Hey! I am using Open Ai text-to-speech API to process the serial input text and convert it into audio. The request is sent to the API, and the audio stream data is received, but while processing the data, I get a Processing timed-out error. I initially had the timeout at 30 seconds and the buffer size of the audio at 1046; now, at 3 minutes of timeout and 4096 as buffer size, I’m still getting the same error. I am confused as to why I keep getting the same error even after that much processing time.
For your information, I’m using the ESP-Wroom-32 Module and Open AI’s Text-to-Speech API.
The goal is to process the audio file and output it through the Bluetooth device that I have.
#include <WiFi.h>
#include <WiFiClientSecure.h>
#include <HTTPClient.h>
#include <ArduinoJson.h>
#include <BluetoothA2DPSink.h>
// Replace with your network credentials
const char* ssid = "ssid";
const char* password = "password";
const char* bluetoothDeviceName = "bluetooth_speaker";
// Replace with your OpenAI API key
const char* serverName = "https://api.openai.com/v1/audio/speech";
const char* apiKey = "API KEY";
// Bluetooth settings
BluetoothA2DPSink a2dp_sink;
WiFiClientSecure client;
HTTPClient http;
const char* root_ca = \
"-----BEGIN CERTIFICATE-----\n" \
"MIICCTCCAY6gAwIBAgINAgPlwGjvYxqccpBQUjAKBggqhkjOPQQDAzBHMQswCQYD\n" \
"VQQGEwJVUzEiMCAGA1UEChMZR29vZ2xlIFRydXN0IFNlcnZpY2VzIExMQzEUMBIG\n" \
"A1UEAxMLR1RTIFJvb3QgUjQwHhcNMTYwNjIyMDAwMDAwWhcNMzYwNjIyMDAwMDAw\n" \
"WjBHMQswCQYDVQQGEwJVUzEiMCAGA1UEChMZR29vZ2xlIFRydXN0IFNlcnZpY2Vz\n" \
"IExMQzEUMBIGA1UEAxMLR1RTIFJvb3QgUjQwdjAQBgcqhkjOPQIBBgUrgQQAIgNi\n" \
"AATzdHOnaItgrkO4NcWBMHtLSZ37wWHO5t5GvWvVYRg1rkDdc/eJkTBa6zzuhXyi\n" \
"QHY7qca4R9gq55KRanPpsXI5nymfopjTX15YhmUPoYRlBtHci8nHc8iMai/lxKvR\n" \
"HYqjQjBAMA4GA1UdDwEB/wQEAwIBhjAPBgNVHRMBAf8EBTADAQH/MB0GA1UdDgQW\n" \
"BBSATNbrdP9JNqPV2Py1PsVq8JQdjDAKBggqhkjOPQQDAwNpADBmAjEA6ED/g94D\n" \
"9J+uHXqnLrmvT/aDHQ4thQEd0dlq7A/Cr8deVl5c1RxYIigL9zC2L7F8AjEA8GE8\n" \
"p/SgguMh1YQdc4acLa/KNJvxn7kjNuK8YAOdgLOaVsjh4rsUecrNIdSUtUlD\n" \
"-----END CERTIFICATE-----\n" \
"";
std::vector<uint8_t> audioData;
String audioUrl;
bool getTTS(String text);
void streamAudio();
void audio_data_callback(const uint8_t *data, uint32_t length);
bool getTTS(String text) {
if (WiFi.status() == WL_CONNECTED) {
Serial.println("WiFi is connected. Sending request to OpenAI TTS...");
http.begin(client, serverName);
http.addHeader("Content-Type", "application/json");
http.addHeader("Authorization", "Bearer " + String(apiKey));
String jsonRequest = "{\"model\":\"tts-1\",\"input\":\"" + text + "\",\"voice\":\"alloy\"}";
Serial.println("Request JSON: " + jsonRequest);
int httpResponseCode = http.POST(jsonRequest);
if (httpResponseCode > 0) {
Serial.print("HTTP Response code: ");
Serial.println(httpResponseCode);
if (httpResponseCode == 200) {
Serial.println("Request sent. Processing response...");
WiFiClient *stream = http.getStreamPtr();
unsigned long startTime = millis();
const unsigned long timeout = 180000; // 60 seconds timeout
const int bufferSize = 4096; // Adjust buffer size as needed
std::vector<uint8_t> buffer(bufferSize);
while (stream->connected() || stream->available()) {
int available = stream->available();
if (available > 0) {
int bytesToRead = min(available, bufferSize);
int bytesRead = stream->readBytes(reinterpret_cast<char*>(buffer.data()), bytesToRead);
audioData.insert(audioData.end(), buffer.begin(), buffer.begin() + bytesRead);
// Reset the timeout timer
startTime = millis();
Serial.print("Bytes read: ");
Serial.println(bytesRead);
}
// Check for timeout
if (millis() - startTime > timeout) {
Serial.println("Response processing timed out.");
http.end();
return false;
}
}
Serial.print("Total audio data received: ");
Serial.println(audioData.size());
http.end();
return true;
} else {
Serial.println("Request failed. Check the response body.");
String response = http.getString();
Serial.println("Response body:");
Serial.println(response);
}
} else {
Serial.print("Error on sending POST: ");
Serial.println(httpResponseCode);
if (httpResponseCode == -1) {
Serial.println("Connection failed - Please check SSL setup or connection.");
} else {
Serial.println("Unexpected error.");
}
}
http.end();
} else {
Serial.println("WiFi Disconnected");
}
return false;
}
void audio_data_callback(const uint8_t *data, uint32_t length) {
if (!audioData.empty()) {
uint32_t bytesToWrite = min(length, (uint32_t)audioData.size());
memcpy((void *)data, audioData.data(), bytesToWrite);
audioData.erase(audioData.begin(), audioData.begin() + bytesToWrite);
}
}
void streamAudio() {
if (!audioData.empty()) {
a2dp_sink.set_stream_reader(audio_data_callback);
a2dp_sink.start("Bluetooth Speaker");
Serial.println("Streaming audio to Bluetooth speaker...");
} else {
Serial.println("No audio data available to stream.");
}
}
void setup() {
Serial.begin(115200);
delay(1000); // Wait for the serial monitor to open
// Connect to WiFi
Serial.print("Connecting to WiFi....");
WiFi.begin(ssid, password);
while (WiFi.status() != WL_CONNECTED) {
delay(1000);
Serial.print(".");
}
Serial.println("Connected to WiFi");
// Test HTTPS connection
client.setCACert(root_ca);
Serial.println("Testing HTTPS connection...");
if (!client.connect("api.openai.com", 443)) {
Serial.println("Connection to API failed!");
} else {
Serial.println("Connected to api.openai.com!");
}
// Test the TTS function
String prompt = "Hello";
if (getTTS(prompt)) {
streamAudio();
} else {
Serial.println("Failed to get TTS");
}
}
void loop() {
// Add any additional logic for your application here
}
