Re: [Hpr] takov751
Voice three, for sure. I used to provide tech support to some people who used speech readers. One fellow (who probably listened at 3x to 4x) needed some help, but he didn't use a video monitor at all. There was no way I could make out what was happening by listening to the synth voice, so I asked if I could use a video monitor. "Sure", he said. "A video monitor is just an assistive device for the visually dependent!" --Bob. On 2022-01-11 13:09, Ken Fallon wrote: Hi Mike, As a TTS engine for reading the screen back to me I am more than happy with it and use it continually during the day. It's not just visually impaired people that rely on TTS. It does that job and does it well. The objection I have to using espeak as the voice of HPR is that it is harsh, unfriendly and not welcoming, its so bad in fact that it makes kids cry. I speak from personal experience. When my kids were small I made a project based on espeak (in English) for them to interact with. It was a disaster. When the espeak voice started speaking they got scared, started to cry, ran away, and never wanted to have anything to do with it again. Over the years the biggest objection to the TTS on HPR has been the espeak voice. It has also been the biggest point of negative feedback I get when trying to promote HPR to potential interviewees or projects. If those are not valid enough reasons then I don't know what will convince you. I can also assure you my desktop wallpaper is the default supplied with my distro. In the past it has been argued that the more natural voices are difficult to understand when sped up. So I took the two most natural voices from the list and posted a side by side comparison to espeak at 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. Can everyone have a listen to this and tell me your preference https://hackerpublicradio.org/tts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.ogg Ken. On 2022-01-11 14:35, Mike Ray wrote: And here was me thinking about posting to the list about how much better it is now with the softer music in the background and a nice punchy eSpeak voice. I still have no idea what the objection to the eSpeak voice is. If you spend as many hours a day coding as I do, and rely on tts to make this possible, then eSpeak is the way forward. Although I know this may only be true for English speakers. Not sure how good eSpeak is at other languages. People who complain about eSpeak are probably the same people who never get any work done because they are constantly fiddling with the desktop wallpaper. :-p On 11/01/2022 10:44, Ken Fallon wrote: Hi All, We got a comment from takov751 via https://matrix.to/#/#hpr:matrix.org Greetings i am a long listener of the shows . And of course planing to make my first show . I would like to ask question regarding tts at the beginning of the show usually I hear the espeak robotic voice . In the workflow have you considered using mimic1 or opentts / Mozillatts or something along those lines ? It’s seems like these would be compatible with licensing as well and bir more realistic voices . A few examples https://hub.docker.com/u/synesthesiam Re-posted with permission Sample voices are here: https://synesthesiam.github.io/opentts @Mike Ray I would like to try and get a happy balance between meeting your needs and having a voice that is friendly. While I love espeak it is not friendly - it literally put my kids in tears when they were younger :-) Could you have a listen to some of the other voices and see if any of them come close to your requirements for TTS FYI I find these two "friendly" * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_ljspeech * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_vctk ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org -- Bob Jonkman Phone: +1-519-635-9413 SOBAC Microcomputer Services http://sobac.com/sobac/ Software --- Office & Business Automation --- Consulting GnuPG Fngrprnt:04F7 742B 8F54 C40A E115 26C2 B912 89B0 D2CC E5EA ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org
Re: [Hpr] takov751
