[Lars Bunch]

Hi,

I have both a Wacom tablet and a Cintiq monitor. The pen that came with the monitor works with the tablet and the pen that came with the tablet works with the monitor. I believe they have RFID chips inside which helps the system to identify which pen you are using, but they do not have to be the pen that came with the device.

Lars

[Lars Bunch]

Hi,

I assume you are going to shoot this footage yourself. If you could shoot this on a greenscreen, then you could just pull the matte from that and you’d have your shot. If you can’t work with a greenscreen, you might try shooting a clean plate (ideally with no traffic or movement in the background) and then, without moving the camera, have the car drive in and the actor walk out of shot.

You could then create a difference matte based on the clean plate. You’d probably have to do some roto to clean it up, but that might be one approach that could work.

If you needed vector lines to work with, after that, you could do and auto trace on the matte.

Lars

[Lars Bunch]

Hi,

Unless the Avids are real old, you should get the Meridian Codec from Avid’s site and render the animations in that format. You MUST make your comp 720×486 (It can be 4×3 or 16×9) at 29.97 fps otherwise it will not render.

This codec supports an alpha channel. Render it out using the “Straight” setting rather than Pre-multiplied. The resulting quicktime will look ugly and wrong, but it will composite properly in the Avid whereas if you render Pre-multiplied, you will get dark ringing in the semi-transparent areas. If you want to check how it composites, bring the rendered animation back into AE and layer it over a test background. It will look right.

If for some reason you can’t use the Meridian codec, render using the “Animation” codec. It will take much longer to import, but some old Avids tend to crash when dealing with Meridian codec files. Again, it should be rendered “Straight”

Hope this helps,

Lars

[Lars Bunch]

Hi Dave,

Every time I answer someone’s question, thinking I know it all, I end up learning something new.

The method you described does the trick and now I don’t have to go through the extra step of downresing to SD before adding the 3/2 pulldown.

Thanks,

Lars

[Lars Bunch]

Hi,

It seems like a perfectly reasonable way to work. I suppose the only issue is making sure the Firestore drive can handle the bandwidth, but if that’s been tested and known to work, then it should be a good way to work without investing in a lot of hardware.

My feeling is that if you want to make sure the footage is more usable as HD in the future, you may want to shoot at 23.98P instead of 29.97P. It’s easy to add the 3/2 pulldown to make it 29.98, but if you shoot at 29.98, you are pretty much stuck with that frame rate as there is no good way to convert it to 23.97.

To a great extent, this is just a personal preference issue, but you might get more value out of your footage if shot at the frame rate that most HD seems to be moving toward. It will mean, however, that you will have to work with different frame rates when you use it in your current SD work. It’s not a big deal for some, but if it is something you don’t want to deal with, then it needs to be considered.

Hope this helps,

Lars

[Lars Bunch]

Hi,

It’s pretty easy in AE. You import your 23.98 footage and then drop it into a 29.97 comp. Render it and all is well. AE will interlace (if that’s what you need) and add the 3/2 pulldown. I’ve been doing a lot of HD compositing in Shake that needs to end up in NTSC 29.97 and my last step is always to convert the HD file sequence in AE.

I have noticed that if your source footage is HD 23.98 and you drop that into an NTSC comp (720×486), re-scaling the HD to fit, AE does not like to add the proper 3/2 pulldown.

Instead, it duplicates one frame for every 4 frames. It achieves the same 29.97 frame rate, but it doesn’t look the same as a proper 3/2 pulldown.

The fix is to render a scaled down version of the HD, but at 23.98 and then bring that back into AE for dropping into a 29.97 comp. Then the pulldown works properly with the interlacing.

Anyway, if you are not re-sizing your source footage, you should be fine with just dropping the 23.98 footage into the 29.97 comp and rendering.

Hopefully if anyone sees something I missed, they will point it out.

Lars

[Lars Bunch]

Hi,

I would guess that the problem is in quicktime since I ran into this problem when looking at conformed files in Shake. Obviously the resolution of the actual file is not changed, but the interpreted size in the file header is changed. Sort of like changing DPI in Photoshop without rescaling the image.

You would think there would be some explicit method for either resizing or NOT resizing files in Cinema Tools. If any Final Cut / Apple experts