Yes. It's normal.
There are programs designed to help you type into a computer. They are better at Speech analysis. Heck, my SmartPhone is better at Speech analysis.
If I talk to my phone, it gets the punctuation right most of the time. Capitalization also.
That's why it's called a Smartphone
Thanks for the replies.
I'm aware that there are several speech to text programs available, however, If you know of one that also includes the embedded time code, used for subtitles.. I would be very much interested..
Oh! That is a horse of a different color. Hmmm. I hadn't though of that. Subtitles?
Take a look at the bottom paragraph of this page: http://help.adobe.com/en_US/Soundbooth/2.0/WS29F692D2-7A49-44db-A109-AE016B766767.html
It says, in short, that inputting the original script helps Premiere Pro get the words right as well as the punctuation. So, if you used a SmartPhone, for example, to analyze the speech, or maybe even Dragon Naturally Speaking, you would then have the text that could be used as a script and then Premiere Pro could use that to help it analyze and you would have the timecode.
Whew! That is a convouted process that might actually be quite helpful. I have no idea how much of a hassle that might be but I get the feeling that it will get you a lot closer to your goal.
Edit: Look at this also: http://help.adobe.com/en_US/story/cs/using/WSeffff8bffc802084-53d9eb2012efc973f75-8000.htm l
Thanks for the links, Steven.
My method for getting an accurate content analysis is sound. that's not my problem.. Its getting the results while keeping the punctuation and capitals is the issue. Using Premiere with an audio clip and exact text script doesn't seem to work. Whether or not "Story" will correct this is uncertain.
Sill, even if I was able to obtain a proper meta data file. The method for bringing this into After Effects to create subtitles, is also another problem. The expression used in the process, simply calls up every 8 words, instead of the logical way of displaying from the beginning of the sentence.
I guess I wish there was a way to convert the content analysis meta data to an.srt file.