如何在 Android 中获得等效的 Google Text-to-Speech 音调

Question

I "created" a custom Google TTS voice by setting a pitch and speech rate here .

The value I chose for pitch is negative (-6.4) and I want to use the exact values in my Android app.

However, TextToSpeech.setPitch(float) cannot accept negative values. Here is the code from android.speech.tts.TextToSeech :

public int setPitch(float pitch) {
  if (pitch > 0.0f) {
    int intPitch = (int)(pitch * 100);
    if (intPitch > 0) {
      synchronized (mStartLock) {
        mParams.putInt(Engine.KEY_PARAM_PITCH, intPitch);
      }
      return SUCCESS;
    }
  }
  return ERROR;
}

I need some help determining the equivalent positive value that would be equivalent to -6.4. It appears I cannot simply extrapolate the value using the relative scales [-20, 20] (default: 0) on the web interface and apparently [25, 400] (default: 100) from the Android library. The latter obtained from com.android.settings.tts.TextToSpeechSettings , available here :

    /**
     * Speech pitch value. TTS pitch value varies from 25 to 400, where 100 is the value for normal
     * pitch. The max pitch value is set to 400, based on feedback from users and the GoogleTTS
     * pitch variation range. The range for pitch is not set in stone and should be readjusted based
     * on user need. This value should be kept in sync with the max value set in tts_settings xml.
     */
    private static final int MAX_SPEECH_PITCH = 400;
    private static final int MIN_SPEECH_PITCH = 25;

Thank you for your time.

Answer 1

I'm not sure how the default 0 (web interface) can equate to 100 (android library), but maybe this formula would work.

You can try using this formula:

A_MIN_SPEECH_PITCH = -20;

A_MAX_SPEECH_PITCH = 20;

B_MIN_SPEECH_PITCH = 25;

B_MAX_SPEECH_PITCH = 400;

a_speech_pitch = // Desired pitch

b_speech_pitch = (B_MAX_SPEECH_PITCH - B_MIN_SPEECH_PITCH) * ((a_speech_pitch - A_MIN_SPEECH_PITCH) / (A_MAX_SPEECH_PITCH - A_MIN_SPEECH_PITCH)) + B_MIN_SPEECH_PITCH;

Test Case #1

a_speech_pitch = -20

b_speech_pitch = (400-25) * ((-20-(-20))/(20-(-20))) + 25 = 25

Test Case #2

a_speech_pitch = 0

b_speech_pitch = (400-25) * ((0-(-20))/(20-(-20))) + 25 = 212.5

Test Case #3

a_speech_pitch = 20

b_speech_pitch = (400-25) * ((20-(-20))/(20-(-20))) + 25 = 400

Test Case #4

a_speech_pitch = -6.4

b_speech_pitch = (400-25) * ((-6.4-(-20))/(20-(-20))) + 25 = 152.5