timonvanhasselt

Not the greatest solution, but a possible workaround with AppleScript. I combined 2 AppleScripts, based on this tutorial: http://apetronix.com/switch-audio-outputs-with-a-keyboard-shortcut-on-os-x/ HDMI and AirSpeaker are stand-ins. Set the currentAudioSource to (do shell...

Any news on the topic of bringing Natural voices to NVDA? The natural voices for Narrator are now publicly available in Win11 version 22h2. Thanks in advance.

That's a pitty. Hopefully in the future. Thanks.

Maybe a role="application" instead of role="main" can help, then screenreaders will switch off the automatic virtual focus. I've tested it with NVDA and Jaws, but I'm not sure if it's...

I've made a possible fix with the help of ChatGPT. The issues seems to be caused by the voice-list.json url that is redirected by HuggingFace. You can test it by...

I'll try 2025.2 beta 2 too. Do you have maybe logs? And what voice did you want to download? Maybe it's something with specific voices also (just want to rule...

I understand. I've mentioned new logs because I've changed quite a bit around downloading too.

@mikebayus Great! @mush42 are you open for [this PR](https://github.com/mush42/sonata-nvda/compare/main...timonvanhasselt:sonata-nvda:main) to solve the issue? Maybe there's a better way, but feel free of course.

I've experienced it again too, also mentioned in #64. I've made some minor changes to make it work again. It's all about the resolving the right URLs of the voicemodels...

@grnowak does it help if you download the [voices.json](https://huggingface.co/rhasspy/piper-voices/raw/main/voices.json) manually to `%appdata%\NVDA\sonata\voices\piper` as piper-voices.json?