diff --git a/DeepDrftAPI/CLAUDE.md b/DeepDrftAPI/CLAUDE.md
index 6d93bc6..a4fa35a 100644
--- a/DeepDrftAPI/CLAUDE.md
+++ b/DeepDrftAPI/CLAUDE.md
@@ -50,21 +50,21 @@ Returns the WAV bytes from the `tracks` vault with HTTP Range support.
 
 ### POST api/track/upload ([ApiKeyAuthorize])
 
-**Authenticated endpoint.** Accepts a raw WAV upload + metadata as `multipart/form-data`, processes the WAV, stores it in the vault, and persists metadata to SQL. Returns the fully persisted `TrackDto` with `Id` populated.
+**Authenticated endpoint.** Accepts a raw audio file upload (.wav, .mp3, .flac) + metadata as `multipart/form-data`, processes the file, stores it in the vault, and persists metadata to SQL. Returns the fully persisted `TrackDto` with `Id` populated.
 
 - **Header `ApiKey`**: required. Validated by `ApiKeyAuthenticationMiddleware`.
 - **Form fields**:
-  - `wav` (`IFormFile`, required): the WAV bytes. File name must end in `.wav`.
+  - `audioFile` (`IFormFile`, required): the audio bytes. File name must end in `.wav`, `.mp3`, or `.flac`.
   - `trackName` (string, required)
   - `artist` (string, required)
   - `album` (string, optional)
   - `genre` (string, optional)
   - `releaseDate` (string, optional, format `YYYY-MM-DD`)
   - `createdByUserId` (long, required): audit trail — who uploaded this track.
-- The upload stream is copied to a `.wav`-suffixed temp file under `Path.GetTempPath()` (the audio processor requires that extension and reads from disk). The temp file is always deleted in a `finally` block — success or failure.
-- `[RequestSizeLimit(1 GB)]` + `[RequestFormLimits(MultipartBodyLengthLimit = 1 GB)]` lift the per-request ceiling above the framework default (~28 MB) so production-sized WAVs are accepted. The body is streamed to the temp file, not buffered in memory.
-- Calls `UnifiedTrackService.UploadAsync`, which orchestrates: `TrackService.AddTrackFromWavAsync` (vault write) → `TrackManager` (SQL persist with `createdByUserId`).
-- Returns 200 with the **persisted** `TrackDto` JSON (Id populated) on success. Returns 400 for missing/invalid form fields. Returns 500 if processing fails.
+- The upload stream is copied to a temp file under `Path.GetTempPath()` with the appropriate extension (`.wav`, `.mp3`, or `.flac`). The audio processor reads from disk and requires the correct extension for format detection. The temp file is always deleted in a `finally` block — success or failure.
+- `[RequestSizeLimit(1 GB)]` + `[RequestFormLimits(MultipartBodyLengthLimit = 1 GB)]` lift the per-request ceiling above the framework default (~28 MB) so production-sized files are accepted. The body is streamed to the temp file, not buffered in memory.
+- Calls `UnifiedTrackService.UploadAsync`, which orchestrates: `TrackContentService.AddTrackAsync` (format-agnostic vault write via router) → `TrackManager` (SQL persist with `createdByUserId`).
+- Returns 200 with the **persisted** `TrackDto` JSON (Id populated) on success. Returns 400 for missing/invalid form fields or unsupported audio format. Returns 500 if processing fails.
 
 ### DELETE api/track/{id:long} ([ApiKeyAuthorize])
 
diff --git a/DeepDrftContent/CLAUDE.md b/DeepDrftContent/CLAUDE.md
index 62205e8..0a1a35e 100644
--- a/DeepDrftContent/CLAUDE.md
+++ b/DeepDrftContent/CLAUDE.md
@@ -74,19 +74,18 @@ public async Task<bool> RegisterResourceAsync(string vaultId, string entryId, Fi
 
 **Callers must check return values.** Do not change this without a deliberate design pass — it's embedded in all FileDatabase tests and client code.
 
-## Audio processor
+## Audio processors
 
-`AudioProcessor.ProcessWavFileAsync(filePath)`:
+Multi-format support via router pattern. All processors live in `DeepDrftContent/Processors/`:
 
-1. Validates the RIFF/WAVE structure and format code.
-2. Accepts standard PCM (audioFormat=1) and WAVE_FORMAT_EXTENSIBLE (audioFormat=0xFFFE) when the SubFormat GUID indicates PCM.
-3. Normalizes EXTENSIBLE-PCM uploads to standard 44-byte PCM WAV before storing in the vault.
-4. Parses the fmt and data chunks.
-5. Extracts duration (sample count / sample rate) and bitrate (file size / duration).
-6. Returns `AudioBinary` with all metadata.
-7. **Fallback**: If parsing fails, logs a warning and returns defaults (180s / 1411 kbps / 44.1 kHz / 16-bit stereo).
+- `AudioProcessor.ProcessWavFileAsync(filePath)`: WAV-specific processor. Validates RIFF/WAVE structure and format code. Accepts standard PCM (audioFormat=1) and WAVE_FORMAT_EXTENSIBLE (audioFormat=0xFFFE) when the SubFormat GUID indicates PCM. Normalizes EXTENSIBLE-PCM uploads to standard 44-byte PCM WAV before storing. Parses fmt and data chunks; extracts duration and bitrate. Returns `AudioBinary` with metadata. On parse failure, logs warning and returns defaults (180s / 1411 kbps / 44.1 kHz / 16-bit stereo). Accepts standard PCM (audioFormat=1), WAVE_FORMAT_EXTENSIBLE with PCM SubFormat (0x0001), IEEE Float SubFormat (0x0003), and Padded 24-in-32 containers; normalizes Float and padded inputs to standard 24-bit PCM before storage.
+- `Mp3AudioProcessor.ProcessMp3FileAsync(filePath)`: MP3 processor. Skips ID3v2 tag, finds first valid MPEG frame sync, decodes frame header (bitrate, sample rate, channels). Reads Xing/Info header for VBR total-frame count (accurate duration); VBRI header as fallback; CBR estimate from file size otherwise. Returns `AudioBinary` with original bytes and `.mp3` extension. On parse failure, falls back to defaults (180s / 320 kbps).
+- `FlacAudioProcessor.ProcessFlacFileAsync(filePath)`: FLAC processor. Validates `fLaC` magic, reads STREAMINFO metadata block (20-bit sample rate, 3-bit channels, 5-bit bits-per-sample, 36-bit total samples — all bit-packed). Computes duration from `totalSamples / sampleRate`; average bitrate from file size. Returns `AudioBinary` with original bytes and `.flac` extension. On parse failure, falls back to defaults (180s / 1411 kbps).
+- `AudioProcessorRouter.ProcessAudioFileAsync(filePath)`: Routes by extension — `.wav` → `AudioProcessor`, `.mp3` → `Mp3AudioProcessor`, `.flac` → `FlacAudioProcessor`. Throws `ArgumentException` for unsupported extensions.
 
-PCM-only. Accepts standard PCM (audioFormat=1), WAVE_FORMAT_EXTENSIBLE (audioFormat=0xFFFE) with PCM SubFormat (0x0001), IEEE Float SubFormat (0x0003), and Padded 24-in-32 containers (wValidBitsPerSample=24 in a 32-bit container). Float and padded-container inputs are normalized to standard 24-bit PCM at storage time via `ConvertFloatTo24BitPcm` and `RepackPaddedContainer` respectively. Other formats (mp3, flac, aac, ogg, m4a) are listed in `MimeTypeExtensions` but not implemented. The processor validates RIFF/WAVE structure — anything else is rejected.
+Vault stores original bytes with correct extension and MIME type (inferred from file extension or content-type header at upload time).
+
+The primary entry point is `TrackContentService.AddTrackAsync(filePath, mimeType)` — format-agnostic. It selects the right processor via `AudioProcessorRouter`, processes the file, generates an entry GUID, stores in vault, returns unpersisted `TrackEntity`. Legacy `AddTrackFromWavAsync(filePath)` is now a shim over `AddTrackAsync` for backward compatibility.
 
 ## Image processor
 
diff --git a/DeepDrftPublic.Client/CLAUDE.md b/DeepDrftPublic.Client/CLAUDE.md
index 4a4fc61..a9d9a85 100644
--- a/DeepDrftPublic.Client/CLAUDE.md
+++ b/DeepDrftPublic.Client/CLAUDE.md
@@ -61,18 +61,27 @@ Both are configured with JSON serializer settings (case-insensitive property mat
 ### Implementation
 - `AudioPlayerService` (abstract base): Lifecycle. Stores current track, playback state, volume. `SelectTrack` throws `NotSupportedException` (buffered path is dead); derived classes override `SelectTrackStreaming`.
 - `StreamingAudioPlayerService` (production): Constructor takes `TrackMediaClient`, `AudioInteropService`, logger. `SelectTrackStreaming`:
-  1. Calls `TrackMediaClient.GetAudioStreamAsync(trackId)`.
-  2. `StreamingAudioPlayerService.StreamAudioAsync` reads chunks (16–64 KB adaptive), pushes each via `AudioInteropService.ProcessStreamingChunkAsync` (JS interop call).
-  3. TypeScript `StreamDecoder` parses WAV header (first chunk), decodes subsequent chunks to `AudioBuffer`s.
+  1. Calls `TrackMediaClient.GetAudioStreamAsync(trackId)`, which returns a response object including `ContentType` (e.g., `audio/wav`, `audio/mpeg`, `audio/flac`).
+  2. `StreamingAudioPlayerService.StreamAudioAsync` reads chunks (16–64 KB adaptive), pushes each via `AudioInteropService.ProcessStreamingChunkAsync(contentType, chunk)` (JS interop call with format hint).
+  3. TypeScript `StreamDecoder` is format-agnostic; delegates format-specific header parsing and chunked decoding to the appropriate `IFormatDecoder` implementation (e.g., `WavFormatDecoder` for WAV, TBD MP3/FLAC decoders for other formats). Decoder parses header (first chunk), decodes subsequent chunks to `AudioBuffer`s.
   4. `PlaybackScheduler` schedules buffers on Web Audio `AudioContext`.
   5. Playback starts as soon as a configurable min buffer count is queued.
-  6. **Seek beyond buffer**: if seek target is past the decoded range, `Seek(position)` calls `TrackMediaClient.GetAudioStreamAsync(trackId, byteOffset)` with a file-absolute byte offset. Client sends `Range: bytes={offset}-`; server responds 206 with raw PCM; decoder retains the parsed WAV header and feeds the continuation directly into the decode pipeline.
+  6. **Seek beyond buffer**: if seek target is past the decoded range, `Seek(position)` calls `TrackMediaClient.GetAudioStreamAsync(trackId, byteOffset)` with a file-absolute byte offset. Client sends `Range: bytes={offset}-`; server responds 206 with raw bytes (same format as original file); decoder retains the parsed header and feeds the continuation directly into the decode pipeline.
 
