Kendell Clark

This is not entirely accurate, this article. I'm not writing to criticize, but to point out all of your options. I'm a blind person, and as such I don't need or necessarily want high quality, which often translates to enormously sized, voices taking up space on my hard drive. To my ears the naturalness sounds more like a sophisticated but not very correct algorithm for running the speech together in an attempt to mimick how a human speaks. The stress is often off, sometimes by only a little, sometimes hilariously so, and it messes up the flow of the text. I'm much more used to voices that use a formant, rather than a concatinative, as the natural voices do, method of speech. It's not as natural but it's faster and certainly smaller. I've been working for the better part of 3 years to improve espeak, available at http://espeak.sf.net, to bring it's US english up to snuff. If you're judging the state of linux tts by how easy festival is to get going then I'm not at all surprised you found it hard. I've never succeeded in getting that piece of software working, and I've tried more than a few times. I'll be more than willing to help improve the state of tts in linux if you'll just direct me on what needs improving. I do, however, take issue with your statement that it's not easy enough. I've never quite understood the windows users complaints that there are no graphical installers, click through wizards, that kind of thing for linux. We do things differently here. What I think you might want is a good natural voice that sounds more human than computer. You should be able to have that. There are commercial quality voices for linux that might fit your needs. I know of two such companies that make them. There's cepstral, from I believe nuance, and ivona, I'm not sure exactly which company manufactures that. I've used windows as new as windows version ten and I have yet to find anything significantly easier or better about that platform that linux cannot do. I find just the opposite as a matter of fact. I don't know if there is any good open source voice dictation software for linux. I've heard that there is, and then I've heard that there is not. If there is not any, or if what's out there isn't good, then that needs to be fixed because you have just as much a right as anyone else to use your computer without paying insanely high prices for software or hardware. I'll be more than willing to help, I can even point you to communities, mostly blind and visually impaired specific it's true, but nevertheless communities where you can get your issues fixed. I've often struggled to get my voice heard when it comes to accessibility. I never imagined that it would be just as hard for you.

I couldn't agree with this article more. I've been struggling to spread the viability of gnu/linux to the mostly blind and visually impaired community. The response has nearly always been overwhelmingly hostile or disinterested. The politest response is usually something like "why would I want to use that when nvda and mush z work fine for me?" It gets worse from there. This isn't a rant comment. I'm just saying that I understand how hard it can be to get someone to look and think outside the box and acknowledge the possibility of a non proprietary solution. Although nvda and mush z are open source, they run only on a closed source OS. I'll keep working. To the person who wrote this article, I'm sorry I can't remember your name, but what can I do better to get more people to at least consider linux as an option? I run into a wall, flounder for a bit while I try to understand what linux isn't doing right, then burn out, try windows, dislike it, go back to linux and the process repeats itself all over again.