my feedback to apple about speach synthisizers

By jay, 18 March, 2013

Forum

iOS and iPadOS

Hello. I wrote some feedback to apple reguarding new ideas for speach synthisizers for the iphone. Please read this, and let me know if sounds like some good ideas. to apple feedback Hello. I was wondering if its possible to add a few speach synthisizers to voice over in the next ios release. The first speach synthisizer I would like to bring to your attention is Alex for the mac. I have used a mac, and really love the Alex voice that comes standard with voice over for the mac. It is verry easy to understand, and it works well with voice over. The other speach synthisizer I would like to mention is E T I Eliquence. E T I Eliquence is used on many windows screen readers, such as Jaws for windows, System Access, Window-eyes, andit can be used with an open source screen reader called NVDA. ETI Eliquence is easy to understand, and should work well with voice over. Also, it is my understanding that you have the licences to the vocalizer voices from nuance comunications. Adding some more of the voices like Tom, and some of the other voices would be great. Thanks for reading this, and have a good day. Please let me know what you think. Thanks. Jay

Options

Comments

This whole topic is very subjective

Hi! The idea of suggesting other voices for i-devices is not a bad one, but which voices should be added is a matter of personal opinion. Personally, I have no wish to have Eloquence on my iPhone, that being a speech synthesiser I have never personally liked, but I know a lot of people like it, so if it was a choice and not a default I'd just avoid it like I already avoid Samantha whom I don't like. I personally wish the Australian IOS voice was Lee and not Karen: I like Karen's voice, but I find Lee more expressive. Also, although I'm not a Mac user myself, I've heard Alex on other people's Macs, and I agree, he is easy to understand. I'm just giving my personal opinion here, without bashing people's preferences for other voices even if I don't like them myself, and I hope others who post to this thread will do the same in a civilised way.

A few considerations here.

Hi. first off, well done Claire. Second, the reason Alex isn't available and will not be, is because of the size. Remember you're carrying a portable device that might be 32gb, but really be more like twenty seven, because of the IOS needed to run. add on to that Alex, who is I think a gb, and you are taking more of your own perceived storage space away. I also dislike Elloquence, but again if you like it, that's certainly fine. I don't care for sam that much either, though I admit she's a little easier on the ears as it were, when she's not in compact mode. I think this really is asking apple to become more of a screen reading company then just adding on a built in, feature as voice over is. again my thoughts alone. Hope others also understand where I'm coming from.

possibility to have more then one hq voice

It should also be great that they add the ability to have more high quality voices and mauybe also the activity thingie we have already on the mac. And when we are speaking about voices, they should really please fix some nuance misspronunciation of some tts voices in some languages like dutch etc. just my 2 cents. :)

that's sound's great

I love that idea for adding new synthesizers to voiceover. I would love to see more synthesizers in voiceover.

My thoughts

Nuance owns the rights to Eloquence, so Apple being able to use it is in theory easy, but they would have to rewrite it for iOS from scratch as I'm pretty sure there's no iOS SDK for eloquence since sighted people don't really use it. I'd love to see the ability for having multiple premium voices, and being able to choose between genders for some of them. Note that if you jailbroke you can replace the voices and get around apple's limitation. Lastly, there are some other TTS's apple could use, Loquendo and Svox. Both have superior voices in other languages (mainly Polish for Loquendo, and svox has an amazing UK voice that you can use in the voice reader app). Both Loquendo and Svox also have been purchased by Nuance and already have iOS SDK's so apple should be able to integrate them easily.

a few more points

Its interesting that I see this conversation now, because earlier yesterday I was watching youtube videos of comparisons between Google Now and Siri, and these were done by sighted people. A lot of the reason why Siri lost these was because of the speech synthisiser. Google Now has a much less robotic sounding voice than siri, and even the mainstream community has noticed that. Its probably not a big deal to Apple, but if people judge Siri, an application that has little to do with the speech synthisiser in the device could be a problem. So basically I'm just trying to say that its not just voice over users that use the speech synthisier and sighted people as well. Another big thing is that most sighted people I know hate Elloquence, I doubt Apple would ever use that speech synthisiser.

love eliquence

I too would like to use eloquence on my I devices. It is good for reading a text book or reading a good novel.