Hi Mike I stand corrected, both amazed and impressed. Beeza LibreOffice - Free and open source office suite: LibreOffice Website<https://www.libreoffice.org> Respects your privacy, and gives you back control over your data From: Hpr on behalf of Mike Ray Sent: 11 January 2022 18:51 To: hpr@hackerpublicradio.org Subject: Re: [Hpr] takov751 I can. As can a lot of blind people who rely on tts. Perhaps not for prose, but certainly for source code. On 11/01/2022 18:49, Nigel Verity wrote: > Hi Ken > > The third voice is head and shoulders above the other two to my ears. Above > 200% speed I struggle to take everything in. By the time you get to 500% it > is just a joke. Surely nobody can follow speech at that speed. > > Beeza > > LibreOffice - Free and open source office suite: LibreOffice > Website<https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.libreoffice.org%2Fdata=04%7C01%7C%7C7e156590b65a43fa181308d9d53399ab%7C84df9e7fe9f640afb435%7C1%7C0%7C637775239860392188%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=ax070r5lReFA2gi3gr57yXpiyqz%2BRPVziDmbvTeUKhU%3Dreserved=0> > Respects your privacy, and gives you back control over your data > > > From: Hpr on behalf of Ken Fallon > > Sent: 11 January 2022 18:09 > To: m...@raspberryvi.org ; hpr@hackerpublicradio.org > > Subject: Re: [Hpr] takov751 > > Hi Mike, > > As a TTS engine for reading the screen back to me I am more than happy > with it and use it continually during the day. It's not just visually > impaired people that rely on TTS. It does that job and does it well. > > The objection I have to using espeak as the voice of HPR is that it is > harsh, unfriendly and not welcoming, its so bad in fact that it makes > kids cry. I speak from personal experience. When my kids were small I > made a project based on espeak (in English) for them to interact with. > It was a disaster. When the espeak voice started speaking they got > scared, started to cry, ran away, and never wanted to have anything to > do with it again. > > Over the years the biggest objection to the TTS on HPR has been the > espeak voice. It has also been the biggest point of negative feedback I > get when trying to promote HPR to potential interviewees or projects. > > If those are not valid enough reasons then I don't know what will > convince you. I can also assure you my desktop wallpaper is the default > supplied with my distro. > > In the past it has been argued that the more natural voices are > difficult to understand when sped up. So I took the two most natural > voices from the list and posted a side by side comparison to espeak at > 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the > coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. > > Can everyone have a listen to this and tell me your preference > https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhackerpublicradio.org%2Ftts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.oggdata=04%7C01%7C%7C7e156590b65a43fa181308d9d53399ab%7C84df9e7fe9f640afb435%7C1%7C0%7C637775239860392188%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=FNw2tezhUdQIfuYsBxcyJ0HC4RN6AjMVGBbYPN4p2Ks%3Dreserved=0 > > Ken. > > > > On 2022-01-11 14:35, Mike Ray wrote: >> >> >> And here was me thinking about posting to the list about how much >> better it is now with the softer music in the background and a nice >> punchy eSpeak voice. >> >> I still have no idea what the objection to the eSpeak voice is. >> >> If you spend as many hours a day coding as I do, and rely on tts to >> make this possible, then eSpeak is the way forward. Although I know >> this may only be true for English speakers. Not sure how good eSpeak >> is at other languages. >> >> People who complain about eSpeak are probably the same people who >> never get any work done because they are constantly fiddling with the >> desktop wallpaper. >> >> :-p >> >> >> >> >> >> On 11/01/2022 10:44, Ken Fallon wrote: >>> Hi All, >>> >>> We got a comment from takov751 via >>> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmatrix.to%2F%23%2F%23hpr%3Amatrix.orgdata=04%7C01%7C%7C7e156590b65a43fa181308d9d53399ab%7C84df9e7fe9f640afb435%7C1%7C0%7C637775239860392188%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=XLtpk3JOSRocE3xz%2BeVgLYJXH4F%2Frq9y
Re: [Hpr] takov751
On Tue, Jan 11, 2022 at 12:11 PM Ken Fallon wrote: > In the past it has been argued that the more natural voices are > difficult to understand when sped up. So I took the two most natural > voices from the list and posted a side by side comparison to espeak at > 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the > coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. > > Can everyone have a listen to this and tell me your preference > https://hackerpublicradio.org/tts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.ogg I rarely listen faster than 2x (I prefer 1x but will speed up if I have a lot of content to get through), so I can imagine someone who deals with audio navigation day after day would have a much more nuanced (and, I think, valuable opinion). That said, here's my feedback: - I found voice #2 the most pleasant of the 3, particularly at 1x - Both voices #2 and #3 were more pleasant than #1 at all speeds - All the voices were intelligible at 1x - At higher speeds, I had the easiest time understanding voice #3, but this could just be due to my own American accent - I'd like to hear from folks like Mike who routinely listen at high speeds ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org
