Found 55 bookmarks
Custom sorting
OpenAI.fm
OpenAI.fm

Test out OpenAI's voices, available through their API. The "vibe" option lets you describe the tone, voice, and delivery. These voices do better at emotions than many of the other AI voices.

The cost is lower than ElevenLabs, but you need to use the API for actual production. I assume they'll create a web interface for use at some point in the future. OpenAI is vague about the data used to train its voices, which I assume means they didn't pay to license voice actors like ElevenLabs does.

4/8/25#AI#audio#text-to-speech
·openai.fm·
OpenAI.fm
Fugatto AI Audio Generator - Create Unique Sounds and Music with AI
Fugatto AI Audio Generator - Create Unique Sounds and Music with AI
This tool isn't available yet, but it promises the ability to generate sound effects and other audio based on text prompts. One especially interesting feature for my work is the ability to change the emotion for an existing audio sample. Plenty of AI voice over tools can do neutral narration well now, but they're all limited for emotional variability. This could be an improvement.
12/24/24#AI#audio#music#voiceover
·fugatto.pro·
Fugatto AI Audio Generator - Create Unique Sounds and Music with AI
Read The Words
Read The Words
Free text-to-speech application that creates mp3s of content from multiple formats. Audio can be shared on a blog as a podcast but not sold commercially. Originally designed to help students with learning disabilities access text more easily.
Imported from Diigo#audio#text-to-speech#podcast
·readthewords.com·
Read The Words
elearninglive.com » Captivate Audio Output Settings Comparison - Part One - Encoding Bitrate
elearninglive.com » Captivate Audio Output Settings Comparison - Part One - Encoding Bitrate

Results of an experiment on the encoding bitrate for Captivate. Mark Siegriest is trying to find the best audio settings to balance sound quality and file size.

The most surprising finding in this is that at the lowest bitrates, the file size actually went up, not down. 32kbps was the smallest file size in his sampling.

So basically from what I can see there is no point in ever going below 32kbps as you’re gaining filesize and losing quality.
Imported from Diigo#captivate#audio#audacity#elearning
·elearninglive.com·
elearninglive.com » Captivate Audio Output Settings Comparison - Part One - Encoding Bitrate
Donald Clark Plan B: Brilliant 35 studies in media and learning
Donald Clark Plan B: Brilliant 35 studies in media and learning
Great summary of research points on our perceptions of media with implications for using media effectively for learning. For example, audio quality matters a lot, but video quality can be low and still effective. Large, wide screens are preferred over higher quality images on smaller screens.
35 psychological studies into the human reaction to media all point towards the simple proposition that people react towards media socially even though, at a conscious level, they believe it is not reasonable to do so. They can't help it. In short, people think that computers are people, which makes e-learning work.
As long as a media technology is consistent with social and physical rules, we will accept it. Read that last part again, 'as long as a media technology is consistent with social and physical rules'. If the media technology fails to conform to these human expectations - we will very much not accept it.
Imported from Diigo#learning#research#multimedia#audio#video#images#elearning#cognition
·donaldclarkplanb.blogspot.com·
Donald Clark Plan B: Brilliant 35 studies in media and learning