Results 1 to 5 of 5

Thread: Lip synch: shouldn't we be resolving video not aud

  1. #1
    Scottness
    Guest Scottness's Avatar

    Post

    We had a flat battery on our crystal synch so it's left me with some lip synching to do manually - I've decided the best way to do it is to cut and paste small sections of audio - like single sentances or statments - under the corresponding video on the time line - in as best fit as there can be -- and then to alter the speed of that portion of the video to get it to fit better -- remembering it's only a small portion of the video you are changing the speed of and only a very small rate - so it doesn't look noticeable.

    But anyway one thing I tried before I got to this was lengthening the sound file to the same as the video file (with out changing it's pitch)- which didn't work of course because the variations in the cameras running were spasmodic - but it showed me an interesting thing -- the actors performance suddenly sounded more metered and lifeless - just with out the emaotional verve of the original -- then I realised of course it's the video that's wrong, not the audio - the audio was recorded on a crystal synched mini disc and is the only accurate record of the performance we now have - so why try and change it? - and it made me think aobut the discussions here about audio resolving software? - well if you have a synch pulse track telling what changes should occur when - wouldn't they be better applied to the video not the audio -- is that possible, or is it just alot easier to do to the audio?

    Scot M

    ------------------

  2. #2
    MovieStuff
    Guest MovieStuff's Avatar

    Post

    Audio is simply more mallable than the video since the video is made up of only 24/25 increments per second (if coming from film). Therefore, any disruption of the film "rate" is going to be more obvious depending on the motion on screen. If there is little or no action, then changing the video to fit the audio would work okay. If the subject is moving quickly, then there are fewer increments that constitute the volume of action flow. Removing or doubling any one of these increments can sometimes look disturbing and un-natural.

    More importantly, in your case, you have the benefit of editing in PAL which means that, later, if you wanted to do a blow up to 16mm from your S8 original, all your audio would match easily on a frame to frame basis. However, if you start screwing around with the video (speeding up and slowing down) then a potential blow up is going to be problematic as you would have to emmulate the same "screwing around" during the blow up that you did on the edit time line. That can be tricky.

    I think it is better to leave the picture alone and adjust the audio to fit, even it means looping some dialog to fix problems in the field. However, you should be able to match the existing audio to the field footage fine, even with the loss of crystal synch. What Premier does NOT have is a "pitch change", which is what you need to re-establish the vitality of the performance. There are a variety of free-ware downloads that will do pitch change that you could use to get people's voices back where they belong. I think you'll find that will make a big difference, especially if the camera was running a bit fast off-synch and the audio now sounds slow and lifeless.

    Roger

    ------------------
    Roger Evans
    MovieStuff
    http://www.afterimagephoto.tv/moviestuff.html

  3. #3
    Actor
    Guest Actor's Avatar
    I've had a few thoughts along this line that I'd like to throw out for your consideration. If you have shot film at 24 fps and had it Rank telecined to 30 fps and you have no intention of ever going back to film (are willing to burn bridges behind you) then every 5th frame of your video is a repeat of the previous frame. The pattern is ABCDDEFGHH etc. This means that for every 10 seconds of video, 2 seconds are repeated frames, evenly spread throughout the clip of course.

    This means that a clip of a falling object has the object stopping for 1/30 second six times per second. Yet it still looks smooth.

    This means if you could theoretically shorten a 10 second clip by up to 2 seconds simply by deleting the repeated frames. I would require an editor that would let you do so and a lot of patience.

    Could you stretch by adding repeat frames, say a pattern of ABBCDDEFFGHH, and still have it look good? I don't know?

    I have not tried either of these. It's something to think about.


    ------------------

  4. #4
    MovieStuff
    Guest MovieStuff's Avatar

    Post

    Well that's a very good point. The 3:2 pulldown is both a curse and a blessing if you need to shorten a take, I guess. If you do not take care in which frames are doubled or cut to alter the speed of a clip, then you could accidentaly double a frame that has ALREADY been doubled during the 3:2 pulldown. Likewise, if you cut out a frame that comes just before the doubled frame, then that will make a definate hicup on screen where the action suddendly jumps. However, as you cleverly point out, judicious editing would allow you to remove the 3:2 pulldown on certain frames to shorten a take without sacrificing motion characteristics.

    Clever boy....

    Roger

    ------------------
    Roger Evans
    MovieStuff
    http://www.afterimagephoto.tv/moviestuff.html

  5. #5
    Scottness
    Guest Scottness's Avatar

    Post

    Well there's one advantage you guys have over us shooting in PAL countries - I shot and ranked at 25fps - which ordinarily would be quite good (until this happens).....

    Scot M

    ------------------

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •