It would be better if you can directly record the sound using the client's microphone, but if that's not possible, a upload dialog similar to the image will be sufficient.
Sure all of questions in the pronunciación tag would benefit from that adition.
I think also the others X Language and Usage sites in the SE network will welcome that feature.
How it works
while writing an answer
First, the editor features a new button:
When the button is clicked, a dialog is shown:
The dialog function basically the same as the image dialog. Bonus points if the server supports automatic windows audio (wma produced by standard sound recorder) to mp3 conversion.
The url of the sound is referenced the same way as a regular link in markdown. I'm not too sure here about if
- Markdown have any standard way to reference a sound (if not it looks like a nice adition, given the html5 standard sound support).
if not, then what's better
- to reference it using the same exclamation sign like the images,
to use other char to identify the link as a sound, using similar syntax, for example:
*[Pronunciación de **RAE**, acento centroaméricano]
while displaying an answer
The markdown to html converter will produce the needed markup to let the user hear the audio, for example:
<figure> <audio controls> <source src="http://a.mp3ur.com/ueDY4.mp3" type="audio/mpeg"> <figcaption>Pronunciación de <strong>RAE</strong>, acento centroaméricano<figcaption> </audio> </figure>
The answer then will look like this: