Accessibility Issue on Mac about voices on Spoken Content

This is a bug in Big Sur. Everyone who is interested in having this issue fixed, please call Apple's accessibility support and open a case on this. This doesn't just affect Chinese voices. This affects any other languages too. For me, the problem is with the male default system voice matching my default language (English US) is used when pronouncing some French words even though I've selected a female French voice as my System Voice. This happens with some words that the broken algorithm "deems" worthy to be pronounced in English. With other words, the correct female French voice is invoked.

I can no longer use the Voiceover feature in macOS for learning foreign languages. I've opened a case with the Apple accessibility department, which was sent to engineering. The reply from engineering is that this feature is working as designed. However, this feature worked differently in all previous macOS versions; in fact, it worked correctly before, when the selected System Voice was always used when the system voice was invoked for selected words. In my example, regardless of what the word looked like, the female French system voice was invoked in previous macOS versions, so even if I were to type up a phrase in English, it would be read in a French way (funny French accent). That was the desired behavior. Now with Big Sur, this Voiceover feature invoked for selected words is a complete and utter mess. It can't be working as it was designed. This is a bug.

I've posted on several help topic pages and I opened a ticket with the Accessibility department but my problem wasn't resolved with a satisfactory solution; just work arounds. I'll say it again here.

System Preferences>Accessibility>Spoken content>Speak selection>"options" doesn't have a way to control which voice is used. The computer randomly chooses the "speak selection" voice. I didn't get a single one of the random voices I wanted. (I know how to add and remove system voices and the computer uses a voice I don't even have selected in the "system voice" window sometimes.) "Speak announcements" has a selection tool under "options" and it works with the selected voice. What is odd is that both "speak items under the pointer" and "speak typing feedback" don't have a way to pick the voice used under "options" but they both use the same voice that is selected on the "system voice" drop down menu under "Spoken content". For some reason "speak selection" doesn't register the chosen voice under the "system voice" drop-down menu.

This is an issue with a 2014 macbook pro running big sur 11.1. Troubleshooting included a clean install of big sur 11.1, a safe boot, and a NVRAM reset. When I got connected to the Accessibility department they said that changing "speak selection" voice wasn't an option and they gave me a few helpful articles for a work around that I may try but I don't really want to use the VoiceOver command or the Voice Control commands because VoiceOver is too cumbersome and you can't customize what is said and Voice Control makes me speak commands and I prefer to be mute when I'm on my computer and listen instead of speak. Plus, if you throw VoiceOver and Voice Control together you are going to overload you computing capabilities and be unable to operate anything that is processor intensive. Also there are sure to be cross talk issues and I don't want a headset.

The staff that I remember names of were Aubrey and Nathanael and they were both excellent resources to work on this problem. They were polite and quick, I wasn't aware of the 24/7 chat support until now and I've got to say; at least apple directs you efficiently with their tech support unlike other manufacturers that have horrible support wait times and worthless tech communities.

But all that being said I would like to voice my disappointment that my solution will likely just bog down my computing capabilities and I would like to urge the engineers to make "Spoken content" fully functional such that the community is satisfied with the updated product. I just found out about this feature but there are some other not so satisfied users out there.

Currently I'm running the "Spoken content" option by using "speak items under the pointer" and "speak announcements" because it doesn't overload my mac and I can run streaming services while working on my art projects. But when I want to have the computer read a long text to speech entry "speak items under the pointer" fails to consistently read long passages. I would like a better long passages solution than VoiceOver because like I said, it just kills the ability to multitask on a laptop.

*Drops mic*

Hey there midnameblank,

We understand that the System Voice you are selecting in System Preferences is different when you try to have text spoken outside of the accessibility settings. We’d like to help however we can. 

To confirm, which voice are you selecting under ‘System Voice’ in the ‘Spoken Content’ section? Does it show that it is downloading the voice still? 

Change the voice your Mac uses to speak text

If this continues after the voice you chose is fully downloaded, then we would recommend reaching out to our Accessibility Support to get the best assistance going forward.

Cheers!

So, on this Christmas Day of 2020, I was able to find a workaround! Enjoy this Christmas present from me to you.

In most native maOS applications in macOS Big Sur, there are two menu items:

Edit > Speech > Start Speaking

Edit > Speech > Stop Speaking

We can leverage these menu items to get macOS Big Sur to pronounce the words we select in that application with the System Voice that we select in System Preferences > Accessibility > Spoken Content: Speak Selection > Options.

Therefore:

• If you use the keyboard shortcut that assigned in System Preferences > Accessibility > Spoken Content: Speak Selection > Options (default is Option + Esc), the behavior or the System Voice that you invoke to read the selected text is inconsistent and often the wrong language/system voice is used in Big Sur.
• If you use the menu item located in your application > Edit > Speech > Start Speaking, then the System Voice that you invoke to read the selected text is always consistent and the correct language/system voice is used in Big Sur. The System Voice invoke in this way is the System Voice that you have selected in System Preferences > Accessibility > Spoken Content: Speak Selection > Options.

It's possible to assign a custom shortcut to the menu Start Speaking located in your application Edit > Speech > Start Speaking.

