What can we expect regarding speech recognition in Scrivener v3 for Windows?

I know that speech recognition capability in Scrivener v1 for Windows was very limited. Does anyone know what the plans are for speech recognition in v3? For example, is Dragon expected to work directly into Scrivener text windows? Will there be a way to issue Scrivener commands directly using speech? And so on…I know things may still be changing, but it would be nice to know what is being planned/considered. Most of what I was able to find online is largely speculation, so an “official” setting of expectations would be most appreciated. [Apologies if I used the wrong forum for this post]

That feature alone would be worth the purchase of scrivener or the 3.0 upgrade. It would be nice to have comparability with Amazon’s Alexa or Google’s Home voice recognition systems.

What does this even mean? Scrivener is extremely unlikely to learn how to turn your lights on or adjust your thermostat.

To the OP, you might want to ask Dragon about their plans for Scrivener compatibility. There isn’t actually much to do on our side: Scrivener just takes the input from whatever source, be it the keyboard, the mouse, or Dragon-generated text.

(For the technically inclined, the Google Assistant SDK is public. You can get it running on any computer you like in a weekend or so.)


I’d settle for Scrivener acquiring sentience and being able to do the dishes.

This might not be a Dragon-side thing, depending on how Qt is building the various UI controls. If I understand correctly, Dragon looks to hook into specific standard Windows control instances. If the corresponding Qt controls are not using those instances, or not inheriting the same objects and classes, then it might be up to the developer to add the appropriate SDK elements so that Dragon can “see” the Scrivener controls in the same way.

This is ringing a bell from what I read some time ago. It was my understanding that the text boxes, etc. need to be implemented in such a way that Dragon (or whatever…) knows they have focus and can manipulate the buffer (inject text into it, cut/paste with it, etc.). If that is correct, then “Dragon support” would need to be provided on the Scrivener side by instrumenting the various Scrivener controls (e.g., text boxes) so that they can be made accessible to other applications (e.g., Dragon). I don’t have any particular experience with this, so I’m just making some guesses.

There is no exposed Scrivener SDK, on either the Mac or Windows. Nor are there plans to create one.


Wrong way round. Dragon SDK.

nuance.com/dragon/for-devel … r-kit.html

This allows Windows devs to add the hooks Dragon needs, if they are using widgets that Dragon doesn’t already support. But, that requires additional licensing, so if Qt’s widgets don’t have the appropriate magic for Dragon to recognize them the way it does basic Windows system widgets, it seems it’s up to Qt to fix their object inheritance.

Ah, yes. And on the Mac – which I’m much more familiar with – Scrivener uses the Mac OS X system tools and therefore gets Dragon’s ability to connect with those tools more or less for free.


I suggest looking at Windows built-in Speech Recognition. Microsoft puts out an SDK with published APIs. This would benefit more Windows users.