From 7c88a1ad732575513e2a11b757543643b2df2c19 Mon Sep 17 00:00:00 2001
From: Jer Miller <jeremie.miller@gmail.com>
Date: Wed, 18 Mar 2026 10:28:46 -0600
Subject: [PATCH] Add .wav to transcribe supported_formats

---
 docs/JOURNAL.md            | 8 ++++----
 docs/OBSERVE.md            | 2 +-
 observe/transcribe/main.py | 2 +-
 observe/utils.py           | 4 ++--
 4 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/docs/JOURNAL.md b/docs/JOURNAL.md
index a1295bcd4..add1e47b1 100644
--- a/docs/JOURNAL.md
+++ b/docs/JOURNAL.md
@@ -25,7 +25,7 @@ solstone transforms raw recordings into actionable understanding through a three
 ┌─────────────────────────────────────┐
 │  LAYER 1: CAPTURES                  │  Raw recordings
 │  (Binary media files)               │  "What was recorded"
-│  - *.flac, *.ogg, *.opus (audio)    │
+│  - *.flac, *.ogg, *.opus, *.wav (audio)    │
 │  - *.webm (video)                   │
 └─────────────────────────────────────┘
 ```
@@ -36,7 +36,7 @@ solstone transforms raw recordings into actionable understanding through a three
 
 | Term | Definition | Examples |
 |------|------------|----------|
-| **Capture** | Raw audio/video recording | `*.flac`, `*.ogg`, `*.opus`, `*.webm` |
+| **Capture** | Raw audio/video recording | `*.flac`, `*.ogg`, `*.opus`, `*.wav`, `*.webm` |
 | **Extract** | Structured data from captures | `*.jsonl` |
 | **Agent Output** | AI-generated narrative summary | `agents/*.md`, `HHMMSS_LEN/*.md` |
 
@@ -180,7 +180,7 @@ Fields:
 - `raw_media_days` (integer or null) – Number of days to retain raw media when mode is `"days"`. Required when `raw_media` is `"days"`, ignored otherwise.
 - `per_stream` (object) – Per-stream overrides keyed by stream name. Each entry supports `raw_media` and `raw_media_days`. Omitted fields inherit from the global retention settings.
 
-"Raw media" means layer 1 capture files only: audio files (`.flac`, `.opus`, `.ogg`, `.m4a`), video files (`.webm`, `.mov`), and screen diffs (`monitor_*_diff.png`).
+"Raw media" means layer 1 capture files only: audio files (`.flac`, `.opus`, `.ogg`, `.m4a`, `.wav`), video files (`.webm`, `.mov`), and screen diffs (`monitor_*_diff.png`).
 
 All layer 2 and layer 3 content is always preserved regardless of retention policy: transcripts (`audio.jsonl`, `screen.jsonl`), agent outputs (`agents/*.md`), speaker labels (`agents/speaker_labels.json`), facet events (`events/*.jsonl`), entity data, segment metadata (`stream.json`), and search index entries.
 
@@ -926,7 +926,7 @@ Audio files are initially written to the day root with the segment key prefix (L
 
 After transcription, audio files are moved into their segment folder:
 
-- `HHMMSS_LEN/*.flac`, `*.m4a`, `*.ogg`, or `*.opus` – audio files moved here after processing, preserving descriptive suffix (e.g., `audio.flac`, `audio.m4a`, `imported_audio.opus`)
+- `HHMMSS_LEN/*.flac`, `*.m4a`, `*.ogg`, `*.opus`, or `*.wav` – audio files moved here after processing, preserving descriptive suffix (e.g., `audio.flac`, `audio.m4a`, `imported_audio.opus`)
 
 Note: The descriptive portion after the segment key (e.g., `_audio`, `_recording`) is preserved when files are moved into segment directories. Processing tools match files by extension only, ignoring the descriptive suffix.
 
diff --git a/docs/OBSERVE.md b/docs/OBSERVE.md
index 396724ef0..d00de1f6c 100644
--- a/docs/OBSERVE.md
+++ b/docs/OBSERVE.md
@@ -18,7 +18,7 @@ Multimodal capture and AI-powered analysis of desktop activity.
 ```
 sol observer (platform-detected capture)
        ↓
-   Raw media files (*.flac, *.ogg, *.opus, *.webm, tmux_*.jsonl)
+   Raw media files (*.flac, *.ogg, *.opus, *.wav, *.webm, tmux_*.jsonl)
        ↓
 sol sense (coordination)
    ├── sol transcribe → audio.jsonl
diff --git a/observe/transcribe/main.py b/observe/transcribe/main.py
index 9245916d8..20507225b 100644
--- a/observe/transcribe/main.py
+++ b/observe/transcribe/main.py
@@ -654,7 +654,7 @@ def main():
         parser.error(f"Audio file not found: {audio_path}")
 
     # Validate supported formats
-    supported_formats = {".flac", ".m4a", ".mp3", ".ogg", ".opus"}
+    supported_formats = {".flac", ".m4a", ".mp3", ".ogg", ".opus", ".wav"}
     if audio_path.suffix.lower() not in supported_formats:
         parser.error(
             f"Unsupported audio format: {audio_path.suffix}. "
diff --git a/observe/utils.py b/observe/utils.py
index eca3cd12b..1020b0b3c 100644
--- a/observe/utils.py
+++ b/observe/utils.py
@@ -52,12 +52,12 @@ def load_audio(raw_path: Path, sample_rate: int = SAMPLE_RATE) -> np.ndarray:
 
     For M4A files from sck-cli (which contain two mono streams: track 0 =
     system audio, track 1 = microphone), all streams are decoded and mixed
-    together. Other formats (.flac, .mp3, .ogg, .opus) are decoded via ffmpeg.
+    together. Other formats (.flac, .mp3, .ogg, .opus, .wav) are decoded via ffmpeg.
 
     Parameters
     ----------
     raw_path : Path
-        Path to audio file (.flac, .ogg, .opus, or .m4a)
+        Path to audio file (.flac, .ogg, .opus, .wav, or .m4a)
     sample_rate : int
         Target sample rate (default: 16000)
 
-- 
2.54.0