Voice banking and Model Talker questions

[Kiwikid]

Hi,

I’m new to this forum although have been reading some of the threads on here recently.

My Father (62) was diagnosed in early November with bulbar-onset ALS. His only obvious symptom at this stage is slurred speech, although it has deteriorated really quite quickly over a couple of months.

We are looking at voice banking and speech devices generally. It seems that Model Talker is a very popular choice but their website is not particularly user-friendly.

We’re not very technically-minded so could someone please explain how it works in practice? I understand the voice downloading process onto the PC, but does that mean that my Father needs to be near a computer to talk or are there other devices which we can transfer the files onto which he can carry around? If so, which are the best? Does he then need to type in phrases every time he wishes to speak or is there a different way of doing it?

Thanks in advance – this forum is a great source of information.

Brendan

 

[BrianD]

Using modeltalker, you can generate a computer voice that attempts to approximate the actual sound of your father's voice. After you download their software, you would then record your father's voice saying a bunch (over 1000 if you can) of different phrases they have identified to try and get all of the voice sounds they need. This is done using a computer and typically a USB head mounted microphone. Once you're done recording of phrases, you would then upload the recordings back to the modeltalker people. They will then generate a synthetic voice for your father which you can download after a few days.

The voice that you download can be used just like any other computer voice and can be used inside a Windows environment directly. More commonly, people will use a specialized communication device such as the Dynavox or Tobii. These specialized devices have software that make it a little easier to type in what you want the device to say, and are expandable in the future to using eye gaze control when someone can no longer use their hands to type.

If you do a search for modeltalker, Dynavox, or Tobii, you should be able to find more information on them.

There are other options available as well, such as an iPad or iPod touch, if you're just looking for speech generation devices, but not necessarily with the modeltalker voice.

I hope this helps answer some of your questions...
Brian

 

[Kiwikid]

Thank-you Brian - this is very helpful.

Brendan

 

[rcharlton]

If your father is not that familiar with computers I would probably not recommend modeltalker. Also if his speech has begun to deteriorate it may be too late.

It is not complicated but you do need to sit in front of the computer the whole time. You don't need to type anything but you do need to navigate around with a mouse.

Modeltalker has an inventory of about 1700 words and phrases. With each phrase you will be given a voice sample then prompted to record using your own voice.

I would think you would need to dedicate a good 2 weeks of 1 or 2 hours a day to complete the inventory.

I finished recording the Model Talker phrase inventory about 4 years ago. I spent about a year off and on doing the recordings. Some phrases I would have to repeat over and over to get right. I became so frustrated that I would put it aside for a few weeks. The problem may have been caused by my noise reducing microphone.

Anyway, once I finished my recordings I received my synthetic voice in less than a week.

The results were variable. Short phrases and those using words directly from my inventory sounded amazing and exactly like me. Longer passages involving a lot sound synthesis were less clear - but still ok.

I don't think model talker is a complete solution as I still need a second program with prepared phrases like etriloquist; however, the process was worth it just to have the ability to say I love you to my wife and daughter using my own voice.

 

[Danijela]

We have downloaded Laurence's voice. Variable results, as pointed out above (OK for shorter phrases). Does anybody know how to slow the voice down? It really seems way to fast to me - compared to Laurence's natural speed of speech. Dani

 

[tammyg]

Sorry about dad, i know what you are going through. We got a machine fthrough the ALS clinic, we never got to do this for my dad, he was mis diagnosed for over 2 years, so we did not know. This might be a little cheaper too. I think they can put in his voice to. My dad used a voice called Paul, ( i wanted him to use the Mexican guy voice but he would not go for it)it broke my heart the first time he called to tell me something on the phone. I cried all day, but I was glad to hear he could still communicate, It takes some time to punch it all in. My son's cell phone(verizon) has a message / notepad read out, that says what the words say. Hang in there

 

[sadiemae]

This is an excellent thread. Rose did it very well.

https://www.alsforums.com/forum/gen...ssistive-technology-clinic-voice-banking.html


This is a great thread on the subject.

 

[BrianD]

Does anybody know how to slow the voice down? It really seems way to fast to me - compared to Laurence's natural speed of speech. Dani

Assuming it is the same as the versions I have...

For "Modeltalker" you change the Options-General Parameters-Speech Rate to a lower number. There is also a Speed variable on that same parameters page that you can set to 0 if it has been changed.

For "Modeltalker 2" you change the Settings-Synthesis-General- Playback Rate and/or Speaking Rate slider bars. Note that I don't know if modeltalker 2 is generally available... I received it recently from them when I asked to get an updated Windows 7 compatible voice, because I recorded my voice so long ago the program I had was only for XP.

 

[Kiwikid]

Thanks all for your input and support - it's much clearer now!

Brendan