Subscribe to GameFromScratch on YouTube Support GameFromScratch on Patreon Join the GFS Discord Server!
30. May 2012


There was an interesting ( to me anyways… ) topic on Reddit today about making games that are accessible to the blind or visual impaired.  After thinking about this for a little bit, I though that there might be a remarkably easy way to add text to speech to a console based game.  Turns out I was correct.



The following is a C# app that makes heavy use of the .NET System.Speech libraries, however it does appear to be available in Mono, so it should work across platforms.


In the end the process is remarkably simple, all I am doing is redirecting Console.Out ( StdOut ) to my custom class, which converts the text to speech.


Let’s take a look, this code is across two classes.  One is our exceedingly simple game:


using System; using System.Collections.Generic; using System.Linq; using System.Text; namespace AudibleTextAdventure { class Program { private static StdoutToSpeech speechOut; static void Main(string[] args) { speechOut = new StdoutToSpeech(); Console.SetOut(speechOut); Console.WriteLine("You are standing in an open field west of a white house, with a boarded front door."); Console.WriteLine("There is a small mailbox here."); bool done = false; while (!done) { Console.Clear(); Console.WriteLine("What would you like to do?"); string currentCommandString = Console.ReadLine().ToLower(); string [] tokens = currentCommandString.Split(new char[] { ' ' }); string currentCommand = tokens.First(); switch (currentCommand) { case "volume": { if (tokens.Length > 1) { if(tokens[1].ToLower() == "up") speechOut.VolumeUp(); if(tokens[1].ToLower() == "down") speechOut.VolumeDown(); } break; } case "quit": done = true; Console.WriteLine("Thank you for playing, Goodbye"); break; case "help": Console.WriteLine("Sorry, you are beyond help"); break; case "changevoice": speechOut.NextVoice(); break; default: Console.WriteLine("I don't know the work \"" + currentCommand + "\""); break; } } } } }



Most of that code is the skeleton of our “game”.  The majority is just getting and displaying strings as well as parsing and handling the commands our game accepts. The only lines of any real consequence here are:

speechOut = new StdoutToSpeech(); Console.SetOut(speechOut);


Here we declare our StdoutToSpeech object we are about to define in a second, and replace the standard output stream with it.  Now lets look at StdoutToSpeech:

using System; using System.Collections.Generic; using System.Linq; using System.Text; using System.Speech; using System.Media; namespace AudibleTextAdventure { class StdoutToSpeech : System.IO.TextWriter { static System.Speech.Synthesis.SpeechSynthesizer synthesizer; static SoundPlayer soundPlayer; static System.IO.TextWriter origOut; static int currentVoice = 0; static List<System.Speech.Synthesis.InstalledVoice> voices; public StdoutToSpeech() { // Grab a reference to Stdout before it's overridden origOut = Console.Out; synthesizer = new System.Speech.Synthesis.SpeechSynthesizer(); // Will bork if no voices found voices = synthesizer.GetInstalledVoices().ToList(); synthesizer.SelectVoice(voices.First().VoiceInfo.Name); // Slow it down a bit synthesizer.Rate = -1; synthesizer.Volume = 5; soundPlayer = new SoundPlayer(); } public override void WriteLine(string value) { // We still want text to show... origOut.WriteLine(value); using (System.IO.MemoryStream mem = new System.IO.MemoryStream()) { synthesizer.SetOutputToWaveStream(mem); synthesizer.Speak(value); soundPlayer.Stream = mem; soundPlayer.Stream.Position = 0; soundPlayer.PlaySync(); } } public void VolumeUp() { if (synthesizer.Volume < 10) { synthesizer.Volume++; this.WriteLine("Volume increased"); } } public void VolumeDown() { if (synthesizer.Volume > 0) { synthesizer.Volume--; this.WriteLine("Volume reduced"); } } public void NextVoice() { currentVoice++; if (currentVoice >= voices.Count) currentVoice = 0; else { synthesizer.SelectVoice(voices[currentVoice].VoiceInfo.Name); this.WriteLine("Voice changed"); } } public override Encoding Encoding { get { throw new Exception("If you are trying to use this as text, you are in for a world of hurt!"); } } } }


In our constructor we create our System.Speech.Sythensis.SpeechSynthesizer, set a voice, default speed and volume for it.  We also instantiate our SoundPlayer which is going to actually play the sound file our synthesizer um… synthesizes.


The key to this class is that it inherits from System.IO.TextWriter, this is the only type that we can set standard out to.  Our class must implement the Encoding:Get method, which if anyone actually called would cause all kinds of problems, so we throw an Exception if it is actually called. In the constructor we take a snapshot of the original standard out, so we will still be able to print to text.


Otherwise most of the work happens in our WriteLine overload.  First we print out to the original standard out, otherwise our command prompt would become audio only!  Next we create our memory stream that our synthesizer is going to record to.  We then render whatever speech was printed via a Console.WriteLine() call, set it as the active stream and play it using our soundPlayer.  We using PlaySync() to make sure it finished playing before continuing execution.  You could use Play(), but it would only render the most recent sentence if a couple lines where written at once.   Also note, we only overrode WriteLine(string), so if you do Console.Write, Console.WriteLine(char[]) or any other version, nothing will happen.  If you want to support more versions, you need to implement.


Otherwise the rest of the code provides a simple interface for increasing and decreasing volume, as well as changing between voices.  Keep in mind though, if you are using Vista or Windows 7, you probably only have on voice installed.  You can however add additional voices ( including different languages ), I download one from here for example ( copy the dlls to your debug folder if you get an error ).



Now when you run the application, you can use the following commands:

quit – exits the “game”

volume up – increases the voices volume

volume down – decreases the voices volume

help – displays a help string

changevoice – changes to the next installed voice, if you have any.



Obviously in a real game you would do substantially more.  In a real game you would also add a great deal more error checking too, as this is just a proof of concept I slapped together!  Obviously use at your own risk, as I offer no warranty.  Then again, I never offer a warranty…



And here it is in action:

Text to speech adventure in action


blog comments powered by Disqus

See More Tutorials on!

Month List