2 Replies Latest reply on Jun 6, 2014 7:33 PM by Ignat Fi

    speech analysis


      ok, I've searched hi and low for real anwers and everything is so vauge here even the help files for it.
        I'm using documentary footage (AVCHD original) to work on some long interviews. 

      (I'm on a macbook pro, 2011, 16gb mem, cs6 cloud

      I really want to try out the speech analysis feature.  I tried it 50 times and after the long analysis process the analysis text field is always empty.  I also tried it at home on my macpro and the same thing:  blank everytime.  I'm ok with it being BAD/ inaccurate but totally blank tells me that something isn't right.

         From what I read it's better if you have a script or reference, so I tried again using a word for word transcription of the whole thing, also through Prelude.  In prelude there is a narrow field (without a label on the interface) and a big large square field (also unlabeled in on the interface).  I don't know if one is for the "title",  and the other for "notes" or what.
        I tried copying all the text of my transcription into each field and after the long analysis it just left my text the way it was but in one solid box, like it was a single word not searchable or anything, and I don't think the analysis happened still.  One of them just left it blank like before, though, I think only the smaller field put it into the speech field after analysis, not married to any timecode either.
        Maybe I should enter it through Adobe Story instead?  but then what's the point of the Speech tansctiption log note feature?  It's really only useful to me now as a searchable edit tool from text, since I'm already transcribing myself first but the analysis part just doesn't happen at all.  Is there some setting I'm missing, does it HAVE to have Mercury playback to work? 
        All the help documents are still vauge about which fields to enter materials into, and youtube videos about it show like it's the easiest thing in the world, but when I follow the directions it always shows up blank in "analysis text" after media encoder is done. 

         The english is extremely clear and simple on both channels (some only have 1 channel audio from a lav mic).  I'm super frustrated, since it seems no one else out there has used this for interviews this way, and it seems designed for just this.  Thanks to anyone who has any experience with this.  any other troubleshooting factor ideas would be helpful, I've tried lots of things.