Dear AppleVis community,
AccessMind is proud to announce the first public beta of Runa TTS, a new text-to-speech engine for iOS that brings high quality, natural sounding, neural voices from ReadSpeaker, a global pioneer in the field of human-like TTS technology, to be used with VoiceOver, as well as other applications that use Apple's speech synthesis APIs.
Enjoy classics such as Paul, Kate and Julie from the Neo speech days, or new voices in 50 languages, including English, Arabic, Hebrew, French, German, Spanish, and many others.
The application also includes some advanced features, such as:
- Voice Profiles, allowing you to configure various settings such as speech rate, pitch, and maximum volume per-voice, and forcing them even if VoiceOver is set to different values.
- Automatic language switching based on Unicode characters for Arabic, Latin, Hebrew, Russian, and Greek even if VoiceOver does not support it in your language or region, configured on a per-voice profile level.
- iCloud sync between devices to make sure all voice profiles and voice settings stay up to date between devices.
- In a future update, Siri shortcuts will be introduced for switching voice profiles on the fly.
Once you download and install the application, focus will land on the “manage voices” tab, allowing you to download voices and filter them by language, quality, etc. 2 qualities are available for each voice: standard and enhanced. We strongly recommend using the enhanced quality.
Every voice will be a switch button, when checked, it will be added to the queue, and a button to install all selected voices will appear at the bottom of the screen, before the tab bar. Activating this button will install all selected voices. If voices are installed, checking them will cause a button to remove all selected voices to appear instead.
Once installed, you can add the voices to Voiceover using the VoiceOver speech settings, including configuring them as primary, and adding them to the rotor.
The settings tab, contains all the advanced features, such as Voice profiles, and iCloud syncing. Both are disabled by default, to avoid confusion.
Important notes
- Due to bugs with iOS 27, language switching by Unicode characters will not work properly if configured through our engine, it will only work through VoiceOver with the way Apple has chosen to implement it. We will communicate with Apple to hopefully get this fixed.
- There are no bundled voices with the app. You have to download the voices you wish to use from the manage voices tab.
Feedback during the public beta
Please send all your questions, comments, and queries, to: adi@accessmind.net
Pricing
Pricing for the final product, including subscription or perpetual licensing models, has not been decided yet. The public beta will be available to try for free at least until September.
Installing the public beta
Please use the following TestFlight link to install the public beta of Runa TTS on your devices:
https://testflight.apple.com/join/bJQ5UpWQ
We look forward to hearing from you!
Comments
My feedback about Runa tts
Hello,
I would like to thank you for this amazing tts, works very well with language detection.
I set my iPhone language to Russian, turned-on Unicode language detection, set the Cyrillic characters to Russian and Latin characters to Turkish.
However, there are some letters in the Turkish alphabet in addition to English letters. They are ş, ı, ğ, ö, ç and ü.
When voiceover encounters with these letters, it stops reading, switches the voice to default Turkish voice and continues reading. Please add the letters mentioned above to the Unicode list, so voiceover doesn't pause and switch language when encountering with these letters.
Also, since I have the use Unicode characters enabled, when I add English language and select that voice from the rotor, voiceover speaks in Turkish anyway, i.e. it continues to use the Latin language instead of switching to English.
If there was a quick way, like a touch gesture to switch between profiles, I'd switch to the default profile and read the English text that way.
Other than these problems, everything works fine. It has enabled me to use my phone in Russian, which is the way that I would like to use my phone and Turkish at the same time.
Thanks, and congrats.
Individual voices
Due to the way in which the licensing works, and for other reasons, individual voices for purchase most likely won't be an option. I am not saying no or never, but the primary option would be one price for all. The price will be reasonable.
very excited for these voicrs
the fact that these older voices have been ported to iOS is amazing! using Paul now, with Beth and Arron on my voices roter as well. I am ready to pay any price when the time comes. great work!
@ Will James
The developers own words are causing the reactions they are getting. Do I expect things like a battery drain absolutely it is beta 1. Do I expect to be gouged for every operating system I have absolutely not. Their own comments prove that is what they are going to do.
"We might offer MacOS versions at some point later if we see that the product takes off well".
