4 Replies Latest reply on Jan 13, 2010 12:20 PM by the_wine_snob

    Workflow & encode advice for 2,000 vid library


      Hello and I hope this is the right forum for my questions...


      We're at the start of a project that will see us capture, encode and distribute several thousand videos to the web. We've spent a few weeks working through different processes - at this early stage we're trying to confirm if we're on the right track or if some part of the workflow should be changed to speed up the whole process. The broad details are:




      Step 1 - capture HD video in Google Earth Pro

      Step 2 - add various sound effects and audio track in PP CS4, encode in ME to h.264 - we've turned to PP CS4 but are new to the app

      Step 3 - upload to web audience, youtube, iphone etc


      Source media...

      - is captured from movie maker within Google Earth Pro

      - format is Windows Media 9, 17 mbps, 1280x720, 25 fps

      - duration is from 2 min to 10 min, avg 4:30

      - library will consist of 2,000 videos


      Question - we can capture the source media at 6mpbs rather than 17mpbs, would the lower bitrate speed up encoding?


      PP encode step...


      - The aim of this step is to add audio tracks and some minor effects to each video, and prepare for delivery into youtube, iphones, etc

      - Output file is currently h.264 MP4, 1280x720, 25 fps, 6mpbs with 10mbps max, VBR

      - We've used PPBM4 to benchmark and after moving from a network output location to local raid, it now takes about 7 mins to encode a 4 min file which seems about right for our hardware specs (available if required).


      Question - given the final renditions will be delivered from 320kbps to 2mbps via youtube and other vid sites, what encode settings should we use to get the best quality for speed balance? Is 2 pass worth it if it doubles the time? Should I use b frames? Are there other settings I either should or shouldn't use?


      Batching up lots of timelines in media encoder is giving us problems too, after 5 or 6 successful jobs it fails with encode errors. It's be great to be able to watch it blow through 50 overnight but not sure what to test.


      Are we on the right track? IF any tips can shave 5% here and 10% there it'll make a big difference with so many videos. Thanks in advance...