Reset READ ALOUD? It's frozen loading new voice - unclickable menus due to endless loading loop?

Question

Reset READ ALOUD? It's frozen loading new voice - unclickable menus due to endless loading loop?

Roger Akers 5

I noticed there were new voices in the Read Aloud options and tried one (they weren't named the usual human names), it was potentially an Ai test voice that slipped through but I'm sorry I can't recall the name -- I feel like it had version numbers. Anyway, since it endlessly tries to "load" the voice, the options to change it is greyed out. Any ideas how to reset to default without going scorched earth? I've reset the browser to default in settings and took a cursory glance in EDGE://FLAGS for "reader", "voice", etc. and saw nothing relevant.

Advice?

3 answers

Your answer

Answer 1

Paulo GM 61,040 Independent Advisor

I see. At this point, we can assume that this issue seems to be on the backend rather than your system or network based on everything you've tried. Submitting all the relevant details will give the team the best chance to identify the problem. You could still try creating a new Edge profile to see if it resets the voice to default. Otherwise, we’ll need to wait for the next update to see if it’s resolved.

Answer 2

Paulo GM 61,040 Independent Advisor

There’s definitely an issue where Read Aloud can get stuck in an endless loading state, regardless of the voice selected. As far as I know, there’s no other place to reset or change the voice outside of the Read Aloud feature itself. In my case, I was able to fix the loading loop by closing Read Aloud and then changing the DNS server address. This likely helps it reach the TTS endpoint properly, since some ISPs’ default DNS servers can have restrictions. After that, the voice options became available, and I could switch to a different voice. To change the DNS server address, go to Edge Settings > Privacy, search, and services > Security > Use Secure DNS > Choose a provider > Click on the text field and select Cloudflare. If Cloudflare doesn’t work, try switching to other public DNS available and see if that helps.

Deleted

This comment has been deleted due to a violation of our Code of Conduct. The comment was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.
Roger Akers 5 Reputation points

2026-01-29T00:29:38.3533333+00:00

I had a reply but clicked off by accident while trying your suggestions. No dice on that. Tried various network settings including resetting network in windows and disabled router security etc. Tried the Secure DNS thru edge and nada. I'm pretty convinced I was (un)lucky enough to see behind the curtain and be able to pick some dev/alpha model that was temporarily on there. Unless a MS edge developer happens to see this, I'll either have to Start a new MS Profile and repair/reset Edge or wait for them to update the backend to where it "resets" everyone's connection anyway.

For now, I'll send feedback using the steps recorder and give em all the info I can. I would be surprised if I wasn't the only one with this problem depending how long those defunct voices were listed for people. My read aloud for Edge iPhone & iPad work great - they all change quickly and read quickly. I'll put in a suggestion to allow us to set a default "voice" in Edge settings anyway. I know it connects online but I don't see any negatives to previewing your default voice and changing it BEFORE it connects to the endpoint - especially on windows 11 which now has Natural Voice TTS for Narrator! I'm rambling but anyway thanks for your suggestions, I tried them all and then some! Not sure if anything will work unless they re-release Supox-Diffuse v4 to the public so I can connect one more time to change it back to Ava - heck I'll take the Trinoids voice at this point.

Answer 3

Paulo GM 61,040 Independent Advisor

Hi,

Edge does not store all natural sounding voices locally as they are streamed from Microsoft online text-to-speech services. This endless loading in Read Aloud usually happens when Edge is unable to reach text-to-speech endpoints.

In this case, try to restart the PC and see if it helps you change the voice to a different one. If it doesn't work, try to switch to a different network connection if possible.

Roger Akers 5 Reputation points

2026-01-28T22:47:46.32+00:00

I understand that but I've tried the solutions. I figured it reached an endpoint online when you change voices but the issue is; when I was looking at it there were some voices listed that were unlike the Human named voices like Aria and Ava (Natural). I assume these were test voices temporarily pushed to production or a glitch!

The list, for example was like this:

Roger (Natural)

Samantha Enhanced

Aria Multilingual

Supox-Diffuse 3.0 <------ not the name but you get the idea.

There were about 7 models listed of a similar syntax. I thought it might be a new Ai voice and tried it out. Since selecting that voice/model, it endlessly tries to connect to it while all options are greyed out. There doesn't seem to be a way to reset this other than within the Read Aloud settings itself. If there were a different avenue to change the voice in Edge settings or even EDGE://flags, I'm certain it would work again. I assume this is also why resetting edge to default values didn't work since all the read aloud settings are walled off on a remote endpoint. Perhaps there's a sync option for my microsoft profile i can nuke? idk, that's why I'm asking here.

EDIT: And to be clear, all the other voices worked perfect before selecting this anomaly voice. I typically used Aria (or Ava) Natural Multilingual. Tested several others from several dialects and no issues with load time or anything. I was hoping someone here could think of a launch option or specific setting that would reset what voice my profile is trying to connect to (since it no longer and probably was never supposed to exist).

Share via

Reset READ ALOUD? It's frozen loading new voice - unclickable menus due to endless loading loop?

3 answers

Your answer