Hvordan bruke fra text til lydfil

Question

Hvordan bruke fra text til lydfil

Ja Erv 0

Hvordan bruke AZUR Speech. Jeg får ikke lagret lydfilen etter å ha laget fra tekst til lydfil

Ja Erv 0 Reputation points

2025-11-27T21:00:20.9833333+00:00

You have to solve the issue by simply make a button which convert text to .wave and save it on the MAC in a folder named Project*

Sridhar M 2,525 Microsoft External Staff Moderator

Hei Ja Erv

Takk for at du kontaktet oss på Microsoft Q&A-forumet.

La oss lage én knapp som tar tekst, konverterer den til WAV og lagrer den på Mac-en din under ~/Documents/Project*

Alternativ A — macOS-app (SwiftUI + REST, ingen SDK-installasjon)

Dette bruker Azure Speech REST API direkte (enklere enn å legge til hele Speech SDK-en). Du får en tekstboks og én knapp: «Konverter til WAV». Appen oppretter ~/Documents/Project* (hvis den mangler) og skriver output.wav.

Nødvendige trinn:

Azure Speech-nøkkelen og regionen din (f.eks. eastus).

Et macOS-utviklingsoppsett (Xcode).

Opprett prosjektet:

Åpne Xcode → Fil → Ny → Prosjekt… → App (macOS / SwiftUI).

Gi det navnet f.eks. AzureTTStoWav.

Koden (SwiftUI):

Erstatt YOUR_SPEECH_KEY og YOUR_REGION før du kjører.

import SwiftUI

struct ContentView: View {
    @State private var inputText: String = "Hello from Azure Speech on macOS!"
    @State private var status: String = ""
    @State private var isWorking: Bool = false
    var body: some View {
        VStack(alignment: .leading, spacing: 12) {
            Text("Text → WAV (Azure Speech)")
                .font(.headline)
            TextEditor(text: $inputText)
                .font(.system(size: 13, design: .monospaced))
                .frame(minHeight: 160)
                .border(Color.gray.opacity(0.3))
            HStack {
                Button(action: convertToWav) {
                    Text(isWorking ? "Converting…" : "Convert to WAV")
                }
                .disabled(isWorking)
                Spacer()
            }
            Text(status)
                .foregroundColor(.secondary)
                .font(.footnote)
                .lineLimit(3)
        }
        .padding(16)
        .frame(minWidth: 560, minHeight: 320)
    }
    private func convertToWav() {
        guard !inputText.trimmingCharacters(in: .whitespacesAndNewlines).isEmpty else {
            status = "Please enter some text."
            return
        }
        isWorking = true
        status = "Calling Azure Speech…"
        // ── Replace these with your credentials ──
        let speechKey = "YOUR_SPEECH_KEY"
        let region = "YOUR_REGION" // e.g., "eastus", "uksouth", "southeastasia"
        // Azure TTS endpoint for region
        let endpoint = "https://\(region).tts.speech.microsoft.com/cognitiveservices/v1"
        // Choose WAV (RIFF PCM) output
        // Other formats: https://learn.microsoft.com/azure/ai-services/speech-service/rest-text-to-speech#audio-outputs
        let outputFormat = "riff-24khz-16bit-mono-pcm"
        // Basic SSML payload (you can change voice)
        let voiceName = "en-US-JennyNeural"
        let ssml =
        """
        <speak version="1.0" xml:lang="en-US">
          <voice xml:lang="en-US" name="\(voiceName)">
            \(xmlEscaped(inputText))
          </voice>
        </speak>
        """
        // Build request
        guard let url = URL(string: endpoint) else {
            status = "Invalid endpoint URL."
            isWorking = false
            return
        }
        var req = URLRequest(url: url)
        req.httpMethod = "POST"
        req.setValue(speechKey, forHTTPHeaderField: "Ocp-Apim-Subscription-Key")
        req.setValue("application/ssml+xml", forHTTPHeaderField: "Content-Type")
        req.setValue(outputFormat, forHTTPHeaderField: "X-Microsoft-OutputFormat")
        req.setValue("AzureTTS-macOS-App", forHTTPHeaderField: "User-Agent")
        req.httpBody = ssml.data(using: .utf8)
        let task = URLSession.shared.dataTask(with: req) { data, response, error in
            DispatchQueue.main.async {
                defer { self.isWorking = false }
                if let error = error {
                    self.status = "Network error: \(error.localizedDescription)"
                    return
                }
                guard let http = response as? HTTPURLResponse else {
                    self.status = "No HTTP response."
                    return
                }
                guard http.statusCode == 200, let audioData = data, !audioData.isEmpty else {
                    let body = String(data: data ?? Data(), encoding: .utf8) ?? "(no body)"
                    self.status = "Azure TTS failed (\(http.statusCode)).\n\(body)"
                    return
                }
                // Save to ~/Documents/Project*/output.wav
                do {
                    let saveURL = try Self.makeProjectFolderAndFileURL(fileName: "output.wav")
                    try audioData.write(to: saveURL, options: .atomic)
                    self.status = "Saved: \(saveURL.path)"
                } catch {
                    self.status = "Save error: \(error.localizedDescription)"
                }
            }
        }
        task.resume()
    }
    private static func makeProjectFolderAndFileURL(fileName: String) throws -> URL {
        let docs = FileManager.default.urls(for: .documentDirectory, in: .userDomainMask).first!
        let projectFolder = docs.appendingPathComponent("Project*", isDirectory: true)
        if !FileManager.default.fileExists(atPath: projectFolder.path) {
            try FileManager.default.createDirectory(at: projectFolder, withIntermediateDirectories: true, attributes: nil)
        }
        return projectFolder.appendingPathComponent(fileName)
    }
    // Minimal XML escape for SSML safety
    private func xmlEscaped(_ s: String) -> String {
        var r = s
        r = r.replacingOccurrences(of: "&", with: "&amp;")
        r = r.replacingOccurrences(of: "<", with: "&lt;")
        r = r.replacingOccurrences(of: ">", with: "&gt;")
        r = r.replacingOccurrences(of: "\"", with: "&quot;")
        r = r.replacingOccurrences(of: "'", with: "&apos;")
        return r
    }
}
#Preview {
    ContentView()

