I have gotten so busy lately that I have gone back to trying to write blog posts and such with Dragon Dictate, the Apple version of the popular Dragon Naturally Speaking voice recognition software. I don't usually do software reviews, so this isn't going to be one, but I will make a few comments about my experience with voice recognition software. In fact, I am going to dictate the entire remaining part of this blog post without making any but very superficial error corrections so you can see how well the program works, sort of. So here it goes:
I have discussed my personal history with voice recognition software a few times in the past. But recent advances bring me back to the subject.
My first introduction to voice recognition software was at the Seattle Worlds Fair in 1962. At that time IBM demonstrated a rudimentary voice recognition system using an IBM Selectric typewriter. Many years later, as I became involved in the desktop computer revolution, I was quick to purchase the first versions of voice recognition software. I found them to be cranky, difficult to use, and practically worthless. Part of the problem was the limitations of the processing power of the computers. Every so often, I would either update my existing programs or purchase new ones. Time after time I was disappointed by the performance of the programs. The accuracy was so low that it was easier to type in the text than dictate it.
Last year, when I was laid up for a time because of a major operation, I once again updated my program and to my surprise found that the software was now quite useful. There were still problems but the difficulties were now at an acceptable level. The main voice recognition program has been, for some time, Dragon NaturallySpeaking. The Macintosh version which I use, is called Dragon Dictate. These programs have finally moved voice recognition out of the cranky tinkering stage into a useful product that does an amazing job of transcribing speech.
Very recently, as I became more involved in writing, I found that it was time to use voice recognition more frequently than I had in the past. Other than a few weird textual mistakes, it has now become a completely useful tool for putting my thoughts into blog posts. I'm sure that all of you will be thrilled to find out that my online production is beginning to increase dramatically and since I can talk when I am half asleep, I am sure some of my posts will become even more incomprehensible than they have been in the past. As a side note, I always wonder if the patrons in the Mesa FamilySearch Library can tell if I am asleep or awake.
Unfortunately, there are only certain types of narratives that lend themselves well to voice recognition software. If you are writing something that has a great deal of citations or complex formatting, voice recognition software can drive you crazy. One of the things that I found absolutely necessary was to obtain a high quality digital microphone. It also helps dramatically to have a quiet environment. If you are competing with TVs, children playing, or other noisy environments the ability of the program to accurately represent speech diminishes dramatically. Most of the versions of Dragon NaturallySpeaking and its Macintosh counterpart Dragon Dictate come with a digital microphone. But I suggest, if you can afford it, to purchase a higher quality headset microphone.
Obviously, if you suffer from a disability that prevents you from using a keyboard, or you have never learned to type, I suggest that you may want to investigate voice recognition as an alternative. In making this suggestion, it is important to understand that using voice recognition software involves some rather steep learning curves with respect to the computer operation. If you are struggling with using a computer, adding the complication of voice recognition software is likely not a solution to your problem.
In 1844 to open the Baltimore Washington telegraph line the first message transmitted was "What hath God wrought?" I suppose we could ask the same question today with computers, webinars, video conferencing, and now, a workable form of voice recognition software.