This is how it's done:

1. Navigate to System Preferences > Keyboard > Shortcuts > App Shortcuts (last on the list of types of shortcuts)
2. Click on the "+" sign at the bottom of the right-hand pane of the window
3. In the new dialog window that opens:

• under Application: select All Applications
• under Menu Title: type in Start Speaking
• place the mouse pointer in the field next to Keyboard Shortcut: and enter the shortcut you want to use to properly invoke the System Voice to read a selected text in every application that has this menu item.

Note: The default shortcut assigned in Accessibility > Spoken Content: Speak Selection > Options (Option + Esc) cannot be used here. So, leave that shortcut alone and come up with a new shortcut that is convenient for you to use. Personally, I chose Cmd Option Spacebar, and this shortcut works perfectly.

• (Optional Step) If you want to create another shortcut for Stop Speaking, repeat the procedure described above but create a new entry under All Applications and name it Stop Speaking. Then assign another shortcut to it.

Note: Personally, I chose (Command Option Control Spacebar).

I hope this procedure will give you something to enjoy at the end of this awful 2020.

I had the same issue. The way I fixed it temporarily is by going to System Preferences>Accessibility>Spoken Content>System Voice and then I unchecked all the voices that I don't like. It will ask you if you want to delete the unchecked selections and I chose yes. For a short amount of time it worked and the computer was only able to pick from the voices I had on the system to run "speak selection".

It *should* only randomly pick from the checked voices that I want but my issue now is that "speak selection" is completely broken and doesn't work with the shortcut keys I have chosen. I'm currently looking to fix that if anyone from apple is reading this.

I have a 2014 MacBook Pro 13 inch with the latest Big Sur software update. I've tried toggling siri on and off, toggling speak selection on and off, restarting, and reinstalling MacOS entirely to fix the bugs I've run into lately but nothing works. Kind of annoyed with this new batch of updates. :[

*rozhasi: You solution is a manageable work around but it doesn't allow for the "show controller" option to control the speech realtime. But it works as a solution until they fix this.

If you guys are interested in having the “Spoken Content” feature work as the Speech feature wiled in previous macOS versions, you should call the Apple Accessibility support line and ask them to create a case and send it to Engineering.

Engineering considers the new way whereby the system decided what language to use to read a selected text the proper behavior. According to them, it’s not a bug but the desired behavior. So, if this behavior is not desired by you, you should make your vice heard. We are at the real risk of having lost this feature forever unless each of us raised our discontent with how they implemented this feature in Big Sur.

I’ve made an effort calling the Apple Accessibility support and making them create a case and send it to Engineering. Yes, it took some time. Now it’s all of yours time to do your part.

The way that Apple bungled up this feature is worth downgrading to Catalina, where this feature works properly. Unfortunately, my new M1-based Mac can’t run Catalina. So, I’m

counting on you guys to push Apple to fix this issue.

What is the number to the engineering department? I'm motivated to make a point to get my monies worth for the mac I bought to do better than other computers. Besides, I'm a young engineer that wouldn't mind networking a little with some people even if it doesn't pan out. I'm currently out of work so I've got the free time to make a stink.

Thanks for posting here. We already know of this problem. That’s why this thread was created. There’s a workaround posted that works in some apps.

My understanding is that Apple support doesn’t read these forums. So, please call Apple Accessibility support phone number, have them open a ticket on this issue and ask them to forward it to Engineering. At this point, if we don’t become a squeaky wheel for Apple, we won’t get the grease.

Posting here will not get this issue solved by Apple.

There is no direct number for the Engineering department.

Tou have to call the Apple Accessibility support, explain to them the problem, and then ask them to create a support ticket and forward it to engineering.

I want to re-iterate. This problem will. It get solved unless each of us demands that it be solved by having a separate support ticket open and having it sent to engineering. So far, the engineering response that I received was that the way it’s working in Big Sur is not a bug but a desired behavior. If this is not a desired behavior for you, make it known directly to Apple. Please.

I finally got your work around working to take care of long selections to read. It works pretty well even without the controller. But I've not been able to test it on webpages or email, just TextEdit. It helps me proofread writing. But now "speak items under the pointer" is not responding so more bugs are in the update than they are admitting to. I get it to work for a short while and then poof, it overloads itself and bugs out. But thanks for the quick fix for reading long selections.

So I'm still problem solving the spoken content problem and I tried making a test user account to see if it is my user account that is the issue and all 4 of the Spoken Content options worked in the test user account. The question is what do I do with this information. I don't want to scrap my current admin account. It could be a non native program I'm running in Mac (I suspect google drive is a possible culprit because it runs in the background) but I want extra data storage for cheap so I don't exactly want to give up that free storage.

The question is what are you supposed to do when the issue you have is resolved by making a test user account?

I will also file a report with apple support. apple saying it works as intended is ridiculous. When I select an English voice and mark an english text it speeks in a German voice (system locale is German). That's a bug any way you look at it.

I´ve found that is affecting Siri voices though not others. I´ve downloaded most for several languages.

*rozhasi: You solution is a manageable work around but it doesn't allow for the "show controller" option to control the speech realtime. But it works as a solution until they fix this.