(OLD) (ARCHIVED) Puppy Linux Discussion Forum Forum Index (OLD) (ARCHIVED) Puppy Linux Discussion Forum
Puppy HOME page : puppylinux.com
"THE" alternative forum : puppylinux.info

This forum can also be accessed as http://oldforum.puppylinux.com
It is now read-only and serves only as archives.

Please register over the NEW forum
https://forum.puppylinux.com
and continue your work there. Thank you.

 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups    
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

The time now is Sat 23 Jan 2021, 19:59
All times are UTC - 4
 Forum index » House Training » Users ( For the regulars )
TTSReader & Balabolka: ebook-to-speech apps that run in WINE
Moderators: Flash, Ian, JohnMurga
This forum is locked: you cannot post, reply to, or edit topics.   This topic is locked: you cannot edit posts or make replies. View previous topic :: View next topic
Page 1 of 3 [32 Posts]   Goto page: 1, 2, 3 Next
Author Message
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Fri 31 May 2019, 11:27    Post subject:  TTSReader & Balabolka: ebook-to-speech apps that run in WINE  

I got into this cause friend was telling me how much she enjoyed the option in Kindle for PC to have it read to her. This option in Kindle for PC of course piggybacks on the built in speech service and voices in windows. I run an older Kindle for PC in WINE (anything after version 1.17 wont run in WINE), it doesnt even offer the speech option and wouldnt work anyway.

So, curious, I tried installing current version Kindle for PC 1.26 in an n-lite'd win10 that I had been playing with, but speech service had been n-lite'd out, so if I tried the text to speech option, it crashed. I installed Balabolka, but without speech service it could find no voices even after I installed couple old MS files it recommended.

So got to looking how to do ebook to speech in linux. I had played with Festival long time ago, but its basically for short text to voice. It does still work by the way. Available in Puppy Package Manager. But its ancient history. Ran across Okular which is KDE pdf reader that also can do epub. It has an option in it for text to speech, but based on Jovie. Jovie (formerly called tts) is deprecated Qt4 program. I tried it in Puppy but no success making it do anything. Ok, I have a computer with Lubuntu 19.04 and apt-get is your friend. Anyway found a recipe to make Okular work with espeak supposedly in current Ubuntu's. Well espeak works. But Okular with espeak doesnt and there is darn little info on making it work. Also seems some repository versions of Okular specifically have text to speech turned off when compiled. That may have been my problem in it not working????

So I look and there is a Gnome screen reader called orca. but didn't want to go down that road. Its not really meant for ebooks. Then I stumble across a CLI program called eBook-speaker that uses espeak. Amazingly its in Ubuntu repositories. Its available in Puppy Package Manager in Xenialpup64, but I was using Lubuntu computer at time so apt-get it. It has a really crude shell menu that was only good enough to navigate to an ebook and start playing it. Supposed to have support for mobi format but I only could get it to do epub. Again didnt spend huge amount time, could be my mobi ebook was corrupt or something. It can be run from terminal with a boat load of options but again I didnt spend time trying to sort them all out. If there is a CLI that deserves a decent gui frontend its this program, nobody really wants to read a book from terminal. It did really well reading, though default voice was Robbie the Robot SAPI-4 kind. Voice was understandable though.

Now I also ran across an old win98/win2000 era freeware windows program called TTSReader 1.03. It was only like 2.5MB download so what the heck. I installed it into my problematic win10 LITE and amazingly it worked, just does text to speech but comes with like ten SAPI-4 voices. Couple of them can be adjusted to be acceptable. Says it supports SAPI-5 but think that depends on having windows speech service and it just can see and use any installed to that. But the neat trick Balabolka could see the voices in TTSReader and use them.

So ok had to try it. I installed both Balabolka and TTSReader in WINE 4.0. Bingo, worked just like they did in win10 LITE, no installing any ancient mystery Microsoft files or anything. Hey you want simplest way to vocalize an epub book on linux with a decent gui, this combo is it. Assuming you already have WINE. Plus two of those TTSReader voices are actually listenable without wanting to pound your head against the wall after five minutes.

