From ec09505f6bed113a6b64586e68d5884c550c1af3 Mon Sep 17 00:00:00 2001
From: Mark Backman <mark@daily.co>
Date: Thu, 3 Jul 2025 13:14:15 -0700
Subject: [PATCH] FishAudioTTSService: Add normalize as InputParam, model_id as
 arg

---
 CHANGELOG.md                     |  9 ++++++---
 src/pipecat/services/fish/tts.py | 19 +++++++++++++++++++
 2 files changed, 25 insertions(+), 3 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 696dd99fd..79b9c04f0 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,6 +9,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Added
 
+- Added `normalize` and `model_id` to `FishAudioTTSService`.
+
 - Added `run_llm` field to `LLMMessagesAppendFrame` and `LLMMessagesUpdateFrame`
   frames. If true, a context frame will be pushed triggering the LLM to respond.
 
@@ -85,9 +87,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Deprecated
 
-- In `FishTTSService`, deprecated `model` and replaced with `reference_id`.
-  This change is to better align with Fish Audio's variable naming and to
-  reduce confusion about what functionality the variable controls.
+- In `FishAudioTTSService`, deprecated `model` and replaced with
+  `reference_id`. This change is to better align with Fish Audio's variable
+  naming and to reduce confusion about what functionality the variable
+  controls.
 
 ## [0.0.73] - 2025-06-26
 
diff --git a/src/pipecat/services/fish/tts.py b/src/pipecat/services/fish/tts.py
index 9d5af657e..598bac0f3 100644
--- a/src/pipecat/services/fish/tts.py
+++ b/src/pipecat/services/fish/tts.py
@@ -58,12 +58,14 @@ class FishAudioTTSService(InterruptibleTTSService):
         Parameters:
             language: Language for synthesis. Defaults to English.
             latency: Latency mode ("normal" or "balanced"). Defaults to "normal".
+            normalize: Whether to normalize audio output. Defaults to True.
             prosody_speed: Speech speed multiplier (0.5-2.0). Defaults to 1.0.
             prosody_volume: Volume adjustment in dB. Defaults to 0.
         """
 
         language: Optional[Language] = Language.EN
         latency: Optional[str] = "normal"  # "normal" or "balanced"
+        normalize: Optional[bool] = True
         prosody_speed: Optional[float] = 1.0  # Speech speed (0.5-2.0)
         prosody_volume: Optional[int] = 0  # Volume adjustment in dB
 
@@ -73,6 +75,7 @@ class FishAudioTTSService(InterruptibleTTSService):
         api_key: str,
         reference_id: Optional[str] = None,  # This is the voice ID
         model: Optional[str] = None,  # Deprecated
+        model_id: str = "speech-1.5",
         output_format: FishAudioOutputFormat = "pcm",
         sample_rate: Optional[int] = None,
         params: Optional[InputParams] = None,
@@ -89,6 +92,7 @@ class FishAudioTTSService(InterruptibleTTSService):
                 The `model` parameter is deprecated and will be removed in version 0.1.0.
                 Use `reference_id` instead to specify the voice model.
 
+            model_id: Specify which Fish Audio TTS model to use (e.g. "speech-1.5")
             output_format: Audio output format. Defaults to "pcm".
             sample_rate: Audio sample rate. If None, uses default.
             params: Additional input parameters for voice customization.
@@ -134,6 +138,7 @@ class FishAudioTTSService(InterruptibleTTSService):
             "sample_rate": 0,
             "latency": params.latency,
             "format": output_format,
+            "normalize": params.normalize,
             "prosody": {
                 "speed": params.prosody_speed,
                 "volume": params.prosody_volume,
@@ -141,6 +146,8 @@ class FishAudioTTSService(InterruptibleTTSService):
             "reference_id": reference_id,
         }
 
+        self.set_model_name(model_id)
+
     def can_generate_metrics(self) -> bool:
         """Check if this service can generate processing metrics.
 
@@ -149,6 +156,17 @@ class FishAudioTTSService(InterruptibleTTSService):
         """
         return True
 
+    async def set_model(self, model: str):
+        """Set the TTS model and reconnect.
+
+        Args:
+            model: The model name to use for synthesis.
+        """
+        await super().set_model(model)
+        logger.info(f"Switching TTS model to: [{model}]")
+        await self._disconnect()
+        await self._connect()
+
     async def start(self, frame: StartFrame):
         """Start the Fish Audio TTS service.
 
@@ -197,6 +215,7 @@ class FishAudioTTSService(InterruptibleTTSService):
 
             logger.debug("Connecting to Fish Audio")
             headers = {"Authorization": f"Bearer {self._api_key}"}
+            headers["model"] = self.model_name
             self._websocket = await websockets.connect(self._base_url, extra_headers=headers)
 
             # Send initial start message with ormsgpack