In otherwords if we make enough off it we'll do it otherwise we won't.
The dev was the one who said it would be one price for all voices. The reaction is what they deserve. They are doing it too themselves.
Bug report
Just thought I'd pop in to say I've encountered a bug. When reading very long strings of text--as in several paragraphs--VoiceOver will pause, then continue on. The pause in and of itself is somewhat irritating, but that alone would be fine. However, I think in this pause VoiceOver stops sensing that it's reading a text, because in the midst of reading the rest my screen will dim, which it otherwise never does when narrating large information chunks. I hope this makes sense and is helpful. Happy to answer more questions as needed.
So, what's wrong with developers making money?
All of the time, I find posts on here; with people complaining that an application costs money. Well, why do you think software developers and software engineers get into this line of work?
ProTip, to make a little profit.
This is Beta 1, and some of you are whining like a bunch of spoiled monkeys.
Regarding the comment about only porting this to macOS if the iOS application is successful; why does that make so many of you mad???
I would not want to make a multi-platform application, if the first itteration which happened to be native to one particular OS, was a failure.
Come on folks, use your brains just a little bit... please?
Here's an idea, how about you folks who are panic-driven about poor quality, lack of variety, or pricing maybe give the developer time? Time to perfect his product. Time to, well, develop?
Just a thought. 🤨
Pricing not equal to beta 1
The pricing is nothing to do with beta 1, oh trust me, I know developers want to make money, I've made money out of software I've made for people specifically, nobody is doing that for free, I can actually agree with the macOS part. Also I have 0 exact issue with the pricing model, I stated that it wouldn't be for me as I'd only use 1 voice and so wouldn't need the rest of them. It's up to a developer how they price a product, and they don't have to agree or disagree, but if there was no feedback, there'd be no conversation.
Repluying to the "Bug Report" comment
I've encountered the same thing when using VO with very long streams of text. I.e. facebook posts. This happened wen using the Kate enhansed voice with a facebook post. This is primarily wy I have these voices on the rotor and use Nikki enhansed as my primary voice. I will continue to do this until this bug is resolved.
This is why we can't have nice things
Brian, thank you for saying this! Dennis, don't fuck this up for the rest of us who want to reward the developer with our hard-earned cash. I get what you say, but have a little fucking tactfulness, please! If you have a Mac, you have a dog in that fight. If you don't, stand back and see what happens.
To Those Comlaining
At least Apple users have the option. Those of us on Android will probably never see these voices in this modern age. We had them once from another company, but they didn’t take off due to the early days and such. Be glad you get to test them for free in the first place!
Completely agree with Brian on
Completely agree with Brian on this. You guys are getting these boys at the test out for free and anybody can just complain about it. That’s going to accomplish absolutely nothing other than to make developers not even bothered to want to develop for you guys. Not cool. It’s very childish behavior to complain about things. This is literally the first batter. There’s always gonna be issue with the first beta that is how it is. Start giving the Church constructive feedback so I’m complaining about everything you don’t have. Start giving mature feedback.
Agreed 100% @wiljames
I agree with your point and understand your perspective. However, I believe the use of profanity is not permitted on this website.
To add to your argument, I concur that we don’t deserve better products. Some individuals will not appreciate that we also require different features. For instance, certain users do not warrant new avenues to utilise a device.
I would purchase this product immediately because the voices are far more pleasant and superior in quality to the vocaliser text to Speach which is excessively robotic for me.
I genuinely hope we can engage in a constructive dialogue in this forum rather than complaining about pricing and other matters. Developers also need to earn a living and I am prepared to pay for this app in particular.
If the individuals do not have constructive criticism to offer, please refrain from posting in this forum as it will be detrimental to the developer who is creating this product and to users like myself who desire a better text-to-speech software.
cannot find the voices
In the add additional voices to roter list, I can't find this tts engine.
Beta 2 is out
Hi folks. I have updated the public beta version with improved battery performance, a new home screen widgit for switching voice profiles on the fly, and some small fixes.
Reacting to feedback
Hi folks,
I have been reading all the feedback provided here for the last several days, and I want to respond to as much as I can of it:
First, I would like to start by thanking everyone who has been providing the feedback, both positive, and negative. Every constructive feedback is highly welcome.