Re: [Hpr] takov751
Hi Ken: I have to say I liked the third voice in your demo. I also like the first but that's because I happen to use it about 16 hours a day and over many years one gets used to listening to any voice. Synthetic voices have come a long way over the years and I'd say that google-tts is very comfortable to listen to. The problem is open source synthetic voices are just now getting pretty good. That third voice wouldn't be a hardship imo to listen to at the beginning and end of hpr shows. kirk/baffled On Tue, 11 Jan 2022, Ken Fallon wrote: Hi Mike, As a TTS engine for reading the screen back to me I am more than happy with it and use it continually during the day. It's not just visually impaired people that rely on TTS. It does that job and does it well. The objection I have to using espeak as the voice of HPR is that it is harsh, unfriendly and not welcoming, its so bad in fact that it makes kids cry. I speak from personal experience. When my kids were small I made a project based on espeak (in English) for them to interact with. It was a disaster. When the espeak voice started speaking they got scared, started to cry, ran away, and never wanted to have anything to do with it again. Over the years the biggest objection to the TTS on HPR has been the espeak voice. It has also been the biggest point of negative feedback I get when trying to promote HPR to potential interviewees or projects. If those are not valid enough reasons then I don't know what will convince you. I can also assure you my desktop wallpaper is the default supplied with my distro. In the past it has been argued that the more natural voices are difficult to understand when sped up. So I took the two most natural voices from the list and posted a side by side comparison to espeak at 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. Can everyone have a listen to this and tell me your preference https://hackerpublicradio.org/tts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.ogg Ken. On 2022-01-11 14:35, Mike Ray wrote: And here was me thinking about posting to the list about how much better it is now with the softer music in the background and a nice punchy eSpeak voice. I still have no idea what the objection to the eSpeak voice is. If you spend as many hours a day coding as I do, and rely on tts to make this possible, then eSpeak is the way forward. Although I know this may only be true for English speakers. Not sure how good eSpeak is at other languages. People who complain about eSpeak are probably the same people who never get any work done because they are constantly fiddling with the desktop wallpaper. :-p On 11/01/2022 10:44, Ken Fallon wrote: Hi All, We got a comment from takov751 via https://matrix.to/#/#hpr:matrix.org Greetings i am a long listener of the shows . And of course planing to make my first show . I would like to ask question regarding tts at the beginning of the show usually I hear the espeak robotic voice . In the workflow have you considered using mimic1 or opentts / Mozillatts or something along those lines ? It’s seems like these would be compatible with licensing as well and bir more realistic voices . A few examples https://hub.docker.com/u/synesthesiam Re-posted with permission Sample voices are here: https://synesthesiam.github.io/opentts @Mike Ray I would like to try and get a happy balance between meeting your needs and having a voice that is friendly. While I love espeak it is not friendly - it literally put my kids in tears when they were younger :-) Could you have a listen to some of the other voices and see if any of them come close to your requirements for TTS FYI I find these two "friendly" * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_ljspeech * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_vctk ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org
Re: [Hpr] takov751
I can. As can a lot of blind people who rely on tts. Perhaps not for prose, but certainly for source code. On 11/01/2022 18:49, Nigel Verity wrote: Hi Ken The third voice is head and shoulders above the other two to my ears. Above 200% speed I struggle to take everything in. By the time you get to 500% it is just a joke. Surely nobody can follow speech at that speed. Beeza LibreOffice - Free and open source office suite: LibreOffice Website<https://www.libreoffice.org> Respects your privacy, and gives you back control over your data From: Hpr on behalf of Ken Fallon Sent: 11 January 2022 18:09 To: m...