From 8acf9a488bc15152be5104a711919323c1d93484 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Aleix=20Conchillo=20Flaqu=C3=A9?= <aleix@daily.co>
Date: Tue, 25 Feb 2025 12:30:54 -0800
Subject: [PATCH] tts: some small HTTP-based services improvements

---
 src/pipecat/services/elevenlabs.py |  8 +++++---
 src/pipecat/services/openai.py     |  4 +++-
 src/pipecat/services/playht.py     | 13 ++++++-------
 src/pipecat/services/rime.py       |  6 +++---
 src/pipecat/services/xtts.py       |  4 +++-
 5 files changed, 20 insertions(+), 15 deletions(-)

diff --git a/src/pipecat/services/elevenlabs.py b/src/pipecat/services/elevenlabs.py
index 3dc42975e..5e70735f8 100644
--- a/src/pipecat/services/elevenlabs.py
+++ b/src/pipecat/services/elevenlabs.py
@@ -570,10 +570,12 @@ class ElevenLabsHttpTTSService(TTSService):
 
                 await self.start_tts_usage_metrics(text)
 
-                yield TTSStartedFrame()
+                # Process the streaming response
+                CHUNK_SIZE = 1024
 
-                async for chunk in response.content:
-                    if chunk:
+                yield TTSStartedFrame()
+                async for chunk in response.content.iter_chunked(CHUNK_SIZE):
+                    if len(chunk) > 0:
                         await self.stop_ttfb_metrics()
                         yield TTSAudioRawFrame(chunk, self.sample_rate, 1)
         except Exception as e:
diff --git a/src/pipecat/services/openai.py b/src/pipecat/services/openai.py
index 6ed3b4612..e49bd3a90 100644
--- a/src/pipecat/services/openai.py
+++ b/src/pipecat/services/openai.py
@@ -530,8 +530,10 @@ class OpenAITTSService(TTSService):
 
                 await self.start_tts_usage_metrics(text)
 
+                CHUNK_SIZE = 1024
+
                 yield TTSStartedFrame()
-                async for chunk in r.iter_bytes(8192):
+                async for chunk in r.iter_bytes(CHUNK_SIZE):
                     if len(chunk) > 0:
                         await self.stop_ttfb_metrics()
                         frame = TTSAudioRawFrame(chunk, self.sample_rate, 1)
diff --git a/src/pipecat/services/playht.py b/src/pipecat/services/playht.py
index 56179e34c..c42a44eaa 100644
--- a/src/pipecat/services/playht.py
+++ b/src/pipecat/services/playht.py
@@ -383,8 +383,6 @@ class PlayHTHttpTTSService(TTSService):
 
         try:
             options = self._create_options()
-            b = bytearray()
-            in_header = True
 
             await self.start_ttfb_metrics()
 
@@ -396,6 +394,8 @@ class PlayHTHttpTTSService(TTSService):
 
             yield TTSStartedFrame()
 
+            b = bytearray()
+            in_header = True
             async for chunk in playht_gen:
                 # skip the RIFF header.
                 if in_header:
@@ -410,11 +410,10 @@ class PlayHTHttpTTSService(TTSService):
                             fh.read(size)
                             (data, size) = struct.unpack("<4sI", fh.read(8))
                         in_header = False
-                else:
-                    if len(chunk):
-                        await self.stop_ttfb_metrics()
-                        frame = TTSAudioRawFrame(chunk, self.sample_rate, 1)
-                        yield frame
+                elif len(chunk) > 0:
+                    await self.stop_ttfb_metrics()
+                    frame = TTSAudioRawFrame(chunk, self.sample_rate, 1)
+                    yield frame
         except Exception as e:
             logger.error(f"{self} error generating TTS: {e}")
         finally:
diff --git a/src/pipecat/services/rime.py b/src/pipecat/services/rime.py
index 60d2d67ef..007ba958e 100644
--- a/src/pipecat/services/rime.py
+++ b/src/pipecat/services/rime.py
@@ -407,10 +407,10 @@ class RimeHttpTTSService(TTSService):
                 yield TTSStartedFrame()
 
                 # Process the streaming response
-                chunk_size = 8192
+                CHUNK_SIZE = 1024
 
-                async for chunk in response.content.iter_chunked(chunk_size):
-                    if chunk:
+                async for chunk in response.content.iter_chunked(CHUNK_SIZE):
+                    if len(chunk) > 0:
                         await self.stop_ttfb_metrics()
                         frame = TTSAudioRawFrame(chunk, self.sample_rate, 1)
                         yield frame
diff --git a/src/pipecat/services/xtts.py b/src/pipecat/services/xtts.py
index 9d2f3dd1e..e275b50b0 100644
--- a/src/pipecat/services/xtts.py
+++ b/src/pipecat/services/xtts.py
@@ -150,8 +150,10 @@ class XTTSService(TTSService):
 
             yield TTSStartedFrame()
 
+            CHUNK_SIZE = 1024
+
             buffer = bytearray()
-            async for chunk in r.content.iter_chunked(1024):
+            async for chunk in r.content.iter_chunked(CHUNK_SIZE):
                 if len(chunk) > 0:
                     await self.stop_ttfb_metrics()
                     # Append new chunk to the buffer.