Let me address the elephant in the room first, the pricing concern and having licenses for individual operating systems:
1. Pricing:
As you might know, and I would like to make this clear in case it is not yet, ReadSpeaker is not an engine we develop ourselves. ReadSpeaker is a commercial company, from which we license the voices, and bring them to be used with the platforms all of us need them to be on, or want them on if possible. So what we develop is the actual application itself that integrates the engine and makes it available to us, but not the engine itself. Therefore, when we price the product, we need to take into account the cost we have to pay ReadSpeaker for each copy we sell, and also make profit for us so that we can make a living, since at the end we have not released this as a free product. The goal here is to keep the pricing reasonable for what we provide, however, due to the licensing model we have with ReadSpeaker, and various other consideration's, we cannot offer an individual voice purchasing option. The price will be somewhere in the 20 to 30-35 dollar range for all the voices at once. Our goal is to always make the product as affordable as possible. I am a blind person myself, all of the developers working on this project are blind. We totally understand the various financial situations all of us experience. We are also a part of this. So we have to balance between making the product affordable, and us making a living, while paying ReadSpeaker. Keep in mind that our market is relatively small as well, and if we take this into account, the licensing model ReadSpeaker offers us is very generous.
2. Individual operating system licenses:
ReadSpeaker has a licensing model that requires having different versions for each operating system, when it comes to the engine license. This means for us, that we have to sell individual versions for each operating system we want to offer the product for. Yes, this also means iOS / macOS even though theoretically in the Apple world one app can work on both, it is not how it works here in terms of the license. So if we decide to release for macOS, we will have to charge an individual fee for it.
At this point, we are going to release versions for iOS, Android, and Windows. If we see that there is a demand for a macOS version, we will gladly do that. But it will require an additional purchase in this case. If we see that enough people want it, and will be willing to pay for it, nothing prevents us from doing it. But we cannot sell 1 version for an affordable price that covers both platforms.
3. Long text reading:
Thanks for the reports about long text reading. We can reliably reproduce the issue of the screen dimming, and we will fix it.
I really hope that this makes the things clear for all of us,and please continue sending feedback.
a suggestion for pricing
Since you have to have people pay for each system for the second system on have you thought about a discounted price? So for Mac OS TV OS and Watch OS as examples if someone purchased the iOS version first maybe discount the price for them on the other systems. I think $20 is a reasonable price.
Re: Windows
Make this work with NVDA and you will have a very satisfied consumer. 😇
@RDKushnir
The second beta is functioning well. I have identified a potential issue where the Bridget Voice does not lower its pitch when capitalising letters or entering any other characters. This is not a significant concern but I wanted to bring it to your attention.
I am also pleased that this is being released for Windows as I intend to use it with Jaws.
Discounted pricing for operating systems
Denis, thanks for razing this. Specifically for Apple operating systems, we willl see if we can offer such option, we are thinking about it. It will be difficult, but we will try. For Windows for example, a discount won't be offerd if purchased for Apple. For Apple, it is going to be iOS/macOS only, but macOS will come only after start of sales for iOS and if we see that there is a demand.
Thanks for this TTs
I’d buy this TTS app just because I can’t stand the native iOS French voices anymore. They haven’t changed them in ages. I’m hoping there will be more French voices, especially female ones, but either way, the Roxane voice I already use is reason enough for me. If it’s not more than €40, that feels reasonable.
Voice Dream Reader
I have the Brazilian Portuguese voice installed on my rotor, and I wanted to use it to read a book in Voice Dream Reader, but the voice isn’t showing up there.
Moderators' Reminder
We realize this topic is generating a lot of strong and passionate reactions. A lively discussion is good, whether you agree with others' opinions or not. With that said, we ask that everyone please keep the following in mind:
@Adi: Thanks and Some Questions
Hi Adi, fantastic explanations of the factors that are going into the product development. Thank you for the insight.
In case anyone is reading my comment and has not seen Adi's reply, here is a direct link: https://www.applevis.com/comment/212984#comment-212984
I personally think US$20-35 is fair given the licensing plus needing the product to actually bring in a profit. But of course I realize that not everyone will agree, and that is okay.
Looking ahead, if this goes well, could your company potentially develop with other engines besides ReadSpeaker? I would love to see someone bring classic DECtalk to iOS, for example. I keep hoping Apple will do it, but so far that hasn't happened and given the results we are seeing from your ReadSpeaker implementation even just in this beta, I could only imagine a classic DECtalk release from your team.
Pricing
$20 for a suite of voices sounds reasonable. I also have yet to play around with beta two but I did update it on my phone via TestFlight. Thank you for all you do for developing this app blind/visually impaired users in mind.
@michael: other engines
Hi Michael,
Thanks very much for this comment and request. We can of course validate all possibilities and think about developing additional engines if there is a demand for them. Dectalk is a good example, I wanted to have Dectalk for iOS myself. It would be something worth looking at. If you have other ideas, commercial or open source, please write them down.
is something I could get behindDectalk
Dectalk is very possibly something I could get behind. Had you stated up front why you needed to charge for different platforms, you wouldn’t have gotten
the reaction you got. I was basing it on other apps. They don’t charge for every system. What did it look like? A money grab. Yes developers deserve to
be paid. I’m not disputing that. I subscribe to apps. I won’t pay for an app if a dev is doing a money grab. I don’t have to charge for Mac; I’m doing
so because I can. Clearly this is not what is going on. That should’ve been made clear up front when you first talked about pricing.
Some Historical Context
I deeply appreciate this app and the quality of the voices. I would gladly purchase Dectalk for use with Voiceover.
Between 2005 and 2010, I purchased Dectalk Access 32 software from GW Micro for use with Jaws. I spent approximately $75 USD. The Dectalk Express cost $1,500 Canadian.
I hope the developer is successful, and
Dectalk
I would absolutely love to have Dectalk on iOS!
Thank you for such an excellent app!
I use English, Japanese and Italian ordinarily, and all voices are working perfectly.
I'm willing to pay a lot for this app; can't wait for the appStore release.
Licensing
Hi,
Thanks for your explanation of how the licenses work, that on my part may have been a slight oversight, which I have no issue admitting.
I'm currently using the Adam voice, but at the moment find myself going between that and the standard Daniel voice on iOS, which I've used for who knows how long.
Coming at it with the updated info, and knowing about it a little more now, I'd say the pricing structure there is fair enough.
I think what I'll do, is use it as a beta, see if I find myself using it more, and then make the decision on release.
I've switched back while reading this thread again now, so maybe I'll give it a full day or 2 of usage without switching back and see how that goes.
The voices are very responsive though, which can't be said for every TTS that's been released to iOS in the past, which is a very very good sign.
And finally, a note, the feedback was never intended to be critical of the product, apparently feedback that people don't agree with is instantly critical, but I was coming at it from only my angle, everyone is free to do as they wish and use the voices or not use them. But putting feedback about the price has its advantages, that full explanation was useful in understanding why and how its going to work going forward, so thank you for that
Small issue
I've noticed a small pronunciation issue when using the UK English or Australian English voices. On this site, the word, "AppleVis", is pronounced as, "Apple Vista". This does not happen with the US English voices.
I have made adjustments with the pronunciation in VoiceOver settings, even setting the parameters for, All, languages, but this seems to be ignored.
Just wanted to point this out, otherwise great work so far. 😎👍
Love the voices
Greetings,
I am really liking these voices. These will work great for reading bookshare books in apps like bookshare reader and speech central. Right now I use voice dream reader because I can use neospeech james. I hope james will be available in Runa TTS eventually. If not Paul is pretty good.
apple vista
This is happening to me as well, using US English Erin
apple vista
This is happening to me as well, using US English Erin
No pitch raise on capital letters
No matter which voice I use, there is no raising of pitch on capital letters. Is this a bug or a limitation? Otherwise, public beta 2 is working great! I have noticed the battery not draining as hard as it did in Beta 1.
Keep up the great work Adi!