Since then I have found maybe way to install a windows SAPI-5 engine from scratch. But no idea right now if it would work. There are some decent SAPI-5 voices out there, though Microsoft intertwined them in registry so you cant just cut and paste, they have to be "installed" so...... Oh in my searching found interesting registry hack somebody came up with to use Cortana voice as a SAPI-5 voice option in win10. Microsoft intentionally isolated it to prevent this.
Back to top
View user's profile Send private message 
nic007


Joined: 13 Nov 2011
Posts: 3444
Location: Cradle of Humankind

PostPosted: Fri 31 May 2019, 11:50    Post subject:  

I've been using the old American-English TruVoice SAPI4 for years. Actually amazing quality if you consider the install file is a little over 1MB. I run it at the speed of 162 WPM. This voice is very good if you take time to make pronunciation corrections. First thing you need to do is to change the ! to . As with this voice the ! leads to unsavoury screeching (screaming) which sounds terrible.. I have read/listened to hundreds of books this way. I use this voice with 2nd Speech Center in WINE and would rate it very good even in comparison to the best commercial voices out there.
Back to top
View user's profile Send private message 
lmemsm

Joined: 27 Jun 2012
Posts: 51

PostPosted: Fri 31 May 2019, 12:23    Post subject:  

I've been using bard for ebook to speech conversion. It only handles epub format. It uses flite for speech. It's a relatively straightforward program so it wouldn't be too hard to debug or add features if needed. For instance, it uses libzip by default and I have a patch to use zlib and minizip (in zlib contrib) instead since zlib is usually on most systems.
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Fri 31 May 2019, 12:25    Post subject:  

2nd speech center looks remarkably like TTSReader 1.30 (not 1.03 as I previously said).....

Quote:
2nd Speech Center Description
Converts any text into spoken words or MP3/WAV audio files! 2nd Speech Center is an award-winning text-to-speech player that lets you listen to documents, emails or web pages instead of reading on screen. In addition to text, 2nd Speech Center also allows you to convert text to MP3 or WAV files for listening later with your portable MP3 player. 2nd Speech Center can directly open Plain Text files (.txt), Microsoft Word files (.doc), PDF files (.pdf), EMail files (.eml), Rich Text files (.rtf), HTML files (.html). For Microsoft Internet Explorer users, 2nd Speech Center supplies a handy toolbar plug-in for Internet Explorer. 2nd Speech Center supports dozens of male, female and robotic voices for over 11 languages. 2nd Speech Center can monitor the Windows Clipboard and automatically process its contents. This feature makes it easy to read web pages, e-mail messages and any text of other applications. Additionally, 2nd Speech Center is integrated with many useful and timesaving features, such as talking alarm clock, talking reminder and powerful command lines for advanced users. And don't worry, the well designed user interface is very easy to use.


Quote:
From SpheNet:
TTSReader is a full-featured, text-to speech software package that allows reading text aloud as well as to wav or mp3 files. TTSReader is freeware, therefore you can download and use it free of charge. Main Features: Intuitive user interface design; Automatic highlighting of currently read text; Reading to wav; Reading to mp3, with adjustable settings; Pronunciation corrections; Support for both SAPI4 and SAPI5 voices; Support for rich text formatting; Skipping of sentences or paragraphs while reading; Auto-reading the clipboard; Reading with control tags; Global hotkeys; Documentation provided for all features. Version 1.30 adds floating toolbar, and Read clipboard hotkey.


So like TTSReader, Second Speech Center doesnt do epub or mobi? You have to convert them to one of supported formats first? I just bet its related to TTSReader somehow.


But anyway thanks for heads up on training it for pronunciation. See if Balabolka offers that too. I like Balabolka better for books though its a much larger program. I was experimenting having Balabolka read some from War and Peace. It was so funny the pronunciation of the Russian names.
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Fri 31 May 2019, 12:34    Post subject:  

lmemsm wrote:
I've been using bard for ebook to speech conversion. It only handles epub format. It uses flite for speech. It's a relatively straightforward program so it wouldn't be too hard to debug or add features if needed. For instance, it uses libzip by default and I have a patch to use zlib and minizip (in zlib contrib) instead since zlib is usually on most systems.


Thank you, this is is interesting: https://github.com/festvox/bard

I hadnt run across this one before. Though did see some mention of flite somewhere. I will have to look at it closer.

It is a shame some of these small projects dont get more attention. And of course usually they just get written for somebody's personal use and so dont get maintained/updated etc.
Back to top
View user's profile Send private message 
lmemsm

Joined: 27 Jun 2012
Posts: 51

PostPosted: Fri 31 May 2019, 13:27    Post subject:  

mouldy wrote:
Thank you, this is is interesting: https://github.com/festvox/bard

I hadnt run across this one before. Though did see some mention of flite somewhere. I will have to look at it closer.

It is a shame some of these small projects dont get more attention. And of course usually they just get written for somebody's personal use and so dont get maintained/updated etc.


Thought it was pretty nice for a lightweight application. Festival lite (flite) is based on festival, but I found it a lot easier to build from source than festival.

I've written to the developer of bard. He seemed very nice. He accepted a few patches I sent him. Got the impression this was a project for a University.

Some of my patches to bard are here (includes support for SDL 2 instead of SDL 1.x):
http://www.distasis.com/cpp/lmports.htm
I do have some later patches but haven't uploaded them yet. Since I'm pretty much the only one who's been using my patches, I haven't been keeping them as up-to-date online as I could.

I typically maintain/update a lot of the software I'm using myself, so not that concerned whether it's actively developed upstream or not. If it is still developed, I'll send my patches to the developers too in case they're interested. If anyone knows of some other interesting small projects that need attention, please mention them.
Back to top
View user's profile Send private message 
nic007


Joined: 13 Nov 2011
Posts: 3444
Location: Cradle of Humankind

PostPosted: Fri 31 May 2019, 19:54    Post subject:  

mouldy wrote:
2nd speech center looks remarkably like TTSReader 1.30 (not 1.03 as I previously said).....

Quote:
2nd Speech Center Description
Converts any text into spoken words or MP3/WAV audio files! 2nd Speech Center is an award-winning text-to-speech player that lets you listen to documents, emails or web pages instead of reading on screen. In addition to text, 2nd Speech Center also allows you to convert text to MP3 or WAV files for listening later with your portable MP3 player. 2nd Speech Center can directly open Plain Text files (.txt), Microsoft Word files (.doc), PDF files (.pdf), EMail files (.eml), Rich Text files (.rtf), HTML files (.html). For Microsoft Internet Explorer users, 2nd Speech Center supplies a handy toolbar plug-in for Internet Explorer. 2nd Speech Center supports dozens of male, female and robotic voices for over 11 languages. 2nd Speech Center can monitor the Windows Clipboard and automatically process its contents. This feature makes it easy to read web pages, e-mail messages and any text of other applications. Additionally, 2nd Speech Center is integrated with many useful and timesaving features, such as talking alarm clock, talking reminder and powerful command lines for advanced users. And don't worry, the well designed user interface is very easy to use.


Quote:
From SpheNet:
TTSReader is a full-featured, text-to speech software package that allows reading text aloud as well as to wav or mp3 files. TTSReader is freeware, therefore you can download and use it free of charge. Main Features: Intuitive user interface design; Automatic highlighting of currently read text; Reading to wav; Reading to mp3, with adjustable settings; Pronunciation corrections; Support for both SAPI4 and SAPI5 voices; Support for rich text formatting; Skipping of sentences or paragraphs while reading; Auto-reading the clipboard; Reading with control tags; Global hotkeys; Documentation provided for all features. Version 1.30 adds floating toolbar, and Read clipboard hotkey.


So like TTSReader, Second Speech Center doesnt do epub or mobi? You have to convert them to one of supported formats first? I just bet its related to TTSReader somehow.


But anyway thanks for heads up on training it for pronunciation. See if Balabolka offers that too. I like Balabolka better for books though its a much larger program. I was experimenting having Balabolka read some from War and Peace. It was so funny the pronunciation of the Russian names.


Yes, 2nd Speech Center is text to speech and does not have a file converter. Been using it for years. I use balabolka's commandline utilities for file conversions and text splitting. 2nd Speech Center's pronunciation editor has worked the best for me and I have tried almost all text-to-speech programs before, also TTS Reader. BTW - ebookbb.com is great for books.
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Sat 01 Jun 2019, 13:23    Post subject:  

lmemsm wrote:
I've been using bard for ebook to speech conversion. It only handles epub format. It uses flite for speech. It's a relatively straightforward program so it wouldn't be too hard to debug or add features if needed. For instance, it uses libzip by default and I have a patch to use zlib and minizip (in zlib contrib) instead since zlib is usually on most systems.


In attempting to compile Bard I get this: configure: error: "Cannot locate CMU Flite specify FLITEDIR explicitly."

Flite 2.1 is already installed so what is it looking for? Is it looking for a source directory used to compile Flite?
Back to top
View user's profile Send private message 
Mike Walsh


Joined: 28 Jun 2014
Posts: 6397
Location: King's Lynn, UK.

PostPosted: Sun 02 Jun 2019, 07:16    Post subject:  

Hey, thanks for the links, guys.

