Users can only use the charset associated with the voice.
e.g. If they use spanish they can only use spanish characters. We need to add validation on the client side to prevents api calls with invalid characters

When the "Text" tab contains no text and the SSML" tab is selected, the app incorrectly displays the error "Please enter the text you would like to synthesize in the text window." and does not submit the request.

Specifying an incorrect SSML input value and downloading the audio file seems to break it.
(I get an error message as expected when I click on 'Speak' instead).

Examples:

• make sure it scales, it could be in the front page of reddit
  • Mainly looking at page weight: Shoot for 2-3mb max and rendering in 5 sec or less.
  • Test with dev tools throttling to 3G speeds and make sure things are still reasonable.
• Add google analytics
• blue-green deployment + travis (see this)
• testing + travis (see this)
• security.js (helmet + express-rate-limitation) + ~~~CSRF (see personality-insights and speech demos)~~~
  • Skipping CSRF protection here because it doesn't work on GET, and we have to use GET to stream to <audio> elements. (It honestly doesn't make a lot of sense in general for unauthenticated apps... but we're employing it for the side-effect of making scraping harder. But, again, it doesn't work on GET requests.)
• package.json should not specify node-engine so that Bluemix will always use the latest one.
• Google RE-Captcha support (make sure design take this into account when designing a demo)
  • Talk to design to add it for existing demos.
• Update travis to send emails when there is a tag release.
• Bluemix deployment tracker and privacy notice

1. Issue: First syllable is lost in created audio after it's downloaded, for all en-US English voices, and some other voices.

Example text:
tomato, tomato, tomato.

Allision voice and most languages
After pressing, "Download" and listening ...
Expected: tomato, tomato, tomato
Actual: mato, tomato, tomato (note: first syllable of first word is missing)

Alexandra

I am getting a challenge to enter ID information to pass through a Watson gateway when trying to use the demo application. In Bluemix Public, I am able to provide my user information and use the service; however, when I try to access the service using a Bluemix dedicated account, my credentials do not work. Assistance would be most appreciated.

{"code":500,"error":"Block-scoped declarations (let, const, function, class) not yet supported outside strict mode"}

How can I fixed this? its about NodeJs version?

Hi, German&James, I've updated the navigation link, service icon and favicon. Please help redeploy the demo :-)

http://181.135.63.86:3003/

Is possible have male for spanish es-US?

need to change the order in the drop down box for the voices. Should be Language (Country code): Name of voice? So for example "English (US): Michael". This way it allows for easier sorting by language.

Hello , Is this demo already using the Language Translation V3 service ?
Or is it still on V2 ?
I cannot find an entry for TTS V3 in the current Watson SDK .
Is there a a way to bypass that ?
Thank you

Thanks for providing!

One quick question: Is the output licensed in a certain way or is it open source as well?

I'm not sure, but from the latency it looks like it is not streaming. I thought this was a recent change by Eric, but maybe not in the published version yet?

When click "deploy to bluemix" button, succeed in

• create project
• clone repository
• config pipeline
  but failed during deploy
  

The play bar seems to be slightly misaligned on the left with the dropdown & download buttons.

How can test all in spanish, only spanish?

When using SSML:
<phoneme alphabet="ipa" ph="nˈaɪtɹəglɪsəɹɨn" ></phoneme> raises the error message: "Language not supported. Please use only ISO 8859 characters"

Long running app can crash due to out of memory error. (With 768 MB).

Should investigate whether this is due to heavy simultaneous usage, a very large input, an ongoing memory leak, or something else.

There is a pause in the middle of the first word "Consapevole" when synthesizing the default Italian text.
This problem seems to only happen when using Firefox. Works ok on Chrome.
Downloading the audio works fine. I only see this problem when the audio is streamed.

The service supports streaming of audio, however on both Chrome and Firefox wait until the entire download is finished before they begin playing, creating a lot of lag. Our old demo did not had this bug (it started playing immediately although the slider position was incorrect, since it didn't know the file size).

In Chrome, I see the option to download the audio here:

This always downloads the audio as 60aa3761-59f1-4519-baca-f029e9a0d8bc.mp3 instead of transcript.mp3 like when I download through this:

Should the GUID be removed?

The "Speak" button in https://text-to-speech-demo.ng.bluemix.net/ is disabled in Safari.
It works fine in Firefox and Chrome.

