src/AquaSynth.Core/IpaGestureExperiment.cs

@@ -196,27 +196,52 @@ public static class IpaGestureExperiment
 
         var duration = Math.Clamp(target.DurationSeconds * variant.DurationScale, 0.03f, 1.5f);
         var intensity = Math.Clamp(variant.IntensityScale, 0.05f, 2f);
-        var sustain = Math.Clamp(duration + 0.12f, 0.15f, 2f);
-        var descriptor = target.Descriptor.Replace(' ', '_');
-        var frequency = DescriptorTokens(target.Descriptor).Contains("vowel") ? 150 : 135;
-
-        return $$"""
-            patch gain=.45
-            morphology name=oral length_cm=17 diameters=.55,.75,1.05,1.45,1.25,.8 tongue_index=3 tongue_diameter=1.4 constriction_index=4 constriction_diameter=1 lip_opening=1.2
-            morphology name=nasal length_cm=12 diameters=.04,.28,.52,.72
-            waveguide_path name=oral_path morphology=oral strategy=thiran order=1 max_delay=4096 loss=.998
-            waveguide_path name=nasal_path morphology=nasal strategy=thiran order=1 max_delay=4096 loss=.997
-            source_port name=folds path=oral_path kind=glottal position=0 pressure=.66 tension=.54 opening=.42 noise=.04 impedance=.32 flow_scale=.12
-            branch_port name=velopharynx from=oral_path from_position=.45 to=nasal_path opening=.01 coupling=1
-            constriction_contact name=contact path=oral_path position=.92 opening=.5 resistance=.45 stored_pressure=.12
-            radiation_load name=mouth path=oral_path kind=lip position=1 aperture=.8 reflection=-.82 impedance=.28
-            probe_timeline name=flow network=voice blocks=12 block_size=64
-            vocal_network name=voice paths=oral_path,nasal_path sources=folds contacts=contact branches=velopharynx radiation=mouth probe=flow
-            phoneme_gesture name={{SafeDslName(target.Id)}} ipa={{target.Ipa}} descriptor={{descriptor}} start=0 dur={{Format(duration)}} intensity={{Format(intensity)}}
-            vocal network=voice freq={{Format(frequency)}} gain=.7 sustain={{Format(sustain)}}
-            """;
+        return Scaffold(
+            [GestureLine(target, 0, duration, intensity)],
+            Math.Clamp(duration + 0.12f, 0.15f, 2f),
+            DescriptorTokens(target.Descriptor).Contains("vowel") ? 150 : 135);
     }
 
+    /// <summary>
+    /// One word on the same scaffold: the targets run back to back, each at its own intensity (a stress mark is a higher
+    /// intensity on the stressed vowel). An empty descriptor leaves it to the DSL's own IPA table. Pitch stays flat at <paramref name="frequency"/>; no contour is authored.
+    /// </summary>
+    public static string BuildWordScript(IReadOnlyList<(IpaGestureExperimentTarget Target, float Intensity)> phones, float frequency = 140, float tailSeconds = 0.12f)
+    {
+        if (phones.Count == 0) throw new ArgumentException("a word needs at least one phone", nameof(phones));
+        var lines = new List<string>();
+        var start = 0f;
+        foreach (var (target, intensity) in phones)
+        {
+            ArgumentException.ThrowIfNullOrWhiteSpace(target.Id);
+            ArgumentException.ThrowIfNullOrWhiteSpace(target.Ipa);
+            var duration = Math.Clamp(target.DurationSeconds, 0.03f, 1.5f);
+            lines.Add(GestureLine(target, start, duration, Math.Clamp(intensity, 0.05f, 2f)));
+            start += duration;
+        }
+
+        return Scaffold(lines, Math.Clamp(start + tailSeconds, 0.15f, 4f), frequency);
+    }
+
+    private static string GestureLine(IpaGestureExperimentTarget target, float start, float duration, float intensity) =>
+        $"phoneme_gesture name={SafeDslName(target.Id)} ipa={target.Ipa} {(target.Descriptor.Length == 0 ? "" : $"descriptor={target.Descriptor.Replace(' ', '_')} ")}start={Format(start)} dur={Format(duration)} intensity={Format(intensity)}";
+
+    private static string Scaffold(IReadOnlyList<string> gestureLines, float sustain, float frequency) => $$"""
+        patch gain=.45
+        morphology name=oral length_cm=17 diameters=.55,.75,1.05,1.45,1.25,.8 tongue_index=3 tongue_diameter=1.4 constriction_index=4 constriction_diameter=1 lip_opening=1.2
+        morphology name=nasal length_cm=12 diameters=.04,.28,.52,.72
+        waveguide_path name=oral_path morphology=oral strategy=thiran order=1 max_delay=4096 loss=.998
+        waveguide_path name=nasal_path morphology=nasal strategy=thiran order=1 max_delay=4096 loss=.997
+        source_port name=folds path=oral_path kind=glottal position=0 pressure=.66 tension=.54 opening=.42 noise=.04 impedance=.32 flow_scale=.12
+        branch_port name=velopharynx from=oral_path from_position=.45 to=nasal_path opening=.01 coupling=1
+        constriction_contact name=contact path=oral_path position=.92 opening=.5 resistance=.45 stored_pressure=.12
+        radiation_load name=mouth path=oral_path kind=lip position=1 aperture=.8 reflection=-.82 impedance=.28
+        probe_timeline name=flow network=voice blocks=12 block_size=64
+        vocal_network name=voice paths=oral_path,nasal_path sources=folds contacts=contact branches=velopharynx radiation=mouth probe=flow
+        {{string.Join("\n", gestureLines)}}
+        vocal network=voice freq={{Format(frequency)}} gain=.7 sustain={{Format(sustain)}}
+        """;
+
     private static IReadOnlyList<IpaGestureMetricSummary> SummarizeMetrics(IReadOnlyList<IpaGestureExperimentMetric> metrics) =>
         metrics
             .Where(metric => metric.Layer.Equals("gesture", StringComparison.OrdinalIgnoreCase))

src/AquaSynth.Core/PatchScript.cs

@@ -1518,6 +1518,7 @@ public static class PatchScript
                 "p" => "voiceless_bilabial_plosive",
                 "b" => "voiced_bilabial_plosive",
                 "m" => "voiced_bilabial_nasal",
+                "n" => "voiced_alveolar_nasal",
                 "s" => "voiceless_alveolar_fricative",
                 "z" => "voiced_alveolar_fricative",
                 "ʃ" => "voiceless_postalveolar_fricative",

tests/AquaSynth.Dsl.Tests/PatchScriptTests.cs

@@ -613,6 +613,47 @@ public sealed class PatchScriptTests
         Assert.True(fricativeControls["/vocal/sources/0/noise"] > nasalControls["/vocal/sources/0/noise"]);
     }
 
+    [Fact]
+    public void BareNIsAnAlveolarNasalNotTheApproximantFallback()
+    {
+        var patch = PatchScript.Parse("""
+            morphology name=oral length_cm=17 diameters=.6,.8,1.2,1.6,1.3,.9 tongue_index=3 tongue_diameter=1.4 constriction_index=4 constriction_diameter=1 lip_opening=1.5
+            morphology name=nasal length_cm=12 diameters=.05,.35,.6,.8
+            waveguide_path name=oral_path morphology=oral loss=.998
+            waveguide_path name=nasal_path morphology=nasal loss=.997
+            source_port name=folds path=oral_path pressure=.7 tension=.55 opening=.45 noise=.05 impedance=.3
+            branch_port name=velopharynx from=oral_path from_position=.45 to=nasal_path opening=.01 coupling=1
+            constriction_contact name=contact path=oral_path position=.92 opening=.5 resistance=.4 stored_pressure=.1
+            radiation_load name=mouth path=oral_path aperture=.8 reflection=-.82 impedance=.28
+            vocal_network name=voice paths=oral_path,nasal_path sources=folds contacts=contact branches=velopharynx radiation=mouth
+            phoneme_gesture name=en ipa=n start=0 dur=.1
+            vocal network=voice freq=140 gain=.2 sustain=.3
+            """);
+
+        Assert.Equal("voiced_alveolar_nasal", patch.PhonemeGestures.Single().Descriptor);
+        Assert.Contains(patch.ControlSplines, spline => spline.Name.StartsWith("en_velum", StringComparison.Ordinal) && spline.SurfacePath == "/vocal/branches/0/opening");
+        var catalog = ControlSurfaceCatalog.FromPatch(patch);
+        Assert.True(catalog.CreateTimeline().ControlValuesAt(.03f)["/vocal/branches/0/opening"] > .5f, "the velum opens for n");
+    }
+
+    [Fact]
+    public void WordScriptRunsGesturesBackToBackAndLeavesDescriptorsToTheDslTable()
+    {
+        var script = IpaGestureExperiment.BuildWordScript(
+        [
+            (new IpaGestureExperimentTarget("w_n", "n", "", .1f), 1f),
+            (new IpaGestureExperimentTarget("w_i", "i", "", .2f), 1.2f),
+            (new IpaGestureExperimentTarget("w_b", "b", "", .1f), 1f)
+        ], frequency: 140);
+
+        var patch = PatchScript.Parse(script);
+        Assert.Equal(["n", "i", "b"], patch.PhonemeGestures.Select(gesture => gesture.Ipa));
+        Assert.Equal([0f, .1f, .3f], patch.PhonemeGestures.Select(gesture => gesture.StartSeconds));
+        Assert.Equal(["voiced_alveolar_nasal", "voiced_close_front_unrounded_vowel", "voiced_bilabial_plosive"], patch.PhonemeGestures.Select(gesture => gesture.Descriptor));
+        Assert.Equal(1.2f, patch.PhonemeGestures[1].Intensity);
+        Assert.Throws<ArgumentException>(() => IpaGestureExperiment.BuildWordScript([]));
+    }
+
     [Fact]
     public void IpaGestureExperimentWritesFrozenRoundBundle()
     {

tools/WordRender/Program.cs

@@ -0,0 +1,87 @@
+using System.Runtime.CompilerServices;
+using System.Runtime.InteropServices;
+using System.Security.Cryptography;
+using System.Text.Json;
+using AquaSynth.Dsl;
+using AquaSynth.Faust;
+
+// Offline probe: one word as a run of IPA gestures on the IpaGestureExperiment scaffold, rendered through the native
+// libfaust path to a mono 16-bit WAV plus script.aqua, source.dsp and manifest.json. It measures peak, rms and NaN;
+// it cannot say how the word sounds.
+// usage: WordRender <outDir> <label> <freq-hz> <ipa>:<ms>:<intensity> ...      e.g. WordRender out nibu 140 n:95:1 i:180:1.2 b:95:1 u:180:.9
+const int Rate = 48000, Frames = 512;
+if (args.Length < 4) { Console.Error.WriteLine("usage: WordRender <outDir> <label> <freq-hz> <ipa>:<ms>:<intensity> ..."); return 2; }
+var outDir = Path.GetFullPath(args[0]);
+var label = args[1];
+var freq = float.Parse(args[2], System.Globalization.CultureInfo.InvariantCulture);
+var phones = args.Skip(3).Select((spec, i) =>
+{
+    var p = spec.Split(':');
+    var ipa = p[0];
+    var ms = float.Parse(p[1], System.Globalization.CultureInfo.InvariantCulture);
+    var intensity = float.Parse(p[2], System.Globalization.CultureInfo.InvariantCulture);
+    return (new IpaGestureExperimentTarget($"{label}_{i}_{ipa}", ipa, "", ms / 1000f), intensity);
+}).ToArray();
+Directory.CreateDirectory(outDir);
+
+var script = IpaGestureExperiment.BuildWordScript(phones, freq);
+File.WriteAllText(Path.Combine(outDir, "script.aqua"), script + "\n");
+var seconds = phones.Sum(p => p.Item1.DurationSeconds) + 0.12f;
+var home = AquaSynthFaustToolchain.Find();
+Console.WriteLine($"faust home: {home ?? "(not found)"}");
+using var compiler = new AquaSynthPatchCompiler(new AquaSynthNativeOptions(SampleRate: Rate));
+// The exporter bakes splines in as offsets summed per surface; the engine's ControlSplineTimeline merges them per surface
+// into one absolute curve. Back-to-back gestures only sequence under the timeline, so the DSP is emitted without splines
+// (surface sliders only) and the timeline drives the sliders block by block, as NativeFaustRuntime.ControlValuesAt does.
+var scripted = PatchScript.Parse(script);
+var timeline = ControlSurfaceCatalog.FromPatch(scripted).CreateTimeline();
+var source = FaustEmitter.Emit(scripted with { ControlSplines = [] }, new FaustExportOptions(label)).Source;
+File.WriteAllText(Path.Combine(outDir, "source.dsp"), source);
+if (!compiler.TryCompileSource(new AquaSynthCompileIdentity(label, label, ""), source, seconds, out var patch0, out var cerr))
+{
+    Console.Error.WriteLine("COMPILE FAILED: " + cerr);
+    return 1;
+}
+
+using var patch = patch0!;
+using var stream = patch.CreateStreamingPatch();
+var buf = GC.AllocateArray<float>(Frames, pinned: true);
+var ptrs = GC.AllocateArray<IntPtr>(1, pinned: true);
+unsafe { ptrs[0] = (IntPtr)Unsafe.AsPointer(ref buf[0]); }
+var outP = Marshal.UnsafeAddrOfPinnedArrayElement(ptrs, 0);
+var blocks = (int)Math.Ceiling(seconds * Rate / Frames);
+var pcm = new short[blocks * Frames];
+double sum = 0; float peak = 0; var nan = false;
+for (var blk = 0; blk < blocks; blk++)
+{
+    var ctl = patch.ControlValuesAt(timeline, blk * Frames / (float)Rate);
+    stream.ProcessBlock(Frames, IntPtr.Zero, outP, ctl);
+    for (var i = 0; i < Frames; i++)
+    {
+        var v = buf[i];
+        nan |= float.IsNaN(v) || float.IsInfinity(v);
+        peak = Math.Max(peak, Math.Abs(v));
+        sum += (double)v * v;
+        pcm[blk * Frames + i] = (short)(Math.Clamp(v, -1f, 1f) * 32767);
+    }
+}
+
+var wav = Path.Combine(outDir, $"{label}.wav");
+using (var w = new BinaryWriter(File.Create(wav)))
+{
+    var db = pcm.Length * 2;
+    w.Write("RIFF"u8); w.Write(36 + db); w.Write("WAVEfmt "u8); w.Write(16); w.Write((short)1); w.Write((short)1);
+    w.Write(Rate); w.Write(Rate * 2); w.Write((short)2); w.Write((short)16); w.Write("data"u8); w.Write(db);
+    foreach (var s in pcm) w.Write(s);
+}
+
+var rms = Math.Sqrt(sum / pcm.Length);
+File.WriteAllText(Path.Combine(outDir, "manifest.json"), JsonSerializer.Serialize(new
+{
+    label, freq, phones = args.Skip(3), seconds, sampleRate = Rate, frames = pcm.Length, peak, rms, nonFinite = nan,
+    wavSha256 = Convert.ToHexString(SHA256.HashData(File.ReadAllBytes(wav))).ToLowerInvariant(),
+    scriptSha256 = Convert.ToHexString(SHA256.HashData(File.ReadAllBytes(Path.Combine(outDir, "script.aqua")))).ToLowerInvariant(),
+    os = RuntimeInformation.OSDescription, faustHome = home
+}, new JsonSerializerOptions { WriteIndented = true }));
+Console.WriteLine($"{wav} {seconds:F2}s peak={peak:F3} rms={rms:F4} nonFinite={nan}");
+return nan ? 1 : 0;

tools/WordRender/WordRender.csproj

@@ -0,0 +1,12 @@
+<Project Sdk="Microsoft.NET.Sdk">
+  <PropertyGroup>
+    <OutputType>Exe</OutputType>
+    <TargetFramework>net10.0</TargetFramework>
+    <ImplicitUsings>enable</ImplicitUsings>
+    <Nullable>enable</Nullable>
+    <AllowUnsafeBlocks>true</AllowUnsafeBlocks>
+  </PropertyGroup>
+  <ItemGroup>
+    <ProjectReference Include="..\..\src\AquaSynth.Faust\AquaSynth.Faust.csproj" />
+  </ItemGroup>
+</Project>