I knew this sort of thing existed, but had no idea all these different versions worked so well under WINE. I've tried TTSReader 1.30; yes, things DO sound a bit 'robotic', but with suitable adjustments, one or two of the female voices actually don't sound too bad.... I mean, I could listen to them for more than just a few seconds without getting too annoyed... Laughing

Balabolka's next on the list; looks pretty professional, I will say. This is version 2.15.0.701; no idea how 'up-to-date' this is, but the copyright shows 2006-2019. I would guess fairly new.

I have this running under my usual WINE 1.7.51; not a recent one, by any means, but it seems to be all functional. Incidentally, this is the 'portable' version, from PortableApps.com. Click on the 'Apps' tab, and it's the very first item under the first section heading, 'Accessibility'. As mikeslr and I have both come to realize, you don't 'install' these in the ~/.wine directory.....you run them from anywhere but WINE.

Just click on the .exe file inside, and they'll fire up, if they're going to work.

---------------------------------

Nic/Mouldy:-

I have a question. Where do you put additional SAPI 4/5 'voices' if you want to add extra ones? I've obtained the extra American/English voices from the creator of 2nd Speech Center (Zerosoft?); they're .msi files, and I haven't a clue what you're supposed to do with them.

I did find a reference somewhere that said as long as they're somewhere in your system, they'll be found, but I'm still not too sure quite where they would go in the WINE directory structure.

Suggestions?


Mike. Wink

_________________
MY 'PUPPY' PACKAGES

Back to top
View user's profile Send private message 
nic007


Joined: 13 Nov 2011
Posts: 3444
Location: Cradle of Humankind

PostPosted: Sun 02 Jun 2019, 09:37    Post subject:  

Mike Walsh wrote:
Hey, thanks for the links, guys.

I knew this sort of thing existed, but had no idea all these different versions worked so well under WINE. I've tried TTSReader 1.30; yes, things DO sound a bit 'robotic', but with suitable adjustments, one or two of the female voices actually don't sound too bad.... I mean, I could listen to them for more than just a few seconds without getting too annoyed... Laughing

Balabolka's next on the list; looks pretty professional, I will say. This is version 2.15.0.701; no idea how 'up-to-date' this is, but the copyright shows 2006-2019. I would guess fairly new.

I have this running under my usual WINE 1.7.51; not a recent one, by any means, but it seems to be all functional. Incidentally, this is the 'portable' version, from PortableApps.com. Click on the 'Apps' tab, and it's the very first item under the first section heading, 'Accessibility'. As mikeslr and I have both come to realize, you don't 'install' these in the ~/.wine directory.....you run them from anywhere but WINE.

Just click on the .exe file inside, and they'll fire up, if they're going to work.

---------------------------------

Nic/Mouldy:-

I have a question. Where do you put additional SAPI 4/5 'voices' if you want to add extra ones? I've obtained the extra American/English voices from the creator of 2nd Speech Center (Zerosoft?); they're .msi files, and I haven't a clue what you're supposed to do with them.

I did find a reference somewhere that said as long as they're somewhere in your system, they'll be found, but I'm still not too sure quite where they would go in the WINE directory structure.

Suggestions?


Mike. Wink

Mike, there are two ways both need WINE running. From WINE menu entry in puppy menu, install via Wine Remove Programs. This method will give you a record of the install and will also enable you to uninstall. The other method is simply double clicking on the install file. Now, you say you have thie MSI files. I don't know how those install in Wine but if you do have problems I have the exe install files for both SAPI4 and The TruVoices so just give a shout and I will upload them. Not sure about the SAPI5 and how compatible that is with Wine but I wouldn't bother with SAPI5 unless you are going to use the commercial voice engines. I suggest you stick with SAPI4 and the TruVoices ( Peter is the best). I've been using this for ages and with a bit of tweaking you can get it to perform at a more than adequate level.

BTW - I released a puplet called tahr-nicOS and with a few extra sfs's. One of them is a Java/Wine combination which includes an older Java and Wine as well as a few windows programmes. 2nd Speech Center is included there and SAPI4 and the Tru Voices are already installed. So that's another option for you, just let me know so I can give further information in this regard.
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Sun 02 Jun 2019, 11:36    Post subject:  

I agree with nic007, Peter is the best of the SAPI 4 voices if you tinker with pitch and speed a bit. IMHO, its as good as lot of the SAPI 5 voices. The female voices with adjustment come close second. The truly neat thing about SAPI 4 is the small size. Oh there are some nice free SAPI 5 voices but not easy to find. The "mobile" SAPI 5 voices built into win10 are maybe the best but you have to do registry hack to free them for general use, then I suspect only in win10, no idea how you would move them since Microsoft binds the voices into the registry. Like say found cool hack to make Cortana voice usable as generic SAPI 5 voice in win10. Microsoft really made the effort to prevent this so not for the faint of heart.

