net - UK (2020-04)

PROJECTS
Text-to-Speech API

There are many great options available for speech synthesis

APIs. Here are a few of the best ones, well worth trying out.

0LFURVRIW$]XUHŜ7H[W7R6SHHFK

https://azure.microsoft.com/en-us/services/cognitive-services/text-to-

speech

Microsoft’s solution boasts over 45 languages supported with

over 80 voices. Its audio is high quality and customisable. If you are

already working with APIs or tools within the Azure ecosystem,

this might be a great choice for you.

,%0şV:DWVRQ7H[WWR6SHHFK

https://www.ibm.com/cloud/watson-text-to-speech

Watson Text-to-Speech enables you to create human-like speech,

via IBM’s cloud API services, much like Google’s model. It boasts

upwards of 190 voices, with plenty of control and languages to suit

your needs. It’s another great option to explore, especially if you

are using IBM’s other cloud services already.

$PD]RQ3ROO\

https://aws.amazon.com/polly/

Amazon Polly transforms text into human speech using its deep

learning system. Polly works as a web service like the other APIs,

making it flexible for use across a wide range of devices and

platforms. It also features speaking styles, such as newscaster

or conversational speaking. These change speaking cadence and

rhythm for a realistic result.

API ALTERNATIVES

RESOURCES

that submit this data in the step after this one. In the

head of the HTML add this PHP code:

<?php

// get incoming vars, set defaults

if(isset($_POST[‘text’])){$text = $_POST[“text”];} else

{$text=”Hello.”;}

if(isset($_POST[‘gender’])){$gender = $_

POST[“gender”];} else {$gender=”female”;}

if(isset($_POST[‘type’])){$type= $_POST[“type”];} else

{$type=”en-GB-Wavenet-B”;}

?>

STEP 10:$'')250),(/'6

Add a few form fields for some voice type options.

We’ll add male or female options, along with language

type and a place for the text to be synthesised.

<textar ea type=”textarea” cols=60 rows=5

name=”text” id=”text” ><?php echo $text ?></

textarea><br/><br />

<select id=”gender” name=”gender”>

<option value=”female”>FEMALE</option>

<option value=”male”>MALE</option>

</select>

<select id=”type” name=”type”>

<option value=”en-GB-Wavenet-D”>English UK

Wavenet D</option>

<option value=”en-US-Wavenet-C”>English US

Wavenet C</option>

<option value=”en-AU-Wavenet-A”>Australian

Wavenet A</option>

<option value=”nl-NL-Wavenet-A”>Dutch Wavenet

A</option>

</select>

STEP 11:(19,5210(179$5,$%/(

The next code goes between the PHP tags, between

results tags where we indicated the php code

would go. You need to set the GOOGLE_APPLICATION_

CREDENTIALS variable. We did this in command line

previously and now can manage it here. Use the

putenv command to set an environment variable. Add

this code next in your PHP code:

putenv(‘GOOGLE_APPLICATION_CREDENTIALS=/Library/

WebServer/Documents/richardmattka_v1/proto/net-text-

to-speech/Text To Speech-bc489c0a7499.json’);

Remember to replace the path and file name with

your JSON private key file.

STEP 12:,1&/8'(7+(6'./,%5$5<

Next, incorporate the SDK library. Add this code in

your PHP code section: