Text-To-Speech

Template - To - Speech

Template-To-Speech versus Text-To-Speech.

Template-To-Speech creates sound files (.wav) from the text where part of the sentence is used as a parameter, (e.g. "You have XX dollar(s) [and YY cent(s)]"). The Template-To-Speech method guaranties that the resulting files will be linguistically and grammatically correct. Language is used as a parameter for the templates. As a result, creating phrases in different languages becomes a simple procedure; you call a template with several parameters (where language is just one of the parameters.)

In contrast, Text-To-Speech converts a given text into a sound file. It doesn't care about syntax or grammar mistakes. Correct or not, text will be converted into sound.

The Template-To-Speech approach is extremely useful for multilingual applications. But even for one language applications it appears to be a very convenient tool for constructing correct sound files. (Try it online.)

Learn more:

You've probably called your bank or credit card company hundreds of times and always heard the announcement: "As of today ... your account balance is ...". Part of the phrase or the whole phrase is generated by a Text-To-Speech engine. Let's face reality ; 99.9% of the time you hear similar phrases about accounts, payments, flight information, etc. This means that you rarely need a full scale Text-To-Speech engine, and if you use it - that means you're wasting money and resources. What you really need is the ability to pronounce phrases of the similar content in different languages. This will mean for your business - a bigger customer base and less resources to handle customers. Everything that you need can be easily done with the Template-To-Speech approach. (Please note that the Template-To-Speech application can have an open interface that allows you to use Text-To-Speech whenever it's required.)

What is the Template-To-Speech approach? It’s like a function in computer language. You specify a function name, language and parameters applicable for that function, and it will return to you sound (.wav) file. For example, you need account information. Just do something like this: AccountInformation (English, Date, Amount) and receive a sound file that will say “As of today, January 25th, your account balance is 25 dollars and 25 cents." Specify another language (Spanish) and you will receive the sound file pronounced in Spanish with the similar contents: “Desde hoy, el 25 de Enero su balance de cuenta es 25 dolores y 25 centavos.”
A lot of people may ask the following question: “Why do we need anything else other than the common Text-To-Speech application?”
Let’s start with an example and you will understand why.
For example, let’s take some very common announcements, such as: “You have 1 message.” Or “You have 25 messages.” The words “message” or “messages” depend on the number of messages and you need to resolve that by constructing an appropriate sentence. In English this particular problem could be easily resolved. But even in English language you will encounter several problems with dates and time. US English versus British English: 3 pm or 15:00, January 1st or 1st of January. What about other languages? Many European languages have masculine and feminine types of words and that can heavily affect the structure of the sentence. If you are going to deal with Eastern European languages, you will find that numbers and words have many different forms depending on many different conditions. If you think that Text-To-Speech can handle that you are wrong. Text-To-Speech is responsible only for converting a given text into a sound file. Nothing else! Creating a correct text that might be converted into a sound file is your responsibility. Can you or your programmer do that? Can you deal with multiple rules and exceptions of the particular language? Is your time worth it? Believe us, implementing the grammar rules of any foreign language is not a simple task!
By the way, Text-To-Speech is not free software. High quality TTS engine that supports different languages, additional channels, and tech support will cost you $$.
But here is the solution. Download our Template-To-Speech application and you will be able to create multilingual telephone/IVR/auto attendant/etc. applications in no time. You might ask absolutely legitimate question. For example, “Our Company sells furniture and we need set of prompts that is very specific for our business. How can we do that?”
Don’t worry. For a very low price we will create a set of templates specific for your business for languages that you will order. You will be able to use them along with the set of common templates that we distribute with our application for free.
Please note that template is Not a just computer translation of an English text into different languages. Each template is created with the help of native speakers, and it reflects the precise contents of the original phrase(s).