CSDK-2246 daily-python: add Deepgram TTS example

Author: aconchillo

demos/README.md

@@ -3,6 +3,7 @@
 Here you can find a few demos that use Daily's Python SDK:
 
 - **audio**: Examples on how to send and receive RAW audio or WAV files.
+- **deepgram**: An example showing how to use Deepgram [Text-To-Speech](https://developers.deepgram.com/docs/text-to-speech) API.
 - **flask**: A demo that uses [Flask](https://flask.palletsprojects.com/) and [Celery](https://docs.celeryq.dev/) to launch multiple concurrent audio bots.
 - **google**: Audio examples using Google [Speech-To-Text](https://cloud.google.com/speech-to-text) and [Text-To-Speech](https://cloud.google.com/text-to-speech) APIs.
 - **gstreamer**: A media player based on [GStreamer](https://gstreamer.freedesktop.org/) that sends a video file into a meeting.

demos/deepgram/deepgram_text_to_speech.py

@@ -0,0 +1,91 @@
+#
+# This demo will join a Daily meeting and, given a text file with senteces (one
+# per line), will translate text into audio using Deepgram's Text-To-Speech API
+# and will send it into the meeting.
+#
+# The demo requires a Deepgram API key set in the DG_API_KEY environment variable.
+#
+# See https://developers.deepgram.com/docs/text-to-speech
+#
+# Usage: python3 deepgram_speech_to_text.py -m MEETING_URL -i FILE
+#
+
+import argparse
+import os
+import time
+
+from daily import *
+from deepgram import (
+    DeepgramClient,
+    SpeakOptions,
+)
+
+parser = argparse.ArgumentParser()
+parser.add_argument("-m", "--meeting", required=True, help="Meeting URL")
+parser.add_argument(
+    "-i",
+    "--input",
+    required=True,
+    help="File with sentences (one per line)")
+args = parser.parse_args()
+
+Daily.init()
+
+# We create a virtual microphone device so we can read audio samples from the
+# meeting.
+microphone = Daily.create_microphone_device(
+    "my-mic", sample_rate=16000, channels=1)
+
+client = CallClient()
+
+print()
+print(f"Joining {args.meeting} ...")
+
+# Join and tell our call client that we will be using our new virtual
+# microphone.
+client.join(args.meeting, client_settings={
+    "inputs": {
+        "microphone": {
+            "isEnabled": True,
+            "settings": {
+                "deviceId": "my-mic"
+            }
+        }
+    }
+})
+
+# Make sure we are joined. It would be better to use join() completion
+# callback.
+time.sleep(3)
+
+sentences_file = open(args.input, "r")
+
+deepgram = DeepgramClient(api_key=os.getenv("DG_API_KEY"))
+
+speak_options = SpeakOptions(
+    model="aura-asteria-en",
+    encoding="linear16",
+    sample_rate="16000",
+    container="none"
+)
+
+print()
+
+for sentence in sentences_file.readlines():
+    print(f"Processing: {sentence.strip()}")
+    print()
+
+    speak_source = {
+        "text": sentence.strip()
+    }
+
+    response = deepgram.speak.v("1").stream(speak_source, speak_options)
+
+    # Send all the audio frames to the microphone.
+    microphone.write_frames(response.stream.read())
+
+# Let everything finish
+time.sleep(2)
+
+client.leave()
+client.release()

demos/google/google_text_to_speech.py

@@ -7,7 +7,7 @@
 #
 # See https://cloud.google.com/text-to-speech/docs/before-you-begin
 #
-# Usage: python3 google_speech_to_text.py -m MEETING_URL
+# Usage: python3 google_speech_to_text.py -m MEETING_URL -i FILE
 #
 
 from daily import *

demos/requirements.txt

@@ -1,4 +1,5 @@
 celery==5.3.6
+deepgram-sdk==3.2.1
 flask==3.0.1
 google-cloud-speech==2.24.0
 google-cloud-texttospeech==2.16.0