@raspberryvi.org ; hpr@hackerpublicradio.org Subject: Re: [Hpr] takov751 Hi Mike, As a TTS engine for reading the screen back to me I am more than happy with it and use it continually during the day. It's not just visually impaired people that rely on TTS. It does that job and does it well. The objection I have to using espeak as the voice of HPR is that it is harsh, unfriendly and not welcoming, its so bad in fact that it makes kids cry. I speak from personal experience. When my kids were small I made a project based on espeak (in English) for them to interact with. It was a disaster. When the espeak voice started speaking they got scared, started to cry, ran away, and never wanted to have anything to do with it again. Over the years the biggest objection to the TTS on HPR has been the espeak voice. It has also been the biggest point of negative feedback I get when trying to promote HPR to potential interviewees or projects. If those are not valid enough reasons then I don't know what will convince you. I can also assure you my desktop wallpaper is the default supplied with my distro. In the past it has been argued that the more natural voices are difficult to understand when sped up. So I took the two most natural voices from the list and posted a side by side comparison to espeak at 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. Can everyone have a listen to this and tell me your preference https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhackerpublicradio.org%2Ftts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.oggdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=LU99WKYC1M8TCvMdKh8hM6N4h%2BalZ9gmdlGY%2FG65VtA%3Dreserved=0 Ken. On 2022-01-11 14:35, Mike Ray wrote: And here was me thinking about posting to the list about how much better it is now with the softer music in the background and a nice punchy eSpeak voice. I still have no idea what the objection to the eSpeak voice is. If you spend as many hours a day coding as I do, and rely on tts to make this possible, then eSpeak is the way forward. Although I know this may only be true for English speakers. Not sure how good eSpeak is at other languages. People who complain about eSpeak are probably the same people who never get any work done because they are constantly fiddling with the desktop wallpaper. :-p On 11/01/2022 10:44, Ken Fallon wrote: Hi All, We got a comment from takov751 via https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmatrix.to%2F%23%2F%23hpr%3Amatrix.orgdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=Sxnt0xVQ%2B1EpCxXi%2FtWfLHlgnsyM4FMpp9X42%2FQY0KY%3Dreserved=0 Greetings i am a long listener of the shows . And of course planing to make my first show . I would like to ask question regarding tts at the beginning of the show usually I hear the espeak robotic voice . In the workflow have you considered using mimic1 or opentts / Mozillatts or something along those lines ? It’s seems like these would be compatible with licensing as well and bir more realistic voices . A few examples https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhub.docker.com%2Fu%2Fsynesthesiamdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=AE86feZPVXgAFElFBdWJkO3f48W6FbaCChLwJd9jBpU%3Dreserved=0 Re-posted with permission Sample voices are here: https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsynesthesiam.github.io%2Fopenttsdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=AmlRQ9dHu3Es2TwBCFCDQO%2BAUFfWza0TmcABN
Re: [Hpr] takov751
Hi Ken The third voice is head and shoulders above the other two to my ears. Above 200% speed I struggle to take everything in. By the time you get to 500% it is just a joke. Surely nobody can follow speech at that speed. Beeza LibreOffice - Free and open source office suite: LibreOffice Website<https://www.libreoffice.org> Respects your privacy, and gives you back control over your data From: Hpr on behalf of Ken Fallon Sent: 11 January 2022 18:09 To: m...@raspberryvi.org ; hpr@hackerpublicradio.org Subject: Re: [Hpr] takov751 Hi Mike, As a TTS engine for reading the screen back to me I am more than happy with it and use it continually during the day. It's not just visually impaired people that rely on TTS. It does that job and does it well. The objection I have to using espeak as the voice of HPR is that it is harsh, unfriendly and not welcoming, its so bad in fact that it makes kids cry. I speak from personal experience. When my kids were small I made a project based on espeak (in English) for them to interact with. It was a disaster. When the espeak voice started speaking they got scared, started to cry, ran away, and never wanted to have anything to do with it again. Over the years the biggest objection to the TTS on HPR has been the espeak voice. It has also been the biggest point of negative feedback I get when trying to promote HPR to potential