Reply To Thread
Results 1 to 10 of 10
  1. Collapse Details
    VSA Wave Motion Analysis vs TrackSkull
    #1
    buckaneerdude's Avatar
    buckaneerdude is offline Bucky Brain Surgeon
    Join Date
    Sep 2008
    Location
    Huntington Beach, CA
    Posts
    887


    I've been using VSA for a few years now and I've used the Wave Motion Analysis feature to create jaw motion for our props. I never got what I would consider great results. Typically I would run the motion analysis several times trying different % settings. After getting that working close to right I had to spend a lot of time editing the tiny details of the jaw movement to fix the more visible errors.

    I picked up MonkeyBasic's Trackskull program and used it to reprogram the head movements for our new props and I was very happy with the results. I was so happy in fact that I went back and recorded new head movements for all of our props. This was way easier than laying things out in VSA. I kept the old jaw movements since I had spent so many hours editing in VSA.

    While editing the jaw movement of a new prop I had made using VSA and Wave Motion Analysis, I thought I'd try using TrackSkull for the jaw movement. I set it up so that the gun trigger on a joystick activated the jaw. At first I didn't like it that much. However, some parts of the bit turned out really well and others were just off. That's when I decided to invest some time in practicing the motion. What I learned was that I expected the jaw to return to the closed position after each syllable and the jaw just wouldn't move that fast. Then the more I practiced, the more I noticed that I could kind of bounce the jaw between 1/4 and 3/4 open and get the jaw to move to each syllable. The more I practiced, the better I got. So, I went back and deleted all 7 of the wave motion analysis jaw tracks and replaced them with ones I made using TrackSkull.
    Reply With Quote
     

  2. Collapse Details
    #2
    soundmankkc's Avatar
    soundmankkc is offline The Great Pumpkin
    Join Date
    Sep 2009
    Location
    Smyrna TN
    Posts
    309
    Bdude, How much practise did it take? I am almost set to start programming my 3 axis w/eyes and jaw and any helpful tips would be greatly appreciated. BTW, did you see the bit I posted on your arm thread about the low priced servo's? Keith
    skating away on the thin ice of a new day
    Reply With Quote
     

  3. Collapse Details
    #3
    buckaneerdude's Avatar
    buckaneerdude is offline Bucky Brain Surgeon
    Join Date
    Sep 2008
    Location
    Huntington Beach, CA
    Posts
    887
    It took a couple of hours to get good at it. Overall, it took a lot less time to do it with TrackSkull than creating a wavemotion analysis in VSA and editing it.
    Reply With Quote
     

  4. Collapse Details
    #4
    Hooked_on_Scares's Avatar
    Hooked_on_Scares is offline The Great Pumpkin
    Join Date
    Apr 2009
    Location
    Los Angeles
    Posts
    436
    Blog Entries
    3
    I haven't used TrackSkull yet, but this last year I did some similar fine-tuning on the way I used WaveMotion. The two biggest challenges I found were...

    1) because of both communication latencies, and mechanical latencies, the jaw movement was always behind the speech.
    2) loud speech is fine, but quieter speech doesn't make the jaw move enough

    The first one's not hard to correct. Just select all the events in (only) that track, and move them by a frame, or two, or three.

    The second one takes a little outside tweaking. I imported the voice track into Adobe Audition, and applied the compander function until the 'quiet' parts were just as loud as the loud parts were. Then, when you run WaveMotion Analysis on it, the jaw moves equally far for quiet and loud parts. Once you're done, you can reimport the un-tweaked audio file back into VSA, and finish the rest of the animation.

    - Hook
    Reply With Quote
     

  5. Collapse Details
    #5
    icemanfred's Avatar
    icemanfred is offline The Great Pumpkin
    Join Date
    Aug 2005
    Posts
    1,173
    Has anyone tried puppeting the jaw with the joystick?

    I think during natural speech, the jaw dosent open and close completly for each syllable.
    But once a sentence is started the jaw varies throughout the sentence then closes.
    Or at least thats the way it appears on SKulltronix original routines. And I think they are among the best.

    Using the joystick would allow you the control over how much the jaw opens .
    I hvent tried this. But have read about it in other forums.
    Reply With Quote
     

  6. Collapse Details
    #6
    buckaneerdude's Avatar
    buckaneerdude is offline Bucky Brain Surgeon
    Join Date
    Sep 2008
    Location
    Huntington Beach, CA
    Posts
    887
    Quote Originally Posted by Hooked_on_Scares View Post
    I haven't used TrackSkull yet, but this last year I did some similar fine-tuning on the way I used WaveMotion. The two biggest challenges I found were...