Kjør det:

Bygg og kjør.
Skriv inn tekst → klikk på Konverter til WAV.
Sjekk ~/Documents/Project*/output.wav.

Merknader:

Hvis du ser en 401/403, bekreft nøkkel/region og at taleressursen din er aktiv og ikke begrenset.
Du kan endre outputFormat til andre formater (MP3 osv.) senere.

Alternativ B — Ett klikk fra Terminal (ingen Xcode)

Hvis du bare vil ha en knapplignende opplevelse, kan du bruke en skallfunksjon og kalle den når som helst. Den legger ut til Azure og skriver en WAV under ~/Documents/Project*/output.wav.

Erstatt YOUR_SPEECH_KEY / YOUR_REGION og stemmenavn etter behov.

#!/usr/bin/env bash
# save as: tts_to_wav.sh ; then: chmod +x tts_to_wav.sh
# usage: ./tts_to_wav.sh "Your text here"
TEXT="$1"
SPEECH_KEY="YOUR_SPEECH_KEY"
REGION="YOUR_REGION"     # e.g., eastus
VOICE="en-US-JennyNeural"
FORMAT="riff-24khz-16bit-mono-pcm"
ENDPOINT="https://${REGION}.tts.speech.microsoft.com/cognitiveservices/v1"
# Ensure ~/Documents/Project*
TARGET_DIR="${HOME}/Documents/Project*"
mkdir -p "${TARGET_DIR}"
OUT="${TARGET_DIR}/output.wav"
# Build SSML (basic)
SSML="<?xml version='1.0'?>
<speak version='1.0' xml:lang='en-US'>
  <voice xml:lang='en-US' name='${VOICE}'>${TEXT}</voice>
</speak>"
# Call Azure TTS
curl -s -X POST "${ENDPOINT}" \
  -H "Ocp-Apim-Subscription-Key: ${SPEECH_KEY}" \
  -H "Content-Type: application/ssml+xml" \
  -H "X-Microsoft-OutputFormat: ${FORMAT}" \
  -H "User-Agent: AzureTTS-macOS-Script" \
  --data "${SSML}" \
  --output "${OUT}"
if [ -s "${OUT}" ]; then
  echo "Saved: ${OUT}"
else
  echo "Failed (empty or error).  echo "Failed (empty or error). Check your key/region/quotas."

