Speech Synthesis & Speech Recognition: Using SAPI 5.1 Using TTS with Delphi CreateOLEObject('SAPI.SpVoice');

 Using TTS with Delphi
https://en.wikipedia.org/wiki/Microsoft_Speech_API

https://en.wikipedia.org/wiki/SASDK

CreateOLEObject('SAPI.SpVoice');

voz := CreateOLEObject('SAPI.SpVoice');
if not VarIsEmpty(voz) then begin
 vozes := voz.getVoices;
 ComboVoz.Clear;
 for i := 0 to vozes.Count - 1 do
  ComboVoz.Items.Add(vozes.item(i).GetDescription);

  Dialogs, StdCtrls, SpeechLib_TLB, OleServer, XPMan, ExtCtrls, ComCtrls, ENGINE,

  TFlatCheckBoxUnit;
    SAPI: TSpVoice;
  PropriedsF: TPropriedsF;   
  SOTokenVoice: ISpeechObjectToken;  // See the MS SAPI SDK for info on
  SOTokenVoices:ISpeechObjectTokens; // registry tokens that hold resources
  VPT:= 'Você selecionou X como voz padrão do sistema.';
 
  SAPI.EventInterests := SVEAllEvents;
  SOTokenVoices := SAPI.GetVoices('','');  // Use the registry tokens
 
    //For each voice, store the descriptor in the TStrings list
    SOTokenVoice := SOTokenVoices.Item(i);
    CBFA.Items.AddObject(SOTokenVoice.GetDescription(0), TObject(SOTokenVoice));
    //Increment descriptor reference count to ensure it's not destroyed
    SOTokenVoice._AddRef;
 
    CBFA.ItemIndex := CBFA.Items.IndexOf(SAPI.Voice.GetDescription(0));
 
  TB.Position := SAPI.Rate;
  EDTEST.TEXT:= TROCA('X', SAPI.Voice.GetDescription(0), VPT);
var SOTokenVoice:  ISpeechObjectToken;
begin
  SOTokenVoice   := ISpeechObjectToken(Pointer(CBFA.Items.Objects[CBFA.ItemIndex]));
  SAPI.Voice := SOTokenVoice;
  EDTEST.TEXT:= TROCA('X', SAPI.Voice.GetDescription(0), VPT);
  if SAPI.Voice = nil then
  SAPI.Speak(EDTEST.Text, 1);
  SpFileStream1: TSpFileStream;
  FileName : TFileName;
  T: TSTRINGLIST;
  I: INTEGER;
  T:= TSTRINGLIST.CREATE;
  T.LoadFromFile(EXTRACTFILEPATH(PARAMSTR(0))+'LISTA.TXT');
  FileName := extractfilepath(paramstr(0))+'RAQUEL\'+T[I];
  SpFileStream1 := TSpFilestream.Create(nil);
  SpFileStream1.Open(FileName, SSFMCreateForWrite, False);
  SAPI.AudioOutputStream := SPFileStream1.DefaultInterface;
  SAPI.Speak(COPY(T[I], 1, POS('.', T[I])-1), SVSFDefault);
  SPFileStream1.Close;
  SpFileStream1.Free;


procedure TForm1.FormClick(Sender: TObject);
var
  SpVoice: Variant;
begin
  SpVoice := CreateOleObject('SAPI.SpVoice');
  SpVoice.Speak('this is a test', 0);
end;


SpVoice GetVoices method (SAPI 5.3) | Microsoft Docs
 
Windows, Messages, SysUtils, Variants, Classes, Graphics, Controls, Forms,
Dialogs, StdCtrls, OleServer, SpeechLib_TLB,ActiveX, ComCtrls,ControlCenter;
 
SpVoice1: TSpVoice;
SpSharedRecoContext1: TSpSharedRecoContext;
ProgressBar1: TProgressBar;
SRGrammar: ISpeechRecoGrammar;
CmdLine:string;
ControlActive:Boolean;
adeSoundControl: TadeSoundControl;
 
SOToken: ISpeechObjectToken;
SOTokens: ISpeechObjectTokens;
SOTokens := SpVoice1.GetVoices('', '');
SOToken := SOTokens.Item(id);
SpVoice1.Voice := SOToken;
SpVoice1.Rate:=speech;
SpVoice1.Volume:=100;
spvoice1.Speak(str,SVSFDefault);
Speecker(StrToInt(Edit2.Text),StrToInt(Edit3.Text),Edit1.Text);
 
 
https://github.com/halilhanbadem/delphi-google-speech-to-text
https://www.youant.net/

https://github.com/SirAlex/delphi-jose-jwt
Delphi JOSE and JWT Library
https://github.com/grijjy/DelphiGoogleAPI
Using the Google Cloud Platform APIs

 
voice: OLEVariant;
  voice := CreateOLEObject ('SAPI.SpVoice');
  voice.Voice := voice.GetVoices.Item(combobox1.ItemIndex); // current voice selected
  voice.volume := tbVolume.position;
  voice.rate := tbRate.position;
  voice.Speak (s, SVSFDefault);
  voice.Speak (s, SVSFlagsAsync);
  repeat  Sleep(100); until  voice.WaitUntilDone(10);
 

TMainForm = class(TForm)
    Button1: TButton;
    procedure Button1Click(Sender: TObject);
    procedure FormCreate(Sender: TObject);
    procedure FormDestroy(Sender: TObject);
  private
    { Private declarations }
  public
    { Public declarations }
   voice: OLEVariant;
   procedure SayIt(const s:string);
  end;

var
  MainForm: TMainForm;

implementation

uses
   ComObj;

{$R *.dfm}
procedure TMainForm.SayIt(const s:string); // s is the string to be spoken
const
  SVSFDefault = 0;
  SVSFlagsAsync = 1;
  SVSFPurgeBeforeSpeak= 2;
begin
  memo1.setfocus;
  voice.Voice := voice.GetVoices.Item(combobox1.ItemIndex); // current voice selected
  voice.volume := tbVolume.position;
  voice.rate := tbRate.position;
  voice.Speak (s, SVSFlagsAsync {or SVSFPurgeBeforeSpeak});
end;


procedure TMainForm.Button1Click(Sender: TObject);
begin
  SayIt('Hello');
end;

procedure TMainForm.FormCreate(Sender: TObject);
begin
  voice := CreateOLEObject('SAPI.SpVoice');
end;

procedure TMainForm.FormDestroy(Sender: TObject);
begin
  voice  := Unassigned;
end;

end.
https://stackoverflow.com/questions/12117883/sapi-with-dephi-async-speech-doesnt-work


https://github.com/kimond/Jarvis-Project/blob/master/pyttsx/drivers/sapi5.py

MSSAM = 'HKEY_LOCAL_MACHINE\\SOFTWARE\\Microsoft\\Speech\\Voices\\Tokens\\MSSam'
MSMARY = 'HKEY_LOCAL_MACHINE\\SOFTWARE\\Microsoft\\Speech\\Voices\\Tokens\\MSMary'
MSMIKE = 'HKEY_LOCAL_MACHINE\\SOFTWARE\\Microsoft\\Speech\\Voices\\Tokens\\MSMike'

MSSam  MSMary MSMike


http://www.blong.com/conferences/dcon2002/speech/SAPI51/SAPI51.htm

https://edn.embarcadero.com/article/29583

 Speech Synthesis & Speech Recognition: Using SAPI 5.1 

procedure TfrmTextToSpeech.FormCreate(Sender: TObject);
var
  I: Integer;
  SOToken: ISpeechObjectToken;
  SOTokens: ISpeechObjectTokens;
begin
  SendMessage(lstProgress.Handle, LB_SETHORIZONTALEXTENT, Width, 0);
  //Ensure all events fire
  SpVoice.EventInterests := SVEAllEvents;
  Log('About to enumerate voices');
  SOTokens := SpVoice.GetVoices('', '');
  for I := 0 to SOTokens.Count - 1 do
  begin
    //For each voice, store the descriptor in the TStrings list
    SOToken := SOTokens.Item(I);
    cbVoices.Items.AddObject(SOToken.GetDescription(0), TObject(SOToken));
    //Increment descriptor reference count to ensure it's not destroyed
    SOToken._AddRef;
  end;
  if cbVoices.Items.Count > 0 then
  begin
    cbVoices.ItemIndex := 0; //Select 1st voice
    cbVoices.OnChange(cbVoices); //& ensure OnChange triggers
  end;
  Log('Enumerated voices');
  Log('About to check attributes');
  tbRate.Position := SpVoice.Rate;
  lblRate.Caption := IntToStr(tbRate.Position);
  tbVolume.Position := SpVoice.Volume;
  lblVolume.Caption := IntToStr(tbVolume.Position);
  Log('Checked attributes');
end;