Boris Soundbite - On Sale Now!

FCPXML? :-D

[David Lawrence] "LOL! not yet, but eventually, right?"

Maybe in a "few" weeks?

:-)

Best,
Andy

It's the same tech, ain't it? Different wrapper?

[David Roth Weiss] "I'm not yet sure what's different or what's the same. All I know at this point is both are powered by the same engine, and mysteriously Soundbite appeared ever so shortly after GET disappeared."

Now if they can just get it working in reverse to generate accurate transcripts...

_______________________
David Lawrence
art~media~design~research
propaganda.com
publicmattersgroup.com
facebook.com/dlawrence
twitter.com/dhl

[David Lawrence] "Now if they can just get it working in reverse to generate accurate transcripts...
"

We absolutely do need a transcription creation tool with enough accuracy to be truly useful. I attempted to use Adobe Speech Search yet again just the other day, and unfortunately, it's every bit as useless for documentary work today as it was 2.5 years ago when I first reviewed it.

Accuracy, even with near-perfect recordings, is hardly 50%, so the transcripts provide just and equal amount interference as they do help. So, simply stated, transcriptions from Speech Search cannot be honestly touted as improving efficiency at all.

David Roth Weiss
Director/Editor/Colorist
David Weiss Productions, Inc.
Los Angeles
http://www.drwfilms.com

Don't miss my new Creative Cow Podcast: Producing Episodic TV with "24" Producer Michael Klick:
http://library.creativecow.net/weiss_roth_david/Podcast-Series-1_Michael-Kl...

POST-PRODUCTION WITHOUT THE USUAL INSANITY ™

Creative COW contributing editor and a forum host of the Business & Marketing and Apple Final Cut Pro forums.

Pretty sure Premiere will do this for you with speech analysis.

bigpine

Soundbite, GET and Avid's Phrasefind are all powered by Nexidia. They own the IP and control licensing costs, which is why Avid has to sell it as an option. It's also the same engine behind Avid's ScriptSync. I don't know why Nexidia changed from AV3 to Boris, but I presume they felt they could get better sales through Boris' channels.

The main difference between Soundbite/GET and Phrasefind is that the former works externally and Phrasefind works inside the Media Composer host. Both work off of matching phonetics in the waveform against a language library. This means matches have a high degree of accuracy, but it's a fundamentally opposite process from actual speech transcription. So, Nexidia looks like magic, whereas Adobe's tool comes across as clunky.

The difference in accuracy is because in Soundbite, GET and Phrasefind, Nexidia can interpolate the areas between recognized words and make matches. It can also deal with words that are misspelled but a phonetic match. Since the "guesses" return valid matches, the process appears near perfect. In the other direction - generating text from such matches - results are gibberish, because in the case of Adobe, it tries to make grammatical and phrase sense out of something that it often sees as nonsense. This gives you hilarious and completely useless results.

Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com

[David Roth Weiss] "Speech Search does work, as it allows you to find and key on those words that are interpreted correctly..... Folks at Adobe suggested the accuracy would improve over time, but it it hasn't changed much, if at all, over the last 2.5 years."

Correct. I once had this conversation with product management at Adobe. They admitted that they oversold the capabilities of the function and should have promoted it as a keyword search tool. That's basically what came in with CS5.5. They also told me that over various years of researching the speech-to-text field, they were told by the experts that perfection was 5-10 years away. They told me they've been given that same future milestone over the course of 10+ years of research. It's always 5-10 years away! ;-)

- Oliver

Oliver Peters Post Production Services, LLC
Orlando, FL
http://www.oliverpeters.com