Forum Replies Created

Page 17 of 19
  • Lars Bunch

    May 10, 2006 at 3:22 pm in reply to: how to approach this

    Hi,

    I have no experience at this kind of elaborate production so the following are just random guesses of how I would start to tackle the problem;

    It seems you could do this by using an array of still cameras positioned along the camera track, thereby capturing a freeze frame of every moment in the sequence from every camera position. One pass is made with no actors on the set so you get a clean plate of the set upon which to composite the actors. In the second pass, you cover everything with greenscreen material and shoot the actors moving naturally. You can then sequence the material into motion footage and key out the actors so that you can place them into the clean set compositing successive freeze frames on top of previous ones.

    The bird is probably CG. The first freeze frame and the last freeze frames are, of course, shot with a motion picture camera.

    While conceptually pretty simple, the execution requires a great deal of planning. I would prefer to use digital cameras in the array so that a system could be arranged to be able to view the motion version of the action on set. This would also reduce the problem of having the scan each frame and then align them properly.

    If you look at the camera move carefully, you’ll notice that, until the final part of the move, it never turns so that you could see the “dolly” track. This ends up requiring a move that looks a bit like a long trucking shot, but is a lot simpler than trying to clean up the cameras that are in frame. This suggests that it was shot rather than created in CG. Obviously you would have to clean up the left part of the frame at the end of the shot, but you’ll notice there are no moving people on the left. The set extension was probably built in CG and tracked using data from the camera rig.

    In order to plan the move, I would want to pre-visualize it in 3D animation and use the motion path of the CG camera as a basis for the rigging upon which the still cameras would be mounted.

    But another approach is to use extremely high quality digital actors and motion capture. But my guess is it was shot with real actors.

    I guess it all depends on whether you prefer film set headaches or computer graphic headaches.

    Lars

  • Lars Bunch

    May 1, 2006 at 2:44 pm in reply to: Text Pixelates During Camera Move

    Hi,

    This may not be the answer, but it might be worth a try… Add a median filter to your text and set it to a really low setting like 1.

    I have found that this helps clean up aliasing in high resolution images and it also might help in text with sharply defined edges.

    Hope this helps,

    Lars

  • Hi,

    Yes, Shake can handle the anamorphic settings. Just make sure your compositor knows that the footage is anamorphic and that the resulting composite should be rendered with the same anamorphic squeeze and everything should work out fine.

    Lars

  • Lars Bunch

    February 23, 2006 at 3:52 pm in reply to: Field dominance question urgent!!!

    Hi,

    I spent months dealing with this stupid problem and what I finally discovered is that Compressor defaults to GUESSING the field dominence. And sometimes it doesn’t guess right. When it guesses the wrong dominence, you will end up with an mpeg file that looks fine on the computer, but stutters when played back on an NTSC monitor.

    When you open Compressor, you get a selection of presets. You can make copies of these presets and then set the field dominence to top or bottom. NTSC uses Lower (AKA bottom, AKA Even) field dominence.

    Save a preset that ALWAYS selects bottom field dominence. This should work for most situations. But as mentioned above, you should do a quick test on a small section of footage.

    Hope this helps.

    Lars

  • Lars Bunch

    February 12, 2006 at 4:07 pm in reply to: Apple 30″ Monitors

    Hi,

    Get the 30″ monitor. I use a 23″ and a 30″ and my feeling is that FCP on a 23″ is okay, but I still feel like I need a second monitor for bins and such. With the 30″ I have plenty of space for most programs (although I’d like to have 2 30″ monitors when running Maya.)

    There is no sluggishness in running such a big monitor. The eveness of illumination and color consistancy is better than on the 23″. The one disconcerting aspect is that sometimes when coming out of sleep mode, you will get green pixels sprinkled throughout the screen. The solution I have found is to shut off the monitor with the button on the side and turn it back on. This always clears up the problem.

    A few things that can take a bit of getting used to… If you sit fairly close to the monitor, you may have to turn your head as you look from one side of the screen to the other. Also, finding your cursor within the vast real estate can be annoying. I use a flat gray background and I still have to move the cursor a bit to see where I’ve left it.

    Anyway, I commonly use Photoshop, Illustrator, FCP, After Effects, Shake and Maya, and as mentioned above, Maya is the only program I commonly use where I still need more screen space.

    Hope this helps,

    Lars

  • Lars Bunch

    February 10, 2006 at 4:05 pm in reply to: 16:9 Rendering/Output

    Hi,

    AE 7 has a direct comp preset for 16×9 uncompressed, but versions 6.5 and eariler do not. But that needn’t stop you from setting up your comp for 16×9.

    Just set up the comp settings to standard D1 NTSC (720×486), but on the pixel aspect ratio, instead of 0.9, set it to 1.2 widescreen.

    Render it out as 8 bit 4:2:2 uncompressed Animation Codec. There is no need to specify 4×3 or 16×9 here. You will end up with a quicktime file that will play back squeezed when displayed with square pixels. When you bring it into FCP check the “anamorphic” field in the bin. Then FCP will display the file with a 1.2 pixel aspect ratio which is what it’s doing with the footage you digitized as 16×9.

    Anyway, that’s what I do and I haven’t run into a problem with it yet.

    Lars

  • Lars Bunch

    February 9, 2006 at 3:46 pm in reply to: sata vs esata

    Hi,

    Please someone correct me if I’m wrong, but here’s what I think I know…

    the SATA connector is designed to be used within the computer box. The eSATA is an EXTERNAL connector designed to run cables to an external device (thus the e prefix)

    The eSATA cables are more expensive, but probably worth it.

    Here is a link to a store that shows the differences… You’ll have to scroll down the page to find it…

    https://www.macgurus.com/productpages/sata/SATACables.php

    Anyway, I have a SATA raid that uses the internal cables and I have to run them out a hole in the back of the computer in order to interface with the RAID. Very annoying. eSATA seems like the appropriate option if you are using an external RAID box.

    Hope this helps,

    Lars

  • Lars Bunch

    February 9, 2006 at 3:32 pm in reply to: animating projector

    Hello,

    The first complication that I see is that the reels in the projector image are seen at a slight angle. This means that if you were to cut them out and then rotate them about their center, they would wobble. Worse yet, the overlap where you can see the far side of the reel would rotate with the near side.

    If I were dealing with this, I would cut out the near side of the reel and tweak it in photoshop so that it was a perfect circle. I would also cut out the far side of the reel and and steal bits to create a perfect circle far side that I could rotate. I would then create two layers that would rotate about their centers using an expression to drive the rotation.

    Actually I’d probably just rebuild the reel in Maya and drop it in, saving me the headache of fixing it in photoshop.

    As for the projector beam, I would create a white solid, mask it into a rough beam shape, use a radial blur centered at the lens nodal point, maybe throw in an animated fractal pattern (before the blur) to give a feel of the beam intensity fluctuating with the image it is projecting… Actually, if you can drive the beam intensity with actual footage, you might get a more realistic effect. Then I would adjust the opacity of the beam layer to get the right amount of beam. You might also want to feather the beam as it gets farther away from the projector to give it more of a sense of falloff.

    I haven’t actually done anything like this so I may be way off base here. I’m just guessing at how I would approach it if I were to do so.

    Hope this helps,

    Lars

  • Lars Bunch

    February 9, 2006 at 3:32 pm in reply to: animating projector

    Hello,

    The first complication that I see is that the reels in the projector image are seen at a slight angle. This means that if you were to cut them out and then rotate them about their center, they would wobble. Worse yet, the overlap where you can see the far side of the reel would rotate with the near side.

    If I were dealing with this, I would cut out the near side of the reel and tweak it in photoshop so that it was a perfect circle. I would also cut out the far side of the reel and and steal bits to create a perfect circle far side that I could rotate. I would then create two layers that would rotate about their centers using an expression to drive the rotation.

    Actually I’d probably just rebuild the reel in Maya and drop it in, saving me the headache of fixing it in photoshop.

    As for the projector beam, I would create a white solid, mask it into a rough beam shape, use a radial blur centered at the lens nodal point, maybe throw in an animated fractal pattern (before the blur) to give a feel of the beam intensity fluctuating with the image it is projecting… Actually, if you can drive the beam intensity with actual footage, you might get a more realistic effect. Then I would adjust the opacity of the beam layer to get the right amount of beam. You might also want to feather the beam as it gets farther away from the projector to give it more of a sense of falloff.

    I haven’t actually done anything like this so I may be way off base here. I’m just guessing at how I would approach it if I were to do so.

    Hope this helps,

    Lars

  • I tested this out and noticed it only worked on the X value so here is an expression that would work on both x and y scale values…

    Xscale=thisComp.layer(index-1).scale[0]+4;
    Yscale=thisComp.layer(index-1).scale[1]+4;
    [Xscale,Yscale]

    Hope this helps,

    Lars

Page 17 of 19

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy