RE: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
Also almost forgot. They sell the demo voices on their site for 29.99. Linux and windows versions. Since I believe what they use is based off festival, perhaps the voices could be made to plug into the existing festival plugin for asterisk? I have been working with app_festival for about a week or so now trying to figure out what is going on with it... The existing app_festival has a serious bug in it that makes it unsuitable for production use. I posted about it before, but the jist is that the more channels that are trying to use app_festival at the same time, the more problems there are -- the channel will abort, and this will result in asterisk abandoning the call. I have tried tracing it with all sorts of things including stepping back through some truly monumental gdb logs. The problem is not on the festival side of things, as the preforking festival keeps up fine with asterisk, and the problem occurs even when using the festival cache (broken in the current code - my patch fixes it) I cannot find the bug after a week of poking at the code. As I intended to use app_festival as a temporary replacement for recorded voice prompts in an AGI application, it was no big deal at first as it works 100% if if you always wait for the speech to finish and only use one channel (fine for testing), but after having the flexibility to do speech synthesis, I can see that it would be a tremendously good application for even the IVR that I am working on... Anyway, I hope this speech project gets off the ground. Staying all OSS is very nice, but after spending a lot of time mucking with it, I'd be easily willing to spend $50 for a great sounding, working solution. I'm still going to be poking at app_festival, though, so if anyone has suggestions or understands some of the internals and wants to work on this with me, please mail me off-list. Having a 100% working and production-ready solution for festival would be good for a number of reasons and I'd like to see this happen too! ~John ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
[Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
Many of you are familiar with how lousy Festival sounds. ATT has a product, NaturalVoices, that sounds much better. There are male female voice fonts for US/UK/Indian English, French, Spanish, and German. I am considering offering a linux-based text-to-speech engine based on the NaturalVoices runtime. An asterisk module would also be provided, making it easy to add natural sounding synthesis to Asterisk applications. You could also use it for other purposes, such as home automation. After discussing royalties with ATT, I have concluded that I can probably offer such a product at the following prices: Runtime - $30 intro price with one voice font one processor Extra voices/languages - $15 each Extra processors - $15 each Depending on demand, the price may rise to $50 at some point. The lower the demand, the higher the price, due to ATT's royalty structure. There is a demo of the synthesis engine here: http://www.naturalvoices.att.com/demos/ If you would be willing to pay for this kind of software, please e-mail me privately (not to the list). Please indicate your purchase timeframe as well as the number of licenses, extra voices, and processors you would want. Regards, Jeff / 'Bicster' on IRC ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
we would buy it even just for emergency prompts generation but we would need different languages, like polish and danish as well :)) On Tuesday 15 July 2003 17:07, Jeff Noxon wrote: Many of you are familiar with how lousy Festival sounds. ATT has a product, NaturalVoices, that sounds much better. There are male female voice fonts for US/UK/Indian English, French, Spanish, and German. I am considering offering a linux-based text-to-speech engine based on the NaturalVoices runtime. An asterisk module would also be provided, making it easy to add natural sounding synthesis to Asterisk applications. You could also use it for other purposes, such as home automation. After discussing royalties with ATT, I have concluded that I can probably offer such a product at the following prices: Runtime - $30 intro price with one voice font one processor Extra voices/languages - $15 each Extra processors - $15 each Depending on demand, the price may rise to $50 at some point. The lower the demand, the higher the price, due to ATT's royalty structure. There is a demo of the synthesis engine here: http://www.naturalvoices.att.com/demos/ If you would be willing to pay for this kind of software, please e-mail me privately (not to the list). Please indicate your purchase timeframe as well as the number of licenses, extra voices, and processors you would want. Regards, Jeff / 'Bicster' on IRC ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users -- Michael Bielicki Managing Director TAAN Consultants Ltd http://www.global-gateway.net/ -- This correspondence is for the named person's use only. It may contain confidential or legally privileged information or both. No confidentiality or privilege is waived or lost by any mistransmission. If you receive this correspondence in error, please immediately delete it from your system and notify the sender. You must not disclose, copy or rely on any part of this correspondence if you are not the intended recipient. Any opinions expressed in this message are those of the individual sender. ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
Wow, I thought they charged for this application on a per-channel basis. I would spend $50 for this in a heartbeat. I would need English, Dutch, German, Spanish, Arabic Russian... -GSR - Original Message - From: Jeff Noxon [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, July 15, 2003 6:07 PM Subject: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis? Many of you are familiar with how lousy Festival sounds. ATT has a product, NaturalVoices, that sounds much better. There are male female voice fonts for US/UK/Indian English, French, Spanish, and German. I am considering offering a linux-based text-to-speech engine based on the NaturalVoices runtime. An asterisk module would also be provided, making it easy to add natural sounding synthesis to Asterisk applications. You could also use it for other purposes, such as home automation. After discussing royalties with ATT, I have concluded that I can probably offer such a product at the following prices: Runtime - $30 intro price with one voice font one processor Extra voices/languages - $15 each Extra processors - $15 each Depending on demand, the price may rise to $50 at some point. The lower the demand, the higher the price, due to ATT's royalty structure. There is a demo of the synthesis engine here: http://www.naturalvoices.att.com/demos/ If you would be willing to pay for this kind of software, please e-mail me privately (not to the list). Please indicate your purchase timeframe as well as the number of licenses, extra voices, and processors you would want. Regards, Jeff / 'Bicster' on IRC ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