My thoughs

Hi, I don't really care all that much for eloquence. I prefer something more human sounding like Vocalizer. Or even The high quality SVox voices. Or Acapela's voices. I use Acapela in VoiceDream all the time.

Samantha

Hey guys, I don't mind Samantha that much, but I wish she'd have taken the cough drop, gum, candy, or whatever she's sucking on out of her mouth before she began her recordings. So is anybody in on this? Who els thinks Sam has something in her mouth? Sometimes it actually get in the way of the esses and zees. Listening to her reminds me of listening to books recorded by Learning Ally, where many readers do the same thing. It's irritating. I wish they'd switch her out for Tom. As for Elocquence, the only thing i liked about it was it's extremely fast response time. The VO voices we have now all have good response times though, so I wouldn't even ask for that one.

Ivona text to speech

Greetings to whole community. The idea of other speech synthesizer is on my mind since long time. Do you think that Apple could consider cooperation with Ivona, for example, to include their TTS in IOS?

How do you replace the voices

How do you replace the voices on the iPhone so you can use the tom voice for voiceover?

Implementing any of the

Implementing any of the suggestions listed here, from alex to eloquence, and particularly the Ivona or Acapela voices would be an improvement, particularly if we could choose between them. The inclusion of Ivona voices is one of the only reasons I am looking into the new touchscreen kindle line when they launch, as Amazon recently purchased Ivona for use on the kindles.

i can agree that we can

i can agree that we can better enjoy vo with more tts voices, giving that i also wrote to apple on behalf of this topic. yet, i agree that adding new voices will most likely use up gb. & then again, atleast we have vo to begin with. at least a company thinks to give a crap about their costomers' satisfaction. at least wwwwee can be more mainstream thanks 2 what we already get from apple. but 2 be less of a party pooper, i have a mac & i agree that alex is a great voice my favorite & sam does sound like she has something in her mouth. lol

Another TTS and what I prefer

Hi all, another suggestion that hasn't been mentioned yet is the Espeak TTS that comes bundled with NVDA. For me personally I would rather have Elquance 1st followed by Espeak. Please let me explain why. The main reason is as well as being blind I am also hearing inpared and interestingly it seems I like robotic voices over natural ones. With elquance and Espeak I can speed it up a bit and still be able to understand what its saying. The other few things why I like these TTS's is the clarity, pronunciation of words and for me it doesn't get tiring when I'm working with these voices. Hearing is also a factor that needs to be considered as well as blindness. I am in no way bashing any company but am just stating a possible fact. Two more things before I rap up my say. 1st a suggestion. Having a place where one can download there voice of choice may be a good way where we can access the voice we want as indeviduals. (like what they did in Lion) Applying that methiod to IOS and adding the additional voices to OSX/IOS should or when they come to be. Finally I think in this case for voiceover users without this meaning to sound arrogant but its up for us as a body to inform Apple what we want or possably need for TTS's. I mean we are the ones that have got to use it all day right? I find this discussion most fascinating and will contribute should it be appropriate. :)

I agree

I agreeall the feedback that we are giving here needs to all go to apple so they can take action.

Add Voice Download / Delete Capabilities.

Hi, The solution is quite simple really, instead of including more voices with iOS out of the box, Apple should give users the option to download any additional voices they might like and delete the ones that they are never going to use, just like on the Mac. The size of each voice should be displayed beside each voice name so that users can be aware of how much space they are using for voices on their devices. Regards!

Maybe th next Applevis Campain?

Agree with what Robert A.M says. Also, I don't know if it has been a topic that has been mentioned in the past but maybe it could be the next Applevis Campain. I don't really know how that works, but its a place where one and all have there say. Just a thought though.

