nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

The same message is not heard in the same way (API != Portal)

dqsi

Hi everyone,

I am using api to create text to speach message. I create it, store it, and use it for send message.

Through the API i generate a message with a "male" voice, but when i receive the call there is a female voice.

This message is stored in the server, so when i retrieve and listen this message i heard a "male" voice !

Can you explain me why the message are not hearded in the same way ?

Regards

Find more posts tagged with

Comments

skamens

Please provide the JobId's for the examples you tried, so we can investigate further.

dqsi

Ok, so the job id is:

67244827

It was sent à 11h56 today with this TTS file:

!messageid_msg_29-56-12_10-56

The content of this file:

<voice gender='male' category='adult'>bonjour un test</voice>

</speak>

Regards

skamens

I'm not sure what you mean by "This message is stored in the server". Do you mean that you use the preview function on the portal to listen to the template before you submit it as a job?

If so, can you provide me the URL you use to access the portal? If not, can you explain further?

dqsi

I'm mean: i use the easylink API for create TTS content and create job. The content is sent to server, i have a response.

At this point the content is stored in your server. Then i create a "job" that use this "content" already stored in the server.

So when i use the API:

1) Create TTS content and upload to the server (in my case !messageid_msg_29-56-12_10-56)

2) Create job that use TTS content (in my case a new job N°67244827 with !messageid_msg_29-56-12_10-56 content)

3) Send Message

A this point when i receive the call, i heard a female voice (this is not normal, cause i put male voice parameter in the content).

For checking if the content is correct, i go to the portal https://myportal.easylink.com/welcome.do and search for the content "!messageid_msg_29-56-12_10-56". I found it, and when i do a preview of this content i heard a male voice (correct i this case).

So my question is: why using API i heard female and through Portal i heard male ?

the problem is more global, because none of parmeters content is interpreted by the API while the portal preview is good (for exemple insert pause in content)

skamens

I'm not 100% sure what's happening, but I think I have an idea.

The particular system where the actual text-to-speech conversion is done is different when you use preview as opposed to when you actually submit your job for delivery. My theory is that the one used for delivery does not have the proper voice definitions loaded, and therefore it is giving you the wrong voice (it would rather give you a female French voice than a male English voice, I believe).

Unfortunately, this is a problem that we can't address via the API forum, since your API functionality is working fine.

Please contact customer service and ask to have a support ticket opened on this issue so they can have the right people investigate the systems involved.

If you provide them the job numbers of your tests they will be able to figure out where the text-to-speech was performed.