"The computer literally speaks to you in the native tongue"
This product is SHAREWARE.
SHAREWARE means ONLY if you like it, do you pay for it. It is a way for you to examine the program before purchase. Shareware is a means for software firms to distribute and advertise their software at no cost to all corners of the globe. The result is that there are no advertising costs to pass on to you the consumer, hence the stuff is much less expensive. Sadly this great system is abused and some people don't pay for the goods, which forces the prices of software up.
You can receive the latest copies of this program in your Internet
mailbox at any point registered or not by sending a short note
to mike@msmith.zynet.co.uk
There are several programs on offer !Dis-le 1,2,&3 (French),
The Interface module is copyright of Simon Huntingdon, but is in the 'public domain'. There is normally a document file which MUST accompany the module for programmers. I have permission from him to exclude it from this program. If you would like this mail me with a disc and a stamp and I will send it to you.
FIDO: 2:254/27.0 'Jick'
The SampPlayer module is copyright of David Radford. Special thanks to him for his permission to let the module 'stand alone'. SampPlayer Module is a part of !Player which plays compressed Amadeus sound samples. !Compress compresses them to 50% of their original size whereas, other compressors would only give you about a 10% savings. !Compress also uncompresses samples. !Player and !Compress are in the 'public domain', but are copyright of David Radford. Send me a disc and a stamp and I will send you these programs free of charge.
FIDO: 2:250/219.0 'David Radford'
The creation of the program:
While I was working in a school one day I saw a foreign language teacher's aide quizzing a student on the meaning of spanish words in a corner of the classroom and the idea for the SayIt series. They were performing a task very similar to what this program does and I thought, "Gee, I can get a robot to do that!"
Getting Help:
The SayIt series has been programmed to respond to Acorn's !Help program found on the App disc in Risc-OS 2 and on the icon bar in Risc-OS 3.
If you need help with SayIt, phone or write, even unregistered users will receive a warm welcome.
Alternatively e-mail mike@msmith.zynet.co.uk
How to work the program:
When the program loads a flag of the respective country will load on the icon bar. Clicking the mouse on that icon will pop up a window.
The computer will say a word. You will have an indefinite period to choose one of the four choices. If you do not hear the word clearly there is a repeat button. You will not be penalised for asking to hear the word repeated.
Tick the answer you think is correct.
If you are right the computer will move on to another word and praise you with a random response.
If you are wrong it will tell you that you were wrong with a random phrase, and NOT move on to the next word.
There are two extra buttons on the screen. One will start a new session. The other will repeat the word so you can either get a better feel for the word or remind you if you have forgotten it, or if you didn't hear it in the first place.
Above these two buttons is another smaller tick-box which will make the foreign word disappear so that you do not have any clues to the meaning of the word other than the sound. It is called the 'difficult' box.
The scoreboard reads correct, wrong, attempts, and percentage in that order from top to bottom. You should try to achieve above 70% at all times. Where I come from (the USA) the grading is as follows:
A.........90% - 100% (excellent)
B.........80% - 89% (good)
C.........70% - 79% (satisfactory)
D.........60% - 69% (poor)
F..........59% or below (failing)
The entire screen is in THE language to give you a feel for working in that language.
The iconbar menu says Info, Earphones, and Quit. The Earphone choice, when ticked, will direct all of SayIt's sound to the headphone socket for classroom or late night study, but will not interrupt the other parts of the Archimedes sound system, like the 'error beep'. The earphone option doesn't work well with a RiscPC series machine.
There used to be a problem where you couldn't run two copies (ie french and german) at the same time. This has now been fixed on all copies of 1.20 +
If you don't have a hard disc or you intend to use this across econet you will find this program a bit slow. A little technique to get perfect performance from this program is to create a RAM disc of 800K and copy the program into it, and then run the RAM disc copy. You will find that you have as close to true human speech as a computer can deliver.
This program pushes the now possible limits of the machine both with memory, disc space, and disc access.
Comments about the people behind the voices:
I have purchased talk tapes in the past where you feel the person speaking definitely has an english accent. I assure you that all speakers in the SayIt series are native to that language, and that the language they speak is their mother tongue. I have been assured that they have no speech impediments and speak a 'main stream' dialect of that language (ie for english you will not have someone with a southern USA drawl).
Catherine (!Dis-le, french)
Born in Belgium. Came to this country (England) with her husband who is going to medical school.
Karin (!Sprich, german)
Born in Germany. Came to this country (England) with her english husband, who after six months now works at the school where SayIt was conceived, completely by accident.
Santiago (!Dilo, spanish)
Born in Spain. Met an english girl and the rest is history.
Mo-chong (!Gong-che Cantonese Chinese)
Hong Kong lass moved with her folks after she married a british man.
Gabby (!Uttala Swedish)
Moved from Sweden to live with her english mum for a while.
You may find it necessary to need to alter the human recordings found in to your own version of the program for certain unspecified needs.
In order to save confusion you must understand you may not sell our program to drive your sound samples, if you intend to sell the samples you produce, you need a licence from us. If you implement your own words they are for your own use. You must understand that even though the product is shareware; we produce a package called SayIt and support all features of it, dictionaries, samples and the program itself. We can not go around fielding technical support and advice for third parties dictionaries and corrupt samples.
Recording the samples
Step 1. Using a suitable microphone record the words you need. We normally record words in groups and then break them apart. When recording the samples remember the higher the quality the faster the input rate. Therefore, sound samples at 47,000 would produce very nice recordings, but you could fit very few words on a disc. We find that machines below a Risc PC record well at 9900K, above a RiscPC MUST be 13000. We record exactly at 13000.
Step 2. Create a dictionary called Temp, save each sound sample in Temp. The name for each sample must use the first 10 letters of each foreign word verbatim.
For example, maison (house in France) would be 'maison'. If the word were Gescheftundgelegenheit, the file name would be 'Gescheftun'.
Compressing the samples
Step 1. In order to work with SayIt all samples must use a special form of compression in order to play. A normal