Welsh voice disappeared
Hi, I was using the Welsh voice, and I upgraded to Beta too. It was in my rotor, it disappeared. I tried to delete it out of my rotor, and then re-add it but Welsh wasn’t even an option. I then tried to delete the Welsh voice in the app, and reinstall it but Welsh still is not showing up. Any idea what might be causing that?
Re: No pitch raise on capital letters
A couple of things might be causing this.
• In Settings, Accessibility, VoiceOver, Verbosity, the option for capital letters may have been switched, to something other than change pitch. Double check this.
• In Settings, Accessibility, VoiceOver, Speech, make sure the option for changing pitch is checked.
HTH.
Dutch voice "Veerle" not working
RE: Dutch (Belgium)
Thanks for this report. We are working on a fix.
Improved latency
Hi folks. We are currently testing a new version internally, in which I was able to improve the delay for long text and latency in general by a big amount, so it is now really close to Vocalizer. I will release it hopefully next week, and after that, I will start working on fixing issues with character pronunciations, etc, for international voices.
French France voices
Regarding the French (France) voices and the pronunciation of the letters on the keyboard, please note that Y is pronounced like “I” rather than “Y,” while I is pronounced like the English letter “I,” not the French “I.”
When are the NVDA Add-on and Windows SAPI5 versions coming?
Hello, I am the Turkish Translator of tiflotecniaVoices Add-on. (I remember Rui with gratitude.) When will the advanced version of this Add-on with the new name and the SAPI 5 versions arrive? Also, will the same translation files be used for the plugin? Thanks.
@umut: private contact
Hello,
Please contact me privately via email and I will explain. The Tiflotecnia voices add-on will continue having Vocalizer only, but our own new product is coming with the ReadSpeaker voices and SAPI5/JAWS support. Please reach out via email and lets connect.
Replying to Adi Kushnir Improved Latency
Hi. Thank you for resolving this. I will start testing more indepth using various voices; Kate, Erin Julie. From what I've seen with beta 2 it works wel.
Thank you again :)
I sent you an e-mail.
Thank you for your quick reply. I sent you a private e-mail.
English voices and 1 of my very busy WhatsApp groups
Good day to all of you
Congratulations creating such an excellent App. Paul, Kate and Sophie voices my favorites. Only one problem. I have 2 very busy WhatsApp groups and after opening 1 of these groups, none of these voices reads the name of the recipiant or message content. I tried this also with Australian, UK and the South-African voices. Same result. Strange, but the problem only occurs in 1 of the groups. No problem though using Tessa or Eloquence.
@blackcat:
Please give the actual text and send feedback privately.
I love those voices
Hello Adi and all, I love it and use it with VoiceOver.
I love the Yael Hebrew voice—she's very clear and responsive.
Regarding the price: I think $20 to $40 makes sense.
In Israel, TTS costs more than that.
AI is amazing, and I'll be happy to see how it improves.
Beta 3 is out
Beta 3 introduces some big changes, one critical for everyone and the rest primarily for advanced users:
1. Latency improvements - we have worked hard on improving the latency a specially for long texts. Please test and let us know what you think.
2. Voice profiles - You can now make each voice profile show in the voices roater, by activating the new Show in voices rotor option when editing/creating a profile. When activated, you can add each voice profile to the VoiceOver rotor, for every locale available in the system. This has been done to overcome Apple's language switching limitations in iOS 26 or lower, iOS 27 has been reported to Apple. We have also renamed automatic language switching to Smart Switch.
3. Default voice - you can now specify a default voice for each profile. If smart switching is disabled for that profile, it will use this default voice all the time.
4. It is now possible to set the default voice profile voice to be used for every script/numbers/punctuation setting of the smart switch feature.
5. It is not possible to set a profile as active anymore since they are only active when chosen as rotor voices.
6. The voice profiles widget has been removed since there is no need for it.
Re Beta 3
Hello.
I have just sent feedback with a possible bug. I may have shot myself in the foot though lol as I have just read the release notes, and the bug I described is actually by design to overcome Apple's limitations, per release notes. Still not sure if the sample should greet you with "hello, my name is" followed by your voice profile, mine being default using the Erin enhanced voice.
Overall seems a great update once you are used to this new capability.