Unfortunately, languages other than English/French/Spanish/German cannot be supported due to the fact that the Natural Voices runtime doesn't support any other languages. On Tue, Jul 15, 2003 at 06:44:02PM +0100, Michael Bielicki wrote: we would buy it even just for emergency prompts generation but we would need different languages, like polish and danish as well :)) ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
--- Jeff Noxon [EMAIL PROTECTED] wrote: Many of you are familiar with how lousy Festival sounds. ATT has a product, NaturalVoices, that sounds much better. There are male female voice fonts for US/UK/Indian English, French, Spanish, and German. Festival only sounds bad because you are using it in a very simple way. It comes with a demo text to speech application that really is no more then a demo but most people just use the demo app and think Festival itself sounds bad. You need to read a bit more. Also at the cmu.edu web site are some tools for building your own voices. These can sound very good and speeak other languages. If you work at it the sound can be very natural. = Chris Albertson Home: 310-376-1029 [EMAIL PROTECTED] Cell: 310-990-7550 Office: 310-336-5189 [EMAIL PROTECTED] KG6OMK __ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
Sorry, I left out the URL. Here it is http://fife.speech.cs.cmu.edu/festival/ If interrested look at demos on the top menu bar. You see what is involved in building new databases for new languages or speaker voices for Festival. When you complain about the sound you are really complaining about the database quality or the quality of in mark-up embedded in the input text. In building a new database first you will need to find a human with a nice sounding voice who speaks the language with the correct regional accent. That may be the hardest part but after that it seems to take about a day or two --- Chris Albertson [EMAIL PROTECTED] wrote: --- Jeff Noxon [EMAIL PROTECTED] wrote: Many of you are familiar with how lousy Festival sounds. ATT has a product, NaturalVoices, that sounds much better. There are male female voice fonts for US/UK/Indian English, French, Spanish, and German. Festival only sounds bad because you are using it in a very simple way. It comes with a demo text to speech application that really is no more then a demo but most people just use the demo app and think Festival itself sounds bad. You need to read a bit more. Also at the cmu.edu web site are some tools for building your own voices. These can sound very good and speeak other languages. If you work at it the sound can be very natural. = Chris Albertson Home: 310-376-1029 [EMAIL PROTECTED] Cell: 310-990-7550 Office: 310-336-5189 [EMAIL PROTECTED] KG6OMK __ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
--- Jared Smith [EMAIL PROTECTED] wrote: Would you mind giving us a few examples on how we can make festival sound better? (Some sample festival configs would be nice!) Jared See my later post where I say Sorry here is the URL http://fife.speech.cs.cmu.edu/festival/ You might want to look at Festvox which is a tool for creating new festival databases and also at sable which is a markup language that allows to you to pass hints to festival like this: As a test of marked-up numbers. Here we have a year SAYAS MODE=date1998/SAYAS, an ordinal SAYAS MODE=ordinal1998/SAYAS, a cardinal SAYAS MODE=cardinal1998/SAYAS, a literal SAYAS MODE=literal1998/SAYAS, and phone number SAYAS MODE=phone1998/SAYAS. So, one way to improve the voice quality is to have the software which generates the text to be spoken embed some mark-up THe software would know if it is generating a phone number, street addres or a error message and could ad the required marks If you poke around you will find a huge body of software that is centered around festival in some way. In fact I would not be surprized at all to find the the ATT voice synthizer is festival based. I see that AtT is partially funding the work done at CMU and festival has no restriction or comercial use (It is NOT GPL'd it's more like X11) On Tue, 2003-07-15 at 14:41, Chris Albertson wrote: --- Jeff Noxon [EMAIL PROTECTED] wrote: Many of you are familiar with how lousy Festival sounds. ATT has a product, NaturalVoices, that sounds much better. There are male female voice fonts for US/UK/Indian English, French, Spanish, and German. Festival only sounds bad because you are using it in a very simple way. It comes with a demo text to speech application that really is no more then a demo but most people just use the demo app and think Festival itself sounds bad. You need to read a bit more. Also at the cmu.edu web site are some tools for building your own voices. These can sound very good and speeak other languages. If you work at it the sound can be very natural. = Chris Albertson Home: 310-376-1029 [EMAIL PROTECTED] Cell: 310-990-7550 Office: 310-336-5189 [EMAIL PROTECTED] KG6OMK __ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users = Chris Albertson Home: 310-376-1029 [EMAIL PROTECTED] Cell: 310-990-7550 Office: 310-336-5189 [EMAIL PROTECTED] KG6OMK __ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
Re: [Asterisk-Users] Poll - Would you pay $30-$50 for high quality speech synthesis?
If you would be willing to pay for this kind of software, please e-mail me privately (not to the list). Please indicate your purchase timeframe as well as the number of licenses, extra voices, and processors you would want. It should be simple, call an executable, with two parameters, the file where the text is and what to name the output sound file. I would pay for that. We could sell at least 10-15 per month. -Matt ___ Asterisk-Users mailing list [EMAIL PROTECTED] http://lists.digium.com/mailman/listinfo/asterisk-users
