synthesis.
-VOICES="sam|Sam|male|en,mary|Mary|female|en,mike|Mike|male|en,keith-bell|Keith Bell|male|en|expression=59+defaults=pitch:27:rate:33:prominences:17:head_size:46:lowering:40:expression:59,kerry-watson|Kerry Watson|female|en|expression=100+defaults=pitch:65:rate:32:prominences:17:head_size:45:lowering:40:expression:100,mike-telephone|Mike (for Telephone)|male|en,mary-telephone|Mary (for Telephone)|female|en,mike-hall|Mike in Hall|male|en,mike-stadium|Mike in Stadium|male|en,mike-space|Mike in Space|male|en,mary-hall|Mary in Hall|female|en,mary-stadium|Mary in Stadium|female|en,mary-space|Mary in Space|female|en"
+# The Keith Bell defaults follow the active preset so the UI's "default" snap points match
+# the values the preset actually writes.
+case "${VESKSLOP_TTS_PRESET:-}" in
+    vox) KB_DEFAULTS="pitch:3:rate:17:prominences:22:head_size:52:lowering:40:expression:59" ;;
+    hev) KB_DEFAULTS="pitch:15:rate:24:prominences:22:head_size:52:lowering:40:expression:59" ;;
+    *)   KB_DEFAULTS="pitch:27:rate:33:prominences:17:head_size:46:lowering:40:expression:59" ;;
+esac
+VOICES="sam|Sam|male|en,mary|Mary|female|en,mike|Mike|male|en,keith-bell|Keith Bell|male|en|expression=59+defaults=$KB_DEFAULTS,kerry-watson|Kerry Watson|female|en|expression=100+defaults=pitch:65:rate:32:prominences:17:head_size:45:lowering:40:expression:100,mike-telephone|Mike (for Telephone)|male|en,mary-telephone|Mary (for Telephone)|female|en,mike-hall|Mike in Hall|male|en,mike-stadium|Mike in Stadium|male|en,mike-space|Mike in Space|male|en,mary-hall|Mary in Hall|female|en,mary-stadium|Mary in Stadium|female|en,mary-space|Mary in Space|female|en"
 
 if [ "${1:-}" = "--voices" ]; then
     echo "$VOICES"
@@ -153,7 +160,21 @@ if [ -f "$TEXTFILE" ]; then
 fi
 
 # Build the engine arguments from the (possibly command-overridden) speed/pitch.
+# Flextalk: rate and pitch ride as sticky vendor tags. Setting them through
+# ITTSAttributes (-s/-p) overrides the speaker profile, which is the bug that made
+# presets sound "too fast and not deep enough". MSTTS has no tags, so it keeps -s/-p.
 SET_ARGS=""
+TAG_PREFIX=""
+if [ -n "$SPEAKER_GUID" ]; then
+    if [ -n "$SPEED" ]; then
+        R=$(awk -v m="$SPEED" 'BEGIN{ if (m==1) exit; v=1/m; if (v<0.25) v=0.25; if (v>4.0) v=4.0; printf "%.3f", v }')
+        [ -n "$R" ] && TAG_PREFIX="${TAG_PREFIX}\\!R${R} "
+    fi
+    if [ -n "$PITCH" ]; then
+        P=$(awk -v st="$PITCH" 'BEGIN{ if (st==0) exit; v=exp(log(2)*st/12); if (v<0.5) v=0.5; if (v>2.0) v=2.0; printf "%.3f", v }')
+        [ -n "$P" ] && TAG_PREFIX="${TAG_PREFIX}\\!%${P} "
+    fi
+else
     if [ -n "$SPEED" ]; then
         S=$(awk -v m="$SPEED" 'BEGIN{printf "%d", 150*m+0.5}')
         [ "$S" -lt 50 ] && S=50
@@ -166,6 +187,7 @@ if [ -n "$PITCH" ]; then
         [ "$P" -gt "$PMAX" ] && P=$PMAX
         SET_ARGS="$SET_ARGS -p $P"
     fi
+fi
 
 # Apply the Flextalk profile before synthesis. Profiles are read at voice load and
 # the host loads the engine per request, so the values are written every call: the
@@ -227,12 +249,21 @@ if [ -n "$SPEAKER_GUID" ]; then
         esac
     done
 
-    REG_KEY="HKLM\\Software\\AT&T\\Watson\\2.0\\FlexTalk\\2.0\\Speakers\\$SPEAKER_GUID"
+    IS_PRESET=0
+    if [ "$ENGVOICE" = "Keith Bell" ]; then
+        case "$PRESET" in
+            vox|hev) IS_PRESET=1 ;;
+        esac
+    fi
+
     # Use printf, not echo: dash's echo interprets \2 in the key path as an
     # escape (STX), which corrupts the key and makes the import a silent no-op.
+    # NetTTS parity: the audible profile goes to both hives. Presets also write the
+    # informational keys from Valve's setup, and point Default Speaker at Keith Bell.
     {
         printf 'Windows Registry Editor Version 5.00\n\n'
-        printf '[%s]\n' "$REG_KEY"
+        for HIVE in 'HKEY_LOCAL_MACHINE' 'HKEY_CURRENT_USER'; do
+            printf '[%s\\Software\\AT&T\\Watson\\2.0\\FlexTalk\\2.0\\Speakers\\%s]\n' "$HIVE" "$SPEAKER_GUID"
             printf '"Pitch"=dword:%08x\n' "$P_PITCH"
             printf '"Speaking Rate"=dword:%08x\n' "$P_RATE"
             printf '"Volume"=dword:%08x\n' "$P_VOLUME"
@@ -240,6 +271,19 @@ if [ -n "$SPEAKER_GUID" ]; then
             printf '"Prominences"=dword:%08x\n' "$P_PROM"
             printf '"Head Size"=dword:%08x\n' "$P_HEAD"
             printf '"Lowering"=dword:%08x\n' "$P_LOWER"
+            if [ "$IS_PRESET" = "1" ]; then
+                printf '"Gender"=dword:00000002\n'
+                printf '"Line Mode"=dword:00000001\n'
+            fi
+            printf '\n'
+        done
+
+        if [ "$IS_PRESET" = "1" ]; then
+            for HIVE in 'HKEY_LOCAL_MACHINE' 'HKEY_CURRENT_USER'; do
+                printf '[%s\\Software\\AT&T\\Watson\\2.0\\FlexTalk\\2.0\\Speakers]\n' "$HIVE"
+                printf '"Default Speaker"="%s"\n\n' "$SPEAKER_GUID"
+            done
+        fi
     } > "$TMP/profile.reg"
 
     # Guard against silent reg-import failures (a malformed key would corrupt
@@ -322,7 +366,7 @@ if [ "$WORDS" = "1" ] && [ -n "$SPEAKER_GUID" ] && [ "$VOX" != "1" ]; then
     INDEX=0
     for word in $(cat "$TXT"); do
         INDEX=$((INDEX + 1))
-        try_run_host "$TMP/w$INDEX.wav" "$word"
+        try_run_host "$TMP/w$INDEX.wav" "${TAG_PREFIX}${word}"
         echo "file '$TMP/w$INDEX.wav'" >> "$TMP/list.txt"
 
         if [ "$GAP" -gt 0 ]; then
@@ -340,7 +384,7 @@ if [ "$WORDS" = "1" ] && [ -n "$SPEAKER_GUID" ] && [ "$VOX" != "1" ]; then
         exit 1
     fi
 else
-    try_run_host "$RAW" "$(cat "$TXT")"
+    try_run_host "$RAW" "${TAG_PREFIX}$(cat "$TXT")"
 fi
 [ -s "$RAW" ] || { echo "engine produced no audio" >&2; exit 1; }