I'm getting an error while trying to start the node server:

"credentials": {
^
SyntaxError: Unexpected token :
at exports.runInThisContext (vm.js:73:16)
at Module._compile (module.js:443:25)
at Object.Module._extensions..js (module.js:478:10)
at Module.load (module.js:355:32)
at Function.Module._load (module.js:310:12)
at Function.Module.runMain (module.js:501:10)
at startup (node.js:129:16)
at node.js:814:3

My app.js looks like the following and the username and password field are filled in properly

{
"credentials": {
"url": "https://stream.watsonplatform.net/text-to-speech/api",
"password": "passwordhere",
"username": "usernamehere"
}
}

Pressing "Speak" results in no user-visible feedback until the entire download is finished and speech begins. This can often take 10-20 seconds, resulting in a bad user experience. As soon as the Speak button is pressed there should be some visual indication that the request was sent and is waiting for the server.

I'm making a voice file that requires emphasis on some words and audio clips to be played, but neither of those features are supported. Do you know either A) when those features will become supported or B) what other programs I can use that do support those features?

var tts_credentials = extend({
  url: 'https://stream.watsonplatform.net/text-to-speech/api',
  version: 'v1',
  username: 'xxxxxxx',
  password: 'xxxxx',
}, bluemix.getServiceCreds('text_to_speech'));

// Create the service wrappers
var textToSpeech = watson.text_to_speech(tts_credentials);

app.get('/synthesize', function(req, res) {
  var transcript = textToSpeech.synthesize(req.query);
  transcript.on('response', function(response) {
	if (req.query.download) {
	  response.headers['content-disposition'] = 'attachment; filename=transcript.ogg';
	}
  });
  transcript.on('error', function(error) {
	console.log('Synthesize error: ', error)
  });
  transcript.pipe(res);
});

I need update this fragment, because i cant hear speak
Tnks

Overview
I checked out demo project from https://github.com/watson-developer-cloud/text-to-speech-nodejs.
I made .env file and input url and apikey that I got from my IBMCloud account.
Then, run locally.
Access from browser and click speak button.
I got 500 error, and there were error message like screenshots.
After few seconds error message disappear.
I retried to click speak button, then run correctlly.
I recognized that 500 error happens only first time when run localserver by "npm start".

Expected behavior
Run locally, then click speak button first time, then run correctlly.

Actual behavior
Run locally, then click speak button first time, there is 500 error and error message.

How to reproduce
Checked out demo project from https://github.com/watson-developer-cloud/text-to-speech-nodejs.
Follow the Readme and run locally, then click speak button.

Screenshots
Before click speak button

After click speak button

Error message

Additional information:

• OS: MacOS Mojave
• Which version of Node are you using?: 12.0.0

Attempting to use input like: "hello" (with the quotation marks) results in the error: "Language not supported. Please use only ISO 8859 characters"

We need to remove css_browser_selector.js since it has a proprietary license

The docs (and my experience a while back) indicate that the service only supports Ogg containers with the Opus codec but the demo says "The audio is returned in the Ogg Vorbis format which..."

(FWIW, I'd love to see both codecs supported - some of the IoT hardware I wanted to do a demo with supports Ogg Vorbis but not Ogg Opus.)

In Firefox (40.0.3 - IBM CCK), if I visit http://text-to-speech-demo.mybluemix.net/ and click the red Speak button, it logs the following:

InvalidStateError: An attempt was made to use an object that is not, or is no longer, usable

The error appears to be coming from the audio.currentTime=0 line.

Chrome appears to work fine.

Just curious whether this could be deployed as a local instance without bluemix account?

Appreciate if somebody point it to documentation of the kind. Thanks!

TTS DEMO -- For all in US English Voices only, after pressing "Speak" button the program begins to synthesize text to speech, but stops before the end when text contains invalid character

Expected: For all voices, program synthesizes all text, with any character, to speech
Actual: For US English voices, program does not synthesize text, with invalid character, to speech

Note:
As per Radek, "This is a bug in the demopage; it works fine when communicating directly with https://stream-d.watsonplatform.net/text-to-speech/api/v1/synthesize?voice=en-US_AllisonVoice"

See details below:
When en-US_AllisonVoice, en-US_LisaVoice, or en-US_MichaelVoice is selected, and the text below is entered, and you press "Speak" the program stops speaking at the point "Stops speaking here" is stated in the text box.
And an audio file is successfully created containing only the part spoken out loud and not the entire text.
When any other voice is selected, the program works correctly and completes successfully by not stopping in the middle of the text.