 ### Interop bridge
 - `AudioInteropService.CreatePlayerAsync` polls `DeepDrftAudio.isReady()` before proceeding; `index.ts` sets `ready = true` after attaching the API to `window`. This guards against slow WASM boot / cache misses.
-- `AudioInteropService.ProcessStreamingChunkAsync(chunk)` calls JS `window.DeepDrftAudio.processStreamingChunk(chunk)` and awaits the Promise.
+- `AudioInteropService.ProcessStreamingChunkAsync(contentType, chunk)` calls JS `window.DeepDrftAudio.processStreamingChunk(contentType, chunk)` and awaits the Promise. The `contentType` parameter is passed through to the format-decoder factory.
 - `AudioInteropService` also manages callback registrations for progress (fired by `PlaybackScheduler`), end-of-playback (fired by `PlaybackScheduler`), and spectrum data (fired by `SpectrumAnalyzer`). Each callback is a `DotNetObjectReference` to a delegate.
 
+### Format decoders (TypeScript)
+New modules in `DeepDrftPublic/Interop/audio/`:
+
+- `IFormatDecoder.ts`: Interface. Defines contract for format-specific decoders: `parseHeader(chunk, offset)` → header metadata; `decodeChunk(chunk, offset)` → `AudioBuffer`.
+- `WavFormatDecoder.ts`: Concrete WAV implementation. Parses RIFF/WAVE structure, fmt and data chunks. All WAV-specific byte-parsing logic lives here. Exported as the default WAV decoder.
+- Future decoders: MP3, FLAC, etc. (TBD).
+
+`StreamDecoder.ts` remains the orchestrator — it accepts the first chunk, selects the right format decoder via factory (based on `contentType`), delegates all format-specific work to it, and chains subsequent chunks through the same decoder instance.
+
 ### Component integration
 - `AudioPlayerProvider.razor` is the cascading host. It injects `IStreamingPlayerService` (resolved to `StreamingAudioPlayerService` in DI), stores it in a cascade with `IsFixed="true"`, and keeps it alive across navigation.
 - `AudioPlayerBar.razor` is the dock UI. It cascades the player, binds buttons to `Play()` / `Pause()` / `Seek()` / `SetVolume()`, and displays current time / duration / progress bar. Minimize-state mutations (`Expand`, `ToggleMinimized`, `Close`) all route through a private `SetMinimized(bool value)` mutator, which guards no-ops, fires the `OnMinimized` callback, and calls `StateHasChanged()`. Subscribes to `IPlayerService.StateChanged` in `OnParametersSet` (reference-guarded, idempotent) and unsubscribes on dispose to re-render itself when the cascade updates.
diff --git a/PLAN.md b/PLAN.md
index 6c816e1..49edd1a 100644
--- a/PLAN.md
+++ b/PLAN.md
@@ -32,6 +32,7 @@ These were flagged during the audit but classified as feature work, not defect f
   - Client side: the JS decoder is currently a WAV byte-walker. For non-WAV, the simplest path is `decodeAudioData` over the full payload (loses streaming-start). The richer path is per-format chunked decoders. Worth a design pass before committing.
 - **Prerequisite:** None functionally, but consider settling **Phase 4 (HTTP Range)** first — native range/cache is much more important for large MP3s than for WAVs.
 - **Constraint:** Spectrum FFT tap currently relies on raw `AudioBuffer`s through `decodeAudioData`. If a future path uses `MediaElementAudioSourceNode` (see 4.1), the FFT tap still works but the early-playback story changes.
+- **Progress:** Wave 1 (foundation) landed on `dev` (2026-06-10). Format router (`Mp3AudioProcessor`, `FlacAudioProcessor`, `AudioProcessorRouter`) and client-side `IFormatDecoder` interface + `WavFormatDecoder` implementation are in place. `TrackContentService.AddTrackAsync` is the new format-agnostic primary method; `AddTrackFromWavAsync` is a shim. Client `StreamDecoder` is now format-agnostic; WAV-specific logic moved to `WavFormatDecoder`. Wave 2 (per-format chunked decoders) and Wave 3 (full wiring) pending.
 
 ### 1.3 Preload / prefetch of the next track