I tried adding engine for SAPI 5 to WINE, but no luck, you cant just add the voices, it needs the engine which is different than one for SAPI 4 voices. But like I mentioned, you install TTSReader 1.30, and it comes with SAPI 4 engine and voices. Balabolka sees those voices and can use them, no Microsoft msi engine/files needed. Thats what makes it so simple. If you can get WINE installed and working, just installing these two freeware programs, you have easy epub to voice.

Now if you did want to install more SAPI 4 voices, and you have a msi file, in WINE you need to run like this "wine msiexec /i whatever-filename.msi" Dont forget that "/i" flag.

Last edited by mouldy on Sun 02 Jun 2019, 12:11; edited 1 time in total
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Sun 02 Jun 2019, 11:52    Post subject:  

Oh, without going to Balabolka website, just from vague memory reading on site, think it was last updated January 2019. So its fresh. Both it and TTSReader say they support SAPI 5, Balabolka strongly recommends SAPI 5, but both assume you are running XP or newer with SAPI 5 engine and at least one SAPI 5 voice installed from getgo. Adding it to WINE I think I read is possible but didnt find clear way to do it.

Oh cant remember name of voice but winVISTA of all things is supposed to have had one of the best SAPI 5 voices built in. I remember it was a female voice. I ran into some old discussion how to move that voice to newer windows. As Microsoft seemed to have canned it. If its possible probably somebody hacked way to do it out there. And maybe a file with it as independent SAPI 5 voice that could be installed elsewhere that has SAPI 5 engine.
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Sun 02 Jun 2019, 12:42    Post subject:  

Ok, name of that voice in VISTA is Microsoft Anna. Apparently she was in win7 too. And there was some way to officially move her to XP too.

Here is a thread that claims to show how to unofficially move her from those systems to win8 and win10. If she can be moved then if you can get SAPI 5 engine in WINE, you could move installed windows voices there too.

https://www.eightforums.com/threads/microsoft-anna-on-windows-8.23916/
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Sun 02 Jun 2019, 13:00    Post subject:  

Oh by way I also installed the "hunspell" dictionary into WINE and Balabolka could use it. Link is on the Balabolka website. Balabolka seems to be able to pronounce somethings better with dictionary access. Outside of pronunciation, not sure how it would be useful to an ebook to voice app.

Saying all this, though WINE would seem the way to go for many wanting simple way to do it, seriously dont overlook native linux programs I mentioned, especially eBook-speaker that uses espeak as plugin. It works, just be really nice if somebody created a decent gui frontend for it. Both ebook-speaker and espeak are installable in Xenialpup64 via Puppy Package manager. they are in the Ubuntu repositories. It was painful enough with apt-get in Lubuntu, so be aware this could require lot dependencies in Puppy.

And if I can ever get it properly compiled, this little Bard app mentioned that uses flite plugin and used by lmemsm looks very promising. there are several voices available for it, comparable to SAPI 4 voices I think. This would seem a very nice program for small distribution like Puppy. There doesnt seem a binary package anywhere, only option is to compile it.
Back to top
View user's profile Send private message 
mouldy


Joined: 04 May 2005
Posts: 664

PostPosted: Sun 02 Jun 2019, 14:28    Post subject:  

This thread might be useful for those wanting SAPI 5 in WINE:

https://nextup.com/forum/viewtopic.php?t=6627

Here is a newer recipe: https://www.reddit.com/r/linux/comments/6z41qb/my_text_to_speech_tts_solution_wine_microsoft/
Back to top
View user's profile Send private message 
Display posts from previous:   Sort by:   
Page 1 of 3 [32 Posts]   Goto page: 1, 2, 3 Next
This forum is locked: you cannot post, reply to, or edit topics.   This topic is locked: you cannot edit posts or make replies. View previous topic :: View next topic
 Forum index » House Training » Users ( For the regulars )
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum


Powered by phpBB © 2001, 2005 phpBB Group
[ Time: 0.9885s ][ Queries: 12 (0.7413s) ][ GZIP on ]