interviewees or projects. If those are not valid enough reasons then I don't know what will convince you. I can also assure you my desktop wallpaper is the default supplied with my distro. In the past it has been argued that the more natural voices are difficult to understand when sped up. So I took the two most natural voices from the list and posted a side by side comparison to espeak at 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. Can everyone have a listen to this and tell me your preference https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhackerpublicradio.org%2Ftts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.oggdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=LU99WKYC1M8TCvMdKh8hM6N4h%2BalZ9gmdlGY%2FG65VtA%3Dreserved=0 Ken. On 2022-01-11 14:35, Mike Ray wrote: > > > And here was me thinking about posting to the list about how much > better it is now with the softer music in the background and a nice > punchy eSpeak voice. > > I still have no idea what the objection to the eSpeak voice is. > > If you spend as many hours a day coding as I do, and rely on tts to > make this possible, then eSpeak is the way forward. Although I know > this may only be true for English speakers. Not sure how good eSpeak > is at other languages. > > People who complain about eSpeak are probably the same people who > never get any work done because they are constantly fiddling with the > desktop wallpaper. > > :-p > > > > > > On 11/01/2022 10:44, Ken Fallon wrote: >> Hi All, >> >> We got a comment from takov751 via >> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmatrix.to%2F%23%2F%23hpr%3Amatrix.orgdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=Sxnt0xVQ%2B1EpCxXi%2FtWfLHlgnsyM4FMpp9X42%2FQY0KY%3Dreserved=0 >> >> >> >> Greetings i am a long listener of the shows . And of course planing >> to make my first show . I would like to ask question regarding tts at >> the beginning of the show usually I hear the espeak robotic voice . >> In the workflow have you considered using mimic1 or opentts / >> Mozillatts or something along those lines ? It’s seems like these >> would be compatible with licensing as well and bir more realistic >> voices . A few examples >> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhub.docker.com%2Fu%2Fsynesthesiamdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000sdata=AE86feZPVXgAFElFBdWJkO3f48W6FbaCChLwJd9jBpU%3Dreserved=0 >> >> >> >> Re-posted with permission >> >> Sample voices are here: >> https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fsynesthesiam.github.io%2Fopenttsdata=04%7C01%7C%7Caa9e5cf113f0411706b908d9d52db85a%7C84df9e7fe9f640afb435%7C1%7C0%7C637775214626255386%7
Re: [Hpr] takov751
Hi Mike, As a TTS engine for reading the screen back to me I am more than happy with it and use it continually during the day. It's not just visually impaired people that rely on TTS. It does that job and does it well. The objection I have to using espeak as the voice of HPR is that it is harsh, unfriendly and not welcoming, its so bad in fact that it makes kids cry. I speak from personal experience. When my kids were small I made a project based on espeak (in English) for them to interact with. It was a disaster. When the espeak voice started speaking they got scared, started to cry, ran away, and never wanted to have anything to do with it again. Over the years the biggest objection to the TTS on HPR has been the espeak voice. It has also been the biggest point of negative feedback I get when trying to promote HPR to potential interviewees or projects. If those are not valid enough reasons then I don't know what will convince you. I can also assure you my desktop wallpaper is the default supplied with my distro. In the past it has been argued that the more natural voices are difficult to understand when sped up. So I took the two most natural voices from the list and posted a side by side comparison to espeak at 150%, 200%, 250%, 300%, 350%, 400%, 450%, and 500%. In my opinion the coqui-tts_en_en_ljspeech is more understandable than espeak at every speed. Can everyone have a listen to this and tell me your preference https://hackerpublicradio.org/tts-espeak-ljspeech-vctk-normal-150-200-250-300-350-400-450-500-percent.ogg Ken. On 2022-01-11 14:35, Mike Ray wrote: And here was me thinking about posting to the list about how much better it is now with the softer music in the background and a nice punchy eSpeak voice. I still have no idea what the objection to the eSpeak voice is. If you spend as many hours a day coding as I do, and rely on tts to make this possible, then eSpeak is the way forward. Although I know this may only be true for English speakers. Not sure how good eSpeak is at other