Jacken’s Blog

Setting up a professional sounding Digital Receptionist in Asterisk@Home

October 14th, 2005 · 7 Comments

Asterisk After setting up my Asterisk@Home server and done all the configuring to get all lines and extensions up and running, it was time for recording the Digital Receptionist messages. My first stab at it wasn’t that great, let’s face it, I’m no voice-over announcer. I wanted a little more professionally sounding recording. So I started to search the net for a way to do it, and I think I found a pretty good solution. There’s a text-to-speech software from a company called Scansoft that is amazing. They also has a demo so you can try it out. If you choose RealSpeak Telecom you get examples that’s 8 khz wav files, just what Asterisk wants. You can only write up to 100 characters in the demo, but I’ll show you how to download and stitch together the demo files to get a longer message that’s usable. There’s a lot of different languages to choose from, but these two are the best sounding English ones.

  • English Australian - Lee
  • English Brittish - Daniel
I live in Sweden, so I have used the Swedish voice. It’s not as good as the english voices (especially the Brittish), but it’s good enough.

So when you’ve chosen the language and the voice you like, enter the text you want to use for your Digital Reception message. You can only enter 100 characters at a time. When the demo voice loads in your browser, listen to see if you like the result. If it sounds bad, try to rephrase using other words instead, and try again. When you get the result you like, copy the URL from the resulting page and download the file to your computer. On a linux or Macintosh computer, the easiest way to do it is to open a terminal window and write

curl -O http://urlofdemo_file

You should now have a .wav file in your home directory.

There probably is an easy way to do this in Windows, but I don’t do Windows. You could try this. But I haven’t tried it.

Repeat the above steps until you have all the speech you need for your voice menu.

Now open the first file you have downloaded with a sound editing program. I would recommend Audacity, a Open Source sound editor, downloadable free for Mac OS X, Linux and Windows.

When you have installed your audio editor, open the first sound file that you saved earlier.

Now it’s time to combine all the sound files you have saved, so open the next sound file. Then choose “Select All” under the Edit-menu.

Selectall

This selects all the audio and it’s time to copy it. Select “Copy” from the Edit-menu.

Copysound

Close the window. You should now see the first sound file you opened. So now let’s move the cursor to the end so we can paste in the sound we copied earlier to the end of this one. Choose “Move Cursor… To track end” under the Edit-menu. Now choose “Paste” from the Edit menu. The copied sound should now be in the first document. Continue to open the next document, select all, close, move the cursor to the end and paste in with all the other documents you saved. Now you have a complete audio file to use for your voice menu.

Check to see if you need to change sampling rate of the document. Down in the left bottom corner of the document you should see the project rate. Make sure it’s 8000, otherwise change it.

Audacity Khz

Now where ready to save the completed audio file. Select “Export As WAV…” from the File-menu.

Audacity Export

Select an appropriate name for the file and save it.

So let’s create an Digital Receptionist in Asterisk@Home.

Open your web browser and navigate to your Asterisk server. Go into the Asterisk Management Portal and select Setup.

Select Digital Receptionist. Click on “Create a new Voice Menu”. Your now going to get a question on which extension you have. Ignore this, because we don’t need to record a message, we have our new one we just created. Just click on Continue.

You should now have the Record Menu on your screen. Here you can dial *77 to record a message, but you also have the choice to upload your own file. And that’s what we’ll do. Press the “Choose file” button. Select the combined sound file you created earlier. Now press the Upload button. Note to Safari users. Sometimes there’s a problem uploading files. Download Firefox and use that instead.

You should now be able to dial *99 to listen to your uploaded voice menu.If everything sounds good, give the menu a name and enter a description of the menu. Press the Continue when done.

Now you need to enter the number of options in your voice menu. I’ve set up my menu so people can leave voice mail to all our four family members, so in my case it’s four. Press continue.

Menuoptions Now you have to describe which phone keys does what. Select your choices and press Continue.

You now have a new Digital Receptionist to use.

This isn’t brain surgery but not everyone have used a sound editor so maybe this helps someone.

Next step, check if there is a possibility to use Scansofts RealSpeak together with Asterisk. Think about the possibilities…

More blog entries about Asterisk

Tags: Asterisk PBX · Technical stuff

7 responses so far ↓

  • 1 James // Nov 15, 2005 at 17:13

    Do you know how to use WAV file recordings as individual voicemail box greetings and annoucements? A@H AMP allows for uploading files but then doesn’t indicate where they go or how to assign to specific extensions.

    Great site!

    Best regards, James

  • 2 Jacken // Nov 15, 2005 at 19:10

    Hi James.

    You can put them there manually. The greeting and busy sounds are stored in /var/lib/asterisk/sounds/vm. In that directory there’s one directory for each extension. So for example, if I want to change the greeting message for extension 200, I would upload a wav file called greet.wav in the directory /var/lib/asterisk/sounds/vm/200 on the asterisk server. If you want to change the busy message, upload a wav file called busy.wav. Make sure the files are in 8khz 16-bit wav format. I would recommend the free Open Source program Audacity.

    Good luck!

    Jacken

  • 3 Han // Nov 25, 2005 at 6:23

    Hi Jack. Im new to this Asterisk@Home server. The problem now is, i cant access to the Asterisk server on Linux platform. It give me an error message saying “Server Error. Could not connect because of networking problems. Contact your system administrator.(CANT_CONNECT)”. Asterisk server is being install on a different pc.

  • 4 Joe Bennett // Jan 30, 2006 at 20:57

    Hi Jacken,

    Thanks for the helpful info. I’m having a problem though. I have the sample rate set correctly, but I still can’t hear the .wav files I upload. Any Ideas?

  • 5 Phil Donaldson // Mar 12, 2006 at 10:11

    Hi, I had a similar problem. For me it was because I was saving the .wav as 8Khz but in 8 bit. Make sure you save it as 8khz and 16 bit. Phil

  • 6 Sridhar // May 21, 2006 at 2:55

    Hello,

    I was trying out the demo but it was nto working. I have found out another place where we can get wav files

    http://naturalvoices.att.com/demos/

    Please include it in your blog if this is helpful.

  • 7 Stephen // Sep 4, 2006 at 0:54

    With the Real Speak demo unavailable for over a month now while they “upgrade”, if you want an alternative to the AT&T one that doesn’t have a character limit you simply need to create a free mail.com email account and email the text you want to your own mail.com email account. Then using MAIL.COM beta mail reader (web based) to read your email it will use ReadSpeaker and you have a choice of different voices. The UK english sounds very similar to the Nuance realspeak voice, however they seem to be unrelated companies.

You must log in to post a comment.