Hi again!
I return with a new book and a little surprise. The book is the solo version of Scarlet Sorcerer. I hope you'll enjoy it.
As for the surprise: I remember somebody asking in another thread about a way of listening the text instead of reading it. I knew that XP and newer windows operation systems had text-to-speech in their control panel, so I decided to have a really good look at it. What I found was really exciting. It seems that the 3.0 framework (that is included both in Vista and Win7) offers both speech synthesis and recognition. After that and finding out how exactly it works, it was quite easy to introduce it into the program.
So as of now, both SS and EE can read the text out loud if you have winXP or higher and a 3.0 framework or higher. But let me warn you: I found the official voice that comes with windows really annoying. To get other voices, you can do one of two things. The first is to go to
eSpeak's web page and download a different voice. It also sounds mechanicaly, but still much better than Sam with whom I was stuck with. The other choice is to get a more natural voice from
AT&T Labs. Unfortunately these are not free and also very large (3-500 MB), but they are much more pleasant to listen to than the other two.
You can turn speech synthesis on and off in the settings menu only BEFORE you open a book. Selecting the voice there doesn't work yet, so you'll have to set the one you want to use in the control panel as the default. It will kick in after you start the single player game.
While it reads the text you can't continue the exploration. At that point you can only close the book, and then the reading will stop.
You can also use speech recognition to navigate in the book. For that, besides the 3.0 framework, you'll also need the speech api. As far as I know, Vista and Win7 contains that. For XP, you'll have to install it manually. Microsoft claims that if you install MS Office, it automatically installs that one too, but I had to get that separately. I ended up installing the whole speech SDK. If you're luckier than me, you might be able to avoid this.
Speech recognition only waits for commands after the reading has stopped. If the reading is turned off, then it'll always wait for you to speak. The recognized words are the following:
Up/North, Down/South, Left/West, Right/East. These are recognized without any problem. If you're really good at it you can also use the commands Northeast, Southeast, Northwest, Southwest. These, however, usually are misinterpreted as simply north or south. I have a theoretical solution for this, but didn't have the time to implement it yet.
You can choose an element from the choice list box by saying the number of the item. One means the first item, two the second, etc. If selected, the available choices are also read by the program, along with the appropriate calling number.
Under winXP, I experienced a problem with the main volume control and the speach recognition engine. Whenever the later was initiated or disposed, it also muted the master volume control. I fixed this problem as much as I could under XP. I don't know yet if it also presents itself under Vista/Win7, but since the solution is completely different there, if it does, the program cannot correct that one yet.
One more thing: since the recognition engine awaits only for the words it is told to expect, it also assumes that it'll receive one of them. So if you say a word similar to the ones that can be accepted, you can expect an unwanted jump somewhere. I'm also working on a visual feedback, to let you know if the program correctly recognized the order or not. Please tell me how and if it worked out for you