Kjør:

./tts_to_wav.sh "Hello from Azure Speech!"

Reference:

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/how-to-speech-synthesis?tabs=browserjs%2Cterminal&pivots=programming-language-csharp

https://learn.microsoft.com/en-us/azure/ai-services/speech-service/rest-text-to-speech?tabs=streaming

Jeg håper dette hjelper. Gi meg beskjed hvis du har flere spørsmål.

Takk!

Sridhar M 2,525 Reputation points Microsoft External Staff Moderator

2025-11-28T19:50:45.4366667+00:00

Hei Ja Erv

Fikk du mulighet til å se over svaret ovenfor? Takk!
Sridhar M 2,525 Reputation points Microsoft External Staff Moderator

2025-12-01T16:20:39.09+00:00

Hei Ja Erv

Vi har ikke hørt fra deg på det siste svaret, og ville bare sjekke om du har en løsning ennå. Hvis du har en løsning, del den med fellesskapet, da den kan være nyttig for andre. Ellers vil vi svare med mer informasjon, og vi vil prøve å hjelpe.

1 answer

Your answer

Ja Erv 0 Reputation points

2025-11-27T21:00:20.9833333+00:00

You have to solve the issue by simply make a button which convert text to .wave and save it on the MAC in a folder named Project*
Sridhar M 2,525 Reputation points Microsoft External Staff Moderator

2025-11-28T19:50:45.4366667+00:00

Hei Ja Erv

Fikk du mulighet til å se over svaret ovenfor? Takk!
Sridhar M 2,525 Reputation points Microsoft External Staff Moderator

2025-12-01T16:20:39.09+00:00

Hei Ja Erv

Vi har ikke hørt fra deg på det siste svaret, og ville bare sjekke om du har en løsning ennå. Hvis du har en løsning, del den med fellesskapet, da den kan være nyttig for andre. Ellers vil vi svare med mer informasjon, og vi vil prøve å hjelpe.

Answer 1

To convert text to an audio file using Azure Speech, you can use the Speech Synthesis capabilities provided by Azure. Here are the general steps you can follow:

Set Up Azure Speech Service: Ensure you have an Azure subscription and have created a Speech resource in the Azure portal. You will need the Speech resource key and region.
Use the Speech Synthesis API: You can use the SpeakTextAsync operation to synthesize speech from text. This allows you to convert a short block of text into audio. For longer texts, consider using the batch synthesis API.
Save the Audio File: To save the audio output as a file, you need to specify the output format when calling the synthesis function. You can use formats like WAV or MP3 depending on your requirements.

Example Code: Here’s a simple example in C# to get you started:

   var config = SpeechConfig.FromSubscription("YourSubscriptionKey", "YourServiceRegion");
   using (var synthesizer = new SpeechSynthesizer(config))
   {
       using (var result = await synthesizer.SpeakTextAsync("Hello, this is a test."))
       {
           if (result.Reason == ResultReason.SynthesizingAudioCompleted)
           {
               // Save the audio to a file
           }
           else
           {
               // Handle error
           }
       }
   }

Check Documentation: For more detailed information on how to synthesize speech and save it as an audio file, refer to the Azure documentation on speech synthesis and the Speech Synthesis Markup Language (SSML).

If you are facing issues saving the audio file, ensure that you are correctly handling the output stream and specifying the correct file path and format.

References:

Share via

Hvordan bruke fra text til lydfil

1 answer

Your answer