Vox Daily The Official Voices.com Blog

IVONA : Text-to-Speech Making Headway?

By Stephanie Ciccarelli

April 10, 2008

Comments (12)


IVONA is the leading text-to-speech software program in Europe and its notoriety is spreading quickly.

Watch this video and let me know what you think.


IVONA TTS

According to this press release posted last week, IVO Software – the manufacturer of IVONA TTS, one of the best text-to-speech (TTS) systems in the world, has begun a global viral-video campaign promoting IVONA TTS.

For the purposes of their campaign IVO prepared a short, 42 seconds long video which has already won 11 honors in the YouTube Science & Technology category. In this video IVONA TTS introduces itself to the world at the same time encouraging a viewer to test it himself or herself.

"IVONA's virtual voices may be used by creative filmmakers. Each day a number of videos with IVONA-generated voice overs are placed on the YouTube" - suggests Lukasz Osowski, President of IVO Software.


Does TTS Pose a Threat to Voice Actors?

Now, text-to-speech is a topic we've touched on before, and it can become a rather heated one at that. As we watch technology advance, it appears that text-to-speech is making progress, but just how much and by whose standards?

Does text-to-speech pose a threat to voiceovers recorded for the web? Let's open this can of worms again.

What do you think?

Best wishes,

Stephanie

Related Topics: ivona, ivona tts, text-to-speech, tts, voice overs, voiceovers, YouTube


Comments


    Yeah...

    Not so much.

    We're smack dab in the middle of the uncanny valley. It's the same problem we're facing with robotics and computer animation. We're really close to replicating humanity, but still years (decades?) away from capturing nuance. We can make still images that are indistinguishable from the real thing, but our attempts at GENERATING performance are still only up to the level of "dancing cadavers" (see Polar Express or Beowulf).

    http://en.wikipedia.org/wiki/Uncanny_Valley

    All of the technology advancement to this point has been incredible, but many scientists are still expecting that a revolutionary jump in computing will have to happen before we can't tell the difference.

    I think it's telling that most of the time we hear robots and computers they're still voiced by actors. Take the recent break out hit Portal. The voice of GLaDOS, easily could've been produced by IVONA, but to really capture the wit and sarcasm of the evil AI, Valve decided to go with actress Ellen McLain (who was incredible in that game).

    I wrote about a similar program called Vocaloid. I think it's cool stuff, but it's still just a toy.

    http://someaudioguy.blogspot.com/2007/07/funny-and-scary-at-same-time-vocaloid.html

    Posted by:

      Would I believe it is a machine?

      Absolutely... you've come a long way baby, but still easily distinguishable from human speech.

      BTW: a machine cannot and will not for long time, be able to emulate the human behaviors and patterns which allow us to immediately identify our favorite voice actors/actresses. What sets guys like LaFontaine apart from the rest isn't just his voice, but his very mannerisms and personality that are translated into his interpretation of the copy.

      Until true AI is a reality, you can't copy that.

      Posted by:
      • Greg
      • April 10, 2008 8:34 PM

        When the computer voices can act, I will worry about it. For now, they are lightyears away from even sounding as good as the HAL 9000 (in the movie "2001: A Space Odyssey"); of course, HAL was actually an actor's voice. I ran this sentence in IVONA:

        "Introducing Scott Scooter Fortney, Voice Actor, Producer, Creative Maniac."

        The result was a choppy monotone read. I then ran the same sentence in the AT&T Labs demo (http://www.research.att.com/~ttsweb/tts/demo.php) and got better results, but still rather choppy.

        So... should we worry? Nah. A good voice actor should be able to read the same line ten different ways. Tell that to the computer program.

        Have a great weekend!
        Scooter

        Posted by:

          I am in agreement with Scott and Greg.

          I don't think we have anything to worry about now... though 10-15 years from now.... who knows?

          It is all choppy and there is no "personal" touch to the sounds -- Which is what VO is all about!!

          Posted by:

            Computers will never be able to replicate emotion to the degree that it's believable -- at least not in our lifetime.

            Sure, for instructions, directions and other technical material, text-to-speech may be viable, however those voices do need to be originally recorded by someone.

            In short, this industry remains well positioned for the foreseeable future.

            Posted by:

              I just typed in a few lines from a theatre show I'm currently doing -- just to see if there was any range of emotion.

              None whatsoever.

              I agree with Scooter that when computer voices can act, then we have something serious to worry about.

              However, there may be folks out there who don't know or can't tell the difference. That's where the rub may come. Look at how our language has been dumbed down. Could our voices face that future? I hope not!!!