Example text in text box:
Italian:
<phoneme alphabet="ipa" ph=".ʤe.ˈla.to"></phoneme> <phoneme alphabet="ibm" ph="1byan.0ko"></phoneme>

English:
<phoneme alphabet="ipa" ph="təˈmeɪ.ɾoʊ"></phoneme> <phoneme alphabet="ipa" ph="təˈmeɪ.ɾoʊ"></phoneme> <phoneme alphabet="ibm" ph=".0tx.1me.0Fo"></phoneme>

Castillian Spanish:
<phoneme alphabet="ipa" ph=".ˈθiŋ.ko"></phoneme> <phoneme alphabet="ibm" ph="1Tin.0ko"></phoneme>

Italian:
<phoneme alphabet="ipa" ph=".ʤe.ˈla.to"></phoneme> <phoneme alphabet="ibm" ph="1byan.0ko"></phoneme>

English:
<phoneme alphabet="ipa" ph="təˈmeɪ.ɾoʊ"></phoneme> <phoneme alphabet="ipa" ph="təˈmeɪ.ɾoʊ"></phoneme> <phoneme alphabet="ibm" ph=".0tx.1me.0Fo"></phoneme>

Castillian Spanish:
<phoneme alphabet="ipa" ph=".ˈθiŋ.ko"></phoneme> <phoneme alphabet="ibm" ph="1Tin.0ko"></phoneme> -- Stops speaking here

North American Spanish:
<phoneme alphabet="ipa" ph=".in.di.ˈβi.ðwo"></phoneme> <phoneme alphabet="ibm" ph=".0e.1DaD"></phoneme>

Brazilian Portuguese

French:

US English

UK English

German:

Alexandra

Please refer to the following document for details and migration instructions https://github.com/IBM-Bluemix/cf-deployment-tracker-client-node/wiki/migrating-to-the-new-metrics-tracker-client.

Add support for MP3 output (doesn't have to be streamed). Currently greatest support for audio support in HTML is through MP3. Ogg is only supported by a few browsers. If we had MP3 support it would be easier to use this service in our product.

We currently use TTS for our product, but the quality of this is much better than anything else out there. We'd like to use this, but our product requires that we know the timings of when words (or phonemes) are said. One of the TTS products we currently use gives us this data and we use it. It'd be nice if this would export that metadata along with the audio file so we could use this service.

We currently parse this from the metadata in an id3v2 tags that are stuck in Comment sections. We currently get something like this:

timed_phonemes
word start_in_ms end_in_ms amplitude
word ...
...

Not married to the format, but just to illustrate what type of information we'd need.

Issue: TTS Demo -- Too many requests, please try again later.
When I make about 5 updates to the text box in a minute, I get an error message: "Too many requests, please try again later."

or when I press the "Speak" button, too often (23 times in a row), I get an error message: "Too many requests, please try again later." (See attachment for a screen shot after this test).

Expected: No error message
Actual: I get an error message: "Too many requests, please try again later."

Alexandra
TTS Demo -- Too many requests, please try again later on June 28 2016.zip

Based on a question in dAnswers: https://developer.ibm.com/answers/questions/186320/text-to-speech-doesnt-like-semi-colons.html

Having just tried out the text to speech demo (English) all was going well until it reached a semi colon (;) in the text.
It was a fair amount of text (several paragraphs) and I though it might might only read the first paragraph so I deleted this character and tried again and it did indeed read past that point and past all other candidate ending points all the way the end end successfully.
I can only therefore deduce that a semi-colon is not supported? Is this intentional or a bug? Are there any other characters that should be stripped out of the text?

This issue was marked as fixed & closed earlier, but I still see it - #57

Example:

1. Select "American English (en-US): Allison (female, expressive, transformable)"
2. Go to the "Expressive SSML"
3. Enter this text: <speak> Hello World
4. Click on Download

I see this (Firefox):

Let me know if you have any questions.
Thanks!

The demo page at reports error

`

Input contains unsupported IPA character 0X00000045

`

for SSML input

<speak version="1.0">
  <phoneme alphabet="ipa" ph="&#x02C8;&#x0045;s">s</phoneme>
</speak>

where ˈEs should correpond to 'Es

Context:
GDPR will go into effect on May 25th; we need the demos to display the following language in order to make them compliant with the law:

This system is for demonstration purposes only and is not intended to process Personal Data. No Personal Data is to be entered into this system as it may not have the necessary controls in place to meet the requirements of the General Data Protection Regulation (EU) 2016/679.

For more information see: https://github.ibm.com/Watson/developer-experience/issues/4342