Saturday, January 20, 2018

Tempermental Butler versus Servant

"Your temperamental butler is listening."

My last post spoke about my helping a friend with Dragon speech recognition software, so she could speak rather than keyboard her story to create an editable text computer file. I frequently speak into my phone to create text rather than use my swype keyboard ( which I love dearly ) but have not done so with my computer. At least recently.

My experience with Irma and Dragon re-whet my appetite for speaking to rather than keyboarding at my computer. I recalled that the capability is built into my version of Windows and I had successfully experimented with it sometime ( likely a long time ) ago. I could see speaking to my computer for at least two things I do virtually daily: making journal type notes in a free program called iDiary; writing … I have, in spurts of success, committed to writing for at least ten minutes, on any topic that seems of interest to me, every day, in either my paper journal or in Microsoft Word. I could easily see speaking as making both tasks easier, more fruitful and, therefore, more rewarding.

It was not hard to resurrect what I knew.  The application is called, appropriately enough, Windows Speech Recognition.  I opened the program and went through the tutorial.  This gave me all kinds of options about commands and what not, all of which may someday be of some interest, but all I wanted to do was speak and obtain editable text as a result.  Doing that was very straight forward … and getting the program to be ready to listen to me or turn that off reminded me very much of how Dragon both worked and looked.

I opened iDiary, told Windows Speech Recognition to listen and spoke. Nothing happened.  Nothing. As though the microphone were disconnected. To say that was disappointing would be a significant understatement.

I thought I would try Microsoft Word. Microsoft makes the speech recognition software I was using and it also makes Microsoft Word … a natural, eh? You would think. You'd be wrong. It was better than using

Figure 01 - The editable text in a pop-up Insert Box window.
iDiary ( nothing happened there, remember? ) but speech did not directly enter text into the Word application. Instead, speech entered text into a pop-up window, which text I could edit and review and then enter into the document by clicking the Insert button … see Figure 01. This is clearly better than my experience with iDiary, but not a great deal better. I got disgusted and left the project for the day.  

I decided to write a blog post about this experience. Wondering if Windows Speech Recognition would work directly with the Google blog creation tool, I tried it. It worked. Perfectly. The spoken words appeared in the tool just as though I had keyboarded them. This was exciting
( "Wow, I can speak my blog post material!" ) and disappointing ( "Why can't I speak and obtain text at all with iDiary??" ) at the same time. 

Being a stubborn techie, I thought I'd try speaking with iDiary once more. This time it worked … the text didn't enter iDiary directly, but via that same pop up window of Figure 01. While I consider this less than ideal, it is far superior to nothing happening. 

I tried Microsoft Word again; guess what?  My spoken word appeared directly in the document that I was creating!  To show you, I am going to dictate a sentence and put it after the colon here just as it came from Windows Speech Recognition: "I dictated this sentence as an example for my block polls." The correct sentence is, "I dictated this sentence as an example for my blog post." Whereas Dragon seemed to be saying, "Speak, your servant is listening," this system seems to be saying, "Your temperamental butler is listening."

This system is not perfect.  There is possibly work that must be done.  I realize that my phone converts my speech into text far more reliably than this configuration does.  I wonder if that is because the microphone I use here is relatively cheap and old or whether the speech recognition software in my phone is much newer and much better … or some combination of both.  Time will tell.  Either way, temperamental or not, I'm excited about this development.

If you would like to comment but don't care to use the comment field, send an email to

No comments:

Post a Comment