Latest post Fri, Jun 19 2009 3:16 AM by blafarm. 6 replies.
Page 1 of 1 (7 items)
Sort Posts: Previous Next
  • Mon, May 18 2009 7:17 PM

    Script Sync Workflow for Behind the Scenes / Reality TV

    I am pretty excited about my behind the scenes workflow using Script Sync but there are still a few issues that could be improved. Just cause I think its cool I will explain what I have set up.

    1. Playback Clip in Avid and verbally transcribe behind the scenes dialog to new wav file using Audacity

    2. Split Audacity stereo track, get rid of empty right channel, and save at mono wav

    3. Use Dragon Naturally Speaking 10 (PC Application) to transcribe wav to text file using batch process

    4. Use Advance Find and Replace to reformat raw text files into readable transcripts.

    (Because this is behind the scenes footage with many different people on camera I announce the speaker when verbally transcribing. Using short titles like LEAD, ACTOR, CINEMA, ASSIST, GRIP, etc. I can search my text output for those terms and reformat it so that a speakers title is BOLD and always starts on a new line. This makes what was a jumbled run on paragraph from Dragon into a very readable transcript)

    5. Import original WAV file from transcript and new text file transcript into Avid

    6. Create a new sequence in Avid for each master clip, add a new audio track, insert the transcript audio in new track, create a new subclip from the now three audio track sequence

    7. Open script file and drag new three audio track subclip into script

    8. Script sync to track 3 (transcribed audio)

    If you had really clean audio you might be able to script sync to your source footage audio, but for my project this is not the case. Often there are a few people talking at the same time, or other work noises happening between takes when much of my behind the scenes footage was shot.

    9. Happily navigate and edit from your script locked video.

    I learned a few things that any clever editor doing this would figure out, but I will save you the hassle and share my knowledge here.

    A, When tagging speakers in the footage use things that the speech recognition software can easily identify and that do not occur in the dialog of your participants. I made the mistake of calling the sound mixer sound, but since lots of people say sound on a movie set, I have mistaken instances where someone speaks sound, but the script looks like the Sound Mixer was say it. Camera is a similar word, but I figure this out early enough to use cinema instead. Proper names can get confused, so I used Lead for the main actor and Actor for anyone else. If i did this in the future I don't think I would break down the people as much, just identifying people by department would have worked fine. When you do the find replace you can make the title anything, so "assist" which Dragon got right most of the time became AD in my printed and imported transcript.

    B, When I imported my audio in the sequence I offset it about two seconds from the true sync point. Since I am speaking everything after I actually hear it in step one, offsetting about 2 seconds puts my words on top of the original speaker most of the time. This method will never be frame accurate, but getting to within a second of a specific line of dialog has been a huge help in re-finding spoken material.

    If AVID wanted to make me really happy they would add the following functionality into Media Composer

    - Add a transcribe engine like Dragon directly into the program and allow me to punch directly into a clip synced track

    - Allow me to batch script sync several clips at once. It's not a huge deal, but with 20 subclips to sync, it would be nice to select them all and go have coffee

    - Allow me to text search any script in the project from one place.

    - Let me print a timecode referenced script using any one of Avid's timecodes

    - Allow me to script sync to any aux audio file (like my wav transcription) using TC or in points etc. so I don't have to do the work of making all the new 3 track subclips.

    I hope I have saved someone at least as much time as it took me to type this. I would say bring on the transcript projects, but I would really rather be doing actual editing (which is why I worked out this system), but if you want to pay me big bucks to come to where you are and help you set this workflow up with your assistants I am ready to go.

    Although there are things I would love to see work a little smother, Avid has created a masterpiece with script sync that I until recently I have never bothered to even investigate. If you are not using it yet, you should be.

  • Mon, May 18 2009 7:41 PM In reply to

    Re: Script Sync Workflow for Behind the Scenes / Reality TV

    seedavework:
    - Allow me to text search any script in the project from one place.

    Not sure if this is the same thing you are asking or not, but you can hit CMD+F with the Script active to search the text.  

    Kenton VanNatten | Avid Editor (for hire)

    "I am not obsessed... I'm detail-oriented"

  • Mon, May 18 2009 7:54 PM In reply to

    Re: Script Sync Workflow for Behind the Scenes / Reality TV

    Actually I could solve two of my problems if I would combine everything into a single script and then reference the individual files (or even create one long file). This would allow me to use find in the singe long transcript which makes more sense.

    Thanks,

  • Tue, May 19 2009 4:01 PM In reply to

    Re: Script Sync Workflow for Behind the Scenes / Reality TV

    and thank YOU for sharing your workflow!

    2 systems: Mobile: 16-inch Macbook Pro 2.4GHz 8-Core i9 w/64 gig RAM, 2TB SSD, OS 10.15.3 running MC 2020 Beta Legacy - MC v5.5.4, OS 10.5.8, QT 7.6, on... [view my complete system specs]
  • Tue, Jun 16 2009 12:46 AM In reply to

    • blafarm
    • Top 500 Contributor
    • Joined on Wed, Jun 21 2006
    • Posts 300
    • Points 3,985

    Re: Script Sync Workflow for Behind the Scenes / Reality TV

    Yes...thank you for sharing that workflow.  Very interesting.

     

    Have you considered:

     

    1.  Create new sequences for all of your master clips -- or if matchback is not a concern, create one new master sequence that contains ALL of your master clips.

     

    2.  Use Avid's Audio Punch-In tool and narrate your descriptions in real time -- and in dead sync to your source material -- onto a new audio track in your Avid timeline.

     

    3.  Export your new narration audio track as a wav file and perform your Dragon speech to text conversion and find/replace formatting.

     

    4.  Then follow the workflow you already described.

     

    Nothing revolutionary here -- but this process would seem to save a couple of steps and provide the additional bonus of being able to easily edit (sucessive punch-ins) your narration without having having to re-record the entire wav all over again.

     

  • Tue, Jun 16 2009 6:33 AM In reply to

    Re: Script Sync Workflow for Behind the Scenes / Reality TV

    Great way to work with Script Sync. Thanks, I know quite a few folks who are waiting on "built in Dragon" too.

    Don't trust the internet...

  • Fri, Jun 19 2009 3:16 AM In reply to

    • blafarm
    • Top 500 Contributor
    • Joined on Wed, Jun 21 2006
    • Posts 300
    • Points 3,985

    Re: Script Sync Workflow for Behind the Scenes / Reality TV

    seedavework,

    Just curious. 

    As you have obviously experimented and thought this novel workflow through very comprehensively -- is there a reason why the suggestions I made earlier in the thread won't work -- or are not as effective as what you are currently doing?

    If so, does it have anything to do with using one of Dragon's approved noise-cancelling microphones?

    Thanks in advance for your feedback.

     

Page 1 of 1 (7 items)

© Copyright 2011 Avid Technology, Inc.  Terms of Use |  Privacy Policy |  Site Map |  Find a Reseller