Closed jbienzms closed 8 years ago
Oh, I forgot that I'm supposed to describe how I plan to solve this problem. I have been able to convert SpeechSyntehsisStream to a byte[] containing WAV data. I then found code to load WAV data into a mono Unity AudioClip. From there, we can assign that audio clip to an Audio Source which allows the voice to be positioned in 3D space. The TextToSpeechManager component does all of this work. The only things the developer needs to do are:
The component takes care of generating the speech using SpeechSysntesizer, converting the output stream to WAV data then on to an AudioClip. It sets the AudioSource.clip property and then calls AudioSource.play. This all happens in about 500ms.
I agree! I've been working on something like this recently as well, but have been stuck at the "convert the WAV data to an AudioClip" step. Sounds like you've solved that, so I think this would be a welcome addition. Feel free to create a pull request for review.
Awesome! Pull request has been submitted. You can review it here:
Updated to new pull request:
Updated to a third pull request:
Awesome! Left some comments. Out of curiosity, why the new Pull Request every time you make changes? If you update your branch the currently standing PR will update to reflect that merge instead.
Can you assign this issue to yourself? That way it's clear from the outside that you're working on it. =)
As for the new pull request, well, to be honest I haven't had a pull request rejected before. The one's I've done were accepted first time so I never had to go back and rework a submission. Good to know it works that way, thank you.
I would be happy to assign the issue to myself, but when I click the Edit button I'm only able to change the title. I don't think I have the permissions to assign the issue. Can you help?
I'm not sure where the Assignees list is filled from, but I wasn't able to find you in the list for some reason.
+1 used it this weekend at the Boston hackathon and it worked very well. Thanks, Jared!
Used at the Boston MS Hololens Hackathon this weekend and it worked perfect for what we needed.
Another +1, this would be great feature to bring in
@jbienzms, have you added a test scene to your pull? Please check out the Contributing.md’s implementation section.
Once your test scene is added and readme updated, I’d be happy to merge.
This was fixed by the merge of #48 correct?
Correct @jwittner. Sorry I forgot to close the issue. Closing now.
Windows 10 supports text to speech via SpeechSynthesizer but this isn't easily used from Unity because of the complexities in converting from SpeechSynthesisStream to AudioClip. I propose a new component called TextToSpeechManager (which I've already written) to do this work for us.