Text-to-Speech Yes or No?

[Noel_Griffiths]

Hello readers,

I dont use TTS but have been asked to explain the concept to one of my colleagues.

I guess these follwing questions are along the lines of asking "how long is a piece of string?", but here goes.

For those of you who do use TSS:

Would you recommend it?

Does it sound like a human voice or is it "robotic?"

Can it handle technical words and words not normally used in day-day conversation?

What are the pros and cons of TTS vs adding your own voice audio?

FYI..the person considering TTS is using Captivate v7.

Noel

[Sreekanth_S]

We use TTS extensively in our projects, which include software demonstrations of all sizes and explanations of the related concepts, and we are quite satisfied with the output.

1.  Yes. I definitely recommend it.

2.  No, it doesn't sound like perfect human voice. But it is not completely robotic also. The best part is you can tweak the tone, speed, etc., using VTML tags.

https://ondemand.neospeech.com/vt_eng-Engine-VTML-v3.9.0-3.pdf‎

3.  Yes, it can handle. You can train the voice to get the pronunciation right using the custom dictionary and VTML tags.

4.  Pros: Easy to maintain and hence cost-effective. If you record using voiceover artist (your colleague or professional) and if the same artist  is not available when the course is due for an minor upgrade, the maintenance cost will be huge.

     Cons: Initial recording time, and hence the cost will be more due to the tweaks and testing time involved.

               It may not be suitable for all types of courses. For example, personally I would not prefer it for a soft skills course as I feel human touch is necessary in such courses.

Sreekanth

[Noel_Griffiths]

Hi Sreekanth_S

Many thanks for your reply.

I was unable to get the link to the PDF to work.

Noel

[Sreekanth_S]

Sorry, some junk characters got copied. Here is the right one. If this doesn't work, just google for "vtml tags in Captivate". You will find it.

https://ondemand.neospeech.com/vt_eng-Engine-VTML-v3.9.0-3.pdf

Sreekanth

[Noel_Griffiths]

That worked....thank you.

Noel

[bobelmore]

Yes, it works.  I have found that only the Bridget voice is good enough to use.  Some of my colleagues use it, but I don't.  While the Bridget voice is quite good, and probably will meet many organizations' requirements, it still sounds a bit robotic to me.

I have my own recording equipment, which is what I use.  I think some people have had good success using headsets too.

This is an area of disagreent I have with some of my colleagues.  I used to use professional voice talent and a recording studio, and that has set my expectations for professional voice-over.

Now I use my own recording equipment, which includes professional quality microphones, a pre-amp and my MacBook pro and Adobe Audition, which gives me mobilty.

Some think this is overkill, but I have been astounded at the poor audio quality in some supposed professional eLearning.

So, TTS gives you a better product than poor audio recording, but to me, it is not yet at a professional level.

[RodWard]

I agree.  I only ever use TTS (usually the Kate voice) as a way to roughly time out how long a module is going to take.  I can roughly synch up on screen objects for animation as a way to show clients approximately how the module will work.

Once voiceover is approved and locked down, I do my own voiceover recording with high quality USB microphones (Blue or Rode).