              Posted by:
              • Jennifer Sloane
              • April 11, 2008 10:47 AM

                Hi Stephanie!

                Extremely interesting post and comments here so far. I was playing around with this Ivona tool and while it is novel and somewhat facinating, I agree with everyone here that it has not come close to replicating and replacing voice talents. Ivona sounds too robotic and monotone, and definitely does not posess the mind and heart of a voice actor. However, I foresee the uses for visually impaired and cell phone text/PDA applications to be beneficial.

                Thanks for listening.
                Bobbin Beam

                Posted by:

                  Hi Stephanie,

                  Sure, TTS will be used...especially on the web, I would think, where it will take some work away from us. But overall, the human voice is just too organic to be effectively replaced by something synthetic. At least in our lifetimes.

                  Best wishes to you.

                  Fred

                  Posted by:

                    I'm just getting caught up with my e-mails and just saw this. Before I comment on whether this technology will be a threat, I listened to the sample provided on the IVONA website which stated:

                    "Hi, I am Jennifer, one of the IVONA Text-To-Speech voices. The IVONA TTS was recognized by the world's famous experts as the best. Did you know that?"

                    The interesting thing was that the phrase "famous experts" came out "famous sexperts." Now, how many producers and directors would allow a live VO get away with that?

                    I also found the voice to lack any character or warmth. I just can't imagine any copy of any length recorded with this technology holding the listener's attention for any length of time. (And if they're writing sentences like this, they'll lose them after the word "technology"! I hope you guys can follow my train of thought.)

                    That being said, I do believe that people who are looking to cut costs and corners will use the technology until they come to realize that they're not getting the return they are looking for. We all know that there are those who would prefer to do things fast and cheap with no concern about quality or affect on the listener. Those are the people who will be using this technology. Than, when their sales numbers aren't improving (or whatever other criteria they're using to measure success), they will hopefully realize that the robotic voice just isn't generating any interest in their product or service and they'll return to using VO talent who can get the results they're looking for. (God, these run-on sentences. I apologize.)

                    Arlene Kahn

                    Posted by:
                    • Arlene Kahn
                    • May 4, 2008 8:59 PM

                      TTS will soon rule for some content types. There are two driving forces:

                      * Accessibility requirements
                      * Localization

                      TTS meets both global, corporate requirements neatly. It makes both possible within budget and on deadline.

                      It is unlikely that it will take all VO work simply because of human discomfort with machine voices, but TTS is coming much sooner than most people expect.

                      Posted by:
                      • steve hammill
                      • September 8, 2011 8:43 PM

                        Some people are so in love with technology that they will use this just so they feel they are cutting edge. It sounds scary to me, big brother-ish. It will make a small dent but it will never be able to truly replicate the beautiful vibrations of a human voice. I think most people will be annoyed and insulted by it.

                        Posted by:
                        • William Babishoff
                        • September 13, 2011 1:44 PM

                          Hi,
                          Before one discusses insults to the ears one should go and compare more than two tts samples from different companies. I feel obligated to note that most companies demoing tts slow the speech way down. I grew up using a screen reader and a speech synth called ETI eloquence. Talk about tone quality issues and robotics that will do it. All of that to say tts isn't a threat to voice actors or to another big group audiobook makers. It might be in a few more years but not at the moment. And oh, btw, I'm using an ivona product on a much higher speed than they use in demo.
                          Thanks,
                          D

                          Posted by:
                          • Daria
                          • April 4, 2014 6:13 AM

Leave a Comment



Recent Articles

YouTube Hits The 60 fps Gas Pedal

Toy Story 4 Set for June 2017 Release

Where Does Pixar Find Its Animators?

Feast On Disney's New Short

DreamWorks, Marvel Studios Play Release-Date Bingo

Audition Improvements Feature Helps Talent, Raises Project Quality

London Poppy Day 2014: Using Voice To Honour

Big Hero 6 = Big News For Animation

SpeechMasterPro Helps Pros Speak Better

When Famous Actors Voice Video Games

   

Animation Solution Center

Radio Advertising Solution Centre

Learn why video animation is more important than ever, how you can use it to gain competitive advantage and what tools are out there to help you make it happen.

Animation Solution Center

Subscribe by Email

About Vox Daily

Vox Daily offers a daily dose of voice acting news, articles, tutorials, interviews, intelligent conversation and business ideas for voice talent and voice actors.


Follow Us

   

Our feed & social options update you with special offers and news as it happens.

New YouTube Video

Watch videos on YouTube