    1) because of both communication latencies, and mechanical latencies, the jaw movement was always behind the speech.
    2) loud speech is fine, but quieter speech doesn't make the jaw move enough

    The first one's not hard to correct. Just select all the events in (only) that track, and move them by a frame, or two, or three.

    The second one takes a little outside tweaking.
    - Hook


    There is a scaling adjustment in wavemotion that will give you more dramatic jaw movement. I have used this many times and used scales of 150% to as much as 500%. You still though have to spend a lot of time editing.

    To answer the question about using a joystick, well yes I use a joystick with TrackSkull. I prefer the Logitec Extreme 3D Pro because it has an axis that works by twisting the stick. This works great for head rotation and then you can use the forward and back motion of the stick for nod and the left and right motion of the stick for tilt.

    I suppose you could use the joystick stick motion for jaw movement instead of the trigger, I've never tried it though. It might be kind of hard to do because of the long range of motion in the stick, however... another Logitec joystick I have has a sort of a wheel (throttle), similar to a thumbwheel that might work with some practice. I think the problem though is moving it fast enough. You might be able to increase the speed by adjusting the sensitivity to 100% in the joystick options. I have suggested that MonkeyBasic look at maybe adding some sensitivity settings within TrackSkull similar to mouse sensitivity settings available in many computer games.

    I might try it later on and see how that works. I'll report back when I do.
    Reply With Quote
     

  7. Collapse Details
    #7
    Hooked_on_Scares's Avatar
    Hooked_on_Scares is offline The Great Pumpkin
    Join Date
    Apr 2009
    Location
    Los Angeles
    Posts
    436
    Blog Entries
    3
    buckaneerdude,

    The scaling function in VSA is useful, as well. But the problem is it amplifies all the motions. If you amp it up enough that the quiet parts of dialog will move the mouth sufficiently, you run the risk of clipping the motion on loud parts (mouth just stays open, or flaps). Companding the audio avoids clipping, and makes it so both loud and quiet parts of the audio will move the mouth in similar ways.

    Sometimes a picture's worth a thousand words. Check out the wikipedia entry for a graphic.

    - Hook
    Reply With Quote
     

  8. Collapse Details
    #8
    Dr Morbius's Avatar
    Dr Morbius is offline Open up and say "Aaaaahh"
    Join Date
    Jan 2006
    Location
    In a house far up on the highest hill where the cemetery ends beyond the winding path.
    Posts
    1,559
    I just use a joystick in VSA and record the jaw on a seperate track using that. I don't see the need for external software. Plus, when I set the deafult setting on the jaw servo in VSA in a closed position, it closes when I release the joy stick. I get REAL accurate jaw motion from VSA that way. BTW I use a logitech joypad that has two joysticks on it.
    Your resident Proptologist.
    www.hauntcast.net
    Reply With Quote
     

  9. Collapse Details
    #9
    buckaneerdude's Avatar
    buckaneerdude is offline Bucky Brain Surgeon
    Join Date
    Sep 2008
    Location
    Huntington Beach, CA
    Posts
    887
    Quote Originally Posted by Dr Morbius View Post
    I just use a joystick in VSA and record the jaw on a seperate track using that. I don't see the need for external software. Plus, when I set the deafult setting on the jaw servo in VSA in a closed position, it closes when I release the joy stick. I get REAL accurate jaw motion from VSA that way. BTW I use a logitech joypad that has two joysticks on it.
    In VSA? Hmmm, looks like there is a VSA feature I overlooked. How do you use a joystick within VSA? And you are using the stick not the trigger?
    Reply With Quote
     

  10. Collapse Details
    #10
    buckaneerdude's Avatar
    buckaneerdude is offline Bucky Brain Surgeon
    Join Date
    Sep 2008
    Location
    Huntington Beach, CA
    Posts
    887
    Ok, I tried the fullsized joystick in VSA and it is way different. For the jaw movement with the joystick, full right closes the jaw and full left fully opens the jaw. I couldn't move it left and right fast enough to get a very accurate representation of jaw movement. I can see how the joypad would work better than a fullsized joystick for this application, especially for someone who is familiar with using gamepad type devices for gaming. I am a total fumble dufus with a gamepad.

    For me, I think TrackSkull and the joystick trigger works best. But if you're familiar with a gamepad and have one that will work on the PC, I'd say give it a try for sure.
    Reply With Quote
     

Reply To Thread

Bookmarks

Posting Permissions

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