languages. People who complain about eSpeak are probably the same people who never get any work done because they are constantly fiddling with the desktop wallpaper. :-p On 11/01/2022 10:44, Ken Fallon wrote: Hi All, We got a comment from takov751 via https://matrix.to/#/#hpr:matrix.org Greetings i am a long listener of the shows . And of course planing to make my first show . I would like to ask question regarding tts at the beginning of the show usually I hear the espeak robotic voice . In the workflow have you considered using mimic1 or opentts / Mozillatts or something along those lines ? It’s seems like these would be compatible with licensing as well and bir more realistic voices . A few examples https://hub.docker.com/u/synesthesiam Re-posted with permission Sample voices are here: https://synesthesiam.github.io/opentts @Mike Ray I would like to try and get a happy balance between meeting your needs and having a voice that is friendly. While I love espeak it is not friendly - it literally put my kids in tears when they were younger :-) Could you have a listen to some of the other voices and see if any of them come close to your requirements for TTS FYI I find these two "friendly" * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_ljspeech * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_vctk ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org
Re: [Hpr] takov751
And here was me thinking about posting to the list about how much better it is now with the softer music in the background and a nice punchy eSpeak voice. I still have no idea what the objection to the eSpeak voice is. If you spend as many hours a day coding as I do, and rely on tts to make this possible, then eSpeak is the way forward. Although I know this may only be true for English speakers. Not sure how good eSpeak is at other languages. People who complain about eSpeak are probably the same people who never get any work done because they are constantly fiddling with the desktop wallpaper. :-p On 11/01/2022 10:44, Ken Fallon wrote: Hi All, We got a comment from takov751 via https://matrix.to/#/#hpr:matrix.org Greetings i am a long listener of the shows . And of course planing to make my first show . I would like to ask question regarding tts at the beginning of the show usually I hear the espeak robotic voice . In the workflow have you considered using mimic1 or opentts / Mozillatts or something along those lines ? It’s seems like these would be compatible with licensing as well and bir more realistic voices . A few examples https://hub.docker.com/u/synesthesiam Re-posted with permission Sample voices are here: https://synesthesiam.github.io/opentts @Mike Ray I would like to try and get a happy balance between meeting your needs and having a voice that is friendly. While I love espeak it is not friendly - it literally put my kids in tears when they were younger :-) Could you have a listen to some of the other voices and see if any of them come close to your requirements for TTS FYI I find these two "friendly" * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_ljspeech * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_vctk -- Michael A. Ray Analyst/Programmer Witley, Surrey, South-east UK He/him "Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away." -- A. de Saint-Exupery ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org
[Hpr] takov751
Hi All, We got a comment from takov751 via https://matrix.to/#/#hpr:matrix.org Greetings i am a long listener of the shows . And of course planing to make my first show . I would like to ask question regarding tts at the beginning of the show usually I hear the espeak robotic voice . In the workflow have you considered using mimic1 or opentts / Mozillatts or something along those lines ? It’s seems like these would be compatible with licensing as well and bir more realistic voices . A few examples https://hub.docker.com/u/synesthesiam Re-posted with permission Sample voices are here: https://synesthesiam.github.io/opentts @Mike Ray I would like to try and get a happy balance between meeting your needs and having a voice that is friendly. While I love espeak it is not friendly - it literally put my kids in tears when they were younger :-) Could you have a listen to some of the other voices and see if any of them come close to your requirements for TTS FYI I find these two "friendly" * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_ljspeech * https://synesthesiam.github.io/opentts/#coqui-tts_en_en_vctk -- Regards, Ken Fallon (PA7KEN,G5KEN) https://kenfallon.com https://hackerpublicradio.org/hosts/ken_fallon ___ Hpr mailing list Hpr@hackerpublicradio.org http://hackerpublicradio.org/mailman/listinfo/hpr_hackerpublicradio.org