On the Floor at NAB
[April 2018] With a million square feet of display booths, no one person can “see it all” at NAB. So, we asked a few folks what they thought were key “finds” that would be of interest to others. Here is part of what Richard Rudman saw.
The classic question one is asked during and after NAB is “Did you see any new stuff?”
NAB has made it easier to answer that question by having most really new stuff corralled in one place, the Futures Park.
I visited that place and found the following: Speech-to-text recognition (StT) that really works!
The Speech to Text Challenge
My most recent experience with (StT) was a few years ago when I fractured my shoulder in two places and had my right arm immobilized in a non-typing position.
I thought Dragonspeak might help me commit words to page faster than hunt-and-peck with one hand, so I bought the latest version. Getting audio into the Dragon program was easy. Editing it to make sense – Quite another story.
In the end, I hunt-and-pecked my way through recovery and never once opened the Dragon app again. Aside from never going near “the Dragon” again, I proved the old adage: writing is easy. Editing is hard.
This year, I found a product called Descript [descript.com]. They bill it as “the world’s first audio word processor.”
The lower part of the Descript software GUI looks on screen just like many audio editors that display waveforms. 2 A nice person name Jason demonstrated what it could do.
After Jason spoke a paragraph with my first name into the system, he sent it off to a service provided by our friends at Google who worked their algorithm magic and send it back in short order. The text paragraph returned was perfectly punctuated and even was smart enough to capitalize the first letter of my name.
Jason repeated the word “so,” so that had to be edited. You can do that in the waveform editor or the text. The revised text is whisked back to Google and returned. Again in very short order, revised.
As they say on their website, “Editing audio is as intuitive as your favorite text editor. Cut, paste, copy and delete — remix your text, and your audio follows.”
Easy to Use
As Jason showed me at the Descript booth, using the product is easy, and it seems to work as advertised. Discript offers an option for occasional users with a fifteen cent/minute charge for Google processing and no software usage fee. If you sign up, Google processing using their app will only cost you ten dollar per month and seven cents a minute for Google processing.
Is Descript for you? Their website has some FAQ’s that might help.
– – –
Richard A. Rudman, President of Remote Possibilities in California, is a frequent contributor to the BDR.
You can contact Richard at firstname.lastname@example.org