Dragon Naturally Speaking

Dragon Naturally Speaking is a software tool for converting speech into text. 

It has been around for more than ten years but never really become mainstream. It promises to speed up content creation ​a lot. So why hasn't it taken the content creation world by Storm? In the early years I think the technology really lagged behind the marketing hype. The computers were not a powerful and the software was less mature. Lots of people tried it and found it did not work for them. 

Dragon has gotten stronger over time​

Computers are more powerful and the software has been improved. I recently found The Writer's Guide to Training Your Dragon and am motivated to give it a try. ​The author has been using Dragon for years and points out the improvements. He and many others got past the startup issues because of health issues that left them no choice. 

Even minimal research quickly shows that learning to use Dragon is hard. The actual technique is not that impossibly difficult, but effective application requires  significant shifts in how you approach your writing. Speaking is not the same as typing and it takes a while to make the shift. 

Listen to this podcast

Scott Baker, the author of the book linked to above, did a really good Podcast in January of 2017. He made two points that I found especially helpful. He recorded himself reading some text exactly as he would for recording it for Dragon. That means speaking that includes punctuation. I found it very useful to actually hear how that sounds, not just reading about it.

Want to save $225?​

Scott also pointed out that an older version of the software is just as good as the most recent release. That is the difference between $73 and $300. In my case that is the difference between affordable and completely out of reach.

Don't expect one pass perfection​

To read the Dragon marketing material  you would think a few minutes of training would have you dictating with 99% accuracy. That isn't going to happen. It's going to take more time than you expected. It is going to require more mental flexibility than you anticipated. It may never be a good fit with your workflow or how your mind works.

The people who stick with it all seem to love​ it. There are Facebook Groups of users who will provide advice and encouragement. So I think it is time for me to give it a try.

​The right use case

I think people get discouraged when they are hoping for near perfection and it doesn't happen for them. It is really discouraging to commit hours ​and have nothing to show for it. 

I am looking for use cases where perfection is not required to have product which still has value. I think video descriptions for YouTube videos might be a good one.

Search engines still rely mostly on text to index content. They are not good at working with images or videos. On the other hand no real person is going to actually read a long description of a video. So recognition errors shouldn't matter much. ​I think that means it would be a good place for practice Dragon output that I don't feel like editing. 

Developing video scripts​

For making screencasts of software in use I prefer to wing it. If  I am reading a script, it sounds like I am reading a script, I am going to try recording a video while I speak to render Dragon audio. ​Then I will see if I can sound "normal" reading the rendered text and syncing the new audio to the screencast. I can see that timing will be an issue but hope some simple editing tricks will let me keep audio and video together.