It all comes down to the code

The vocalizer voices could be made different, such as in lion, with about an hour of someone's time. The problem with other engines is that they are not as easy. Yes, acapela already has an SDK for IOS, but this is directed at apps, not the internal workings of the operating system. It also doesn't work anything like the vocalizer system. Same with the others. Eloquence was written by different people who don't currently work at nuance, in an older programming language for a different platform a decade and a half ago, not to mention that as pleasant as it might be, it sounds like a synthesizer produced with technology fifteen years old. It is not what you expect to hear from a device that is so new and flashy like the apple devices. Also, it supports seven languages. Vocalizer has those, plus many more in western Europe, plus Eastern Europe and Asia. Acapela, as nice as their voices are, doesn't have the capability in Eastern European languages yet, and only introduced Japanese and Chinese voices about a year ago. They expect Korean soon, but no Hindi, no Indonesian, no Thai. Apple is better off, at least for now, keeping vocalizer because of its international support and because it doesn't require too much upkeep. I do agree about the whole voice plethora being made available, however.

Sending the request to apple

Sending the request to apple for the ability to pick and choose and change VO voices is a great one. While it is also absolutely true that we are and should be grateful for all the work of apple for including voiceOver at all, the ability to pick and choose voices would be a highly useful one. Does anyone want to or know how to organize a petition or email form to subscribe to send to apple

My opinion on alternate voices

I've been an Android user for years and just switched to the iPhone about 2 weeks ago. The only thing I truly miss about Android is the option to switch voices. My preference was the Acapela voices. It would be amazing to have some other options for VoiceOver.

yes we should send a request

Yes we should request for apple to add this feature in the next major iOS update.

What are some of the voices that Android supports?

Dear Brook, you mentioned Android and I have never used that OS. I was just wondering what other voices does Android have over IOS? Can you use Espeak does it have any robotic ones out there? Look forward to hearing from you.

Voices on Android

Hey Daniel, Android has several options when it comes to TTS voices. My favorite, as I mentioned, was Acapela. There are countless voices you can pick from (at one point, I had 6 downloaded and kept rotating them to find my favorite). Another TTS option is SVox, which has several voices you can pick from (my favorite was Victoria, a UK voice). I didn't like them as much as Acapela, but they were okay. Yet another option are the Ivona voices; their US voice (Ivona Kendra) worked well and was pretty responsive. You can also use ESpeak, although I couldn't deal with that voice and so uninstalled it in a matter of minutes. There's also Google's TTS, and, if someone is running an older version of Android, they could use Loquendo Susan. But for some reason, Loquendo was never updated past Android 2.3, so that's not an option with most newer phones. And lastly, if Mobile Accessibility is on the phone, its primary voice is the same as the US voice in the iPhone. But you can also change this to the system default voice so you can use whichever voice you prefer. I'd say the only robotic-sounding voice is ESpeak; and I wasn't a huge fan of Google's TTS either, especially when there are so many other options out there. Sorry for the long post, but I hope that helped answer your questions!

very helpful

Hi Brook, by no means your post wasn't long. In fact it was very informative! Quite a range of choices on the tabal for voices on that platform. My final question to you is in your use of Android voices, what are the male voices that you have used. If at all any. Don't worry if you have not used male voices but I just wanted to know what's out there in male voices for the Android system.

The voice I chose for Android

Actually, the voice I chose as my primary voice was a male voice, Acapela Will. I tried Acapela Ryan but found Will easier to listen to. A voice that seemed popular and that Acapela just added to Android is Graham; I haven't tried that one, but it was a voice that seemed to be requested often.

Re: A Few Considerations Here

I certainly agree about Alex. I just got my Mac Book Air last week and really like him. I wonder how he is able to breathe. But I also like the other voices that are installed on my machine. I like Eloquence too, but another possibility might be eSpeak. I realize a lot of people don't like it that much, but it is open-source and from what I've read it is also rather small. I like eSpeak, but personal preference is what it boils down to. BTW, please forgive any possible spelling mistakes here. The spelling correction thing keeps popping up, and I don't yet know how to turn that off. I will try to edit this post.

Apple won't be able to add

Apple won't be able to add Espeak into iOS themselves because of its license. Espeak is under the GPL, which in short lets you distribute the program, charge for it even, but only if the source is freely available. That's the first problem with GPL. The second is that if you want to use another GPL library in a program, it also has to be GPL. While we're on the topic of Alex and TTS sizes, with every new OS X release this is becoming less and less of a problem. As I recall, now alex is about 3 or 400 MB. You might think that's a lot, until you consider that the premium Vocalizer voices are also pretty big, 270MB for Samantha for example, add another 200 for the 2 Siri voices, and more for any premium voices you download yourself. I guess Apple just doesn't want to port the Macintalk synthesiser over to iOS because they already have one that's perfectly good and works.

So how can you tell Alex's file size?

Hi, how did you find out the file size of Alex then? I would like to know how to find out this information. Thanks Daniel

Go to

Go to system/library/speech/voices. Find alex.speechvoice, then press space to open quick look, which should give you the file size.

my 2 sents

while a noble idea here, i would be very surprised if you have any success with this. apple's business model is tell the customer what they can and can't do within the IOS ecosystem. they implement heavy restrictions on just about everything within the IOS operating system. hence a whole booming community for jailbraking to get around those limitations and restrictions. if any apple person takes time to investigate this they will probably think, "they already have a voice for lots of languages. Why do they want more? all the ones they have work fine." one huge plus for android and windows is the fact that you can change your synthesizer. from something like eliquence, to espeak, or to an ivona voice. while apple has loosened their stranglehold on the user a little bit in IOS 7, its still rather tight in other areas. i am in full support of this project, and i would love for this to succeed, but i highly doubt if it does. sorry for being kind of a negative nancy on the subject, but its the unfortunate reality of the situation your in while using the IOS operating system. if you want the customization go to an operating system like android.

Things I don't like about "real sounding' voices.

Hi. I don't like that human or real sounding voices, have no excitement to there voices. For example; Oh what a cold day outside, I better wrap up or I'm going to freeze! A Human sounding voice, doesn't read the last words as freeze! it will read it as freeze. So basicly no inflection. That's just what I think and sorry if I offend anyone.

A Humorous Suggestion

Amid all the arguments about synth preferences and discussion of why this or that synth is not likely to be implemented in the iOS world, I have a tongue-in-cheek suggestion: all this discussion misses the mark! Apple should develop a Lauren Bacall synthesizer! Talk about one's phone making one, er, salivate! :-) :-)

Apple's point?

I think apple's point is to make their devices as accessible as they can with the best possible voices they can license. This makes it easier to control what the voices can do over a long time. It makes no sense to develop a single killer voice that can only do a few things. In the accessibility realm of software and services development, novelty voices have no place.

I agree. I want to get my

I agree. I want to get my stuff done I don't want to play with voices. hence why I removed the novelty voices from my computer. I have no use fo them. My computer is for me to do actual work not to have fun and play with.

Have to Disagree Slightly

Hi. I'm revisiting this topic now that I've had my Mac for almost a year. I sort of disagree about the novelty voices. While some of them can get a bit annoying at times, I am one of those people who likes to switch to them a lot. So is my sister. Perhaps Apple meant for those voices to be used mainly when gaming? We both tend to switch back to a more normal-sounding voice whenever a sighted person uses our computers for whatever reason. For instance, my sister likes to switch back to Alex whenever she is with her life-skills tutor and they are at a training session at the Apple store. I do agree though that a lot of the more human-sounding voices tend to lack expression. Not all of them though. But like I mentioned in my first reply here, personal preference is what it boils down to. If you can still get work done while listening to the Good News voice, for instance, then you have every right to stick with that one for however long you want.

Apple response

Sorry to revive this thread, but I am extremely curious. Now over a year had past. And we know alex is available in newer devices.

To those who had written to apple, do you guys got any response?

It's not gonna happen

It's not gonna happen. It's just not practical to include every single voice that everyone likes on all these devices.

re: It's not gonna happen

But at least the voices that many (most?) people want.

Storage

I think it boils down to storage. It's fine to have a few 500 mb voices on a computer but on an IOS device where storage is limited and has a huge premium for the larger capacities... They're just trying to squeeze the optimum into this package, unfortunately, from this thread I gather that 'optimum' is very subjective indeed.

Alex is already here

Alex is already in IOS 9 for the IOS devices. It was brought on when uiOS 8 came out.

To Wayne

See the above posts. Yes it was brought on in iOS8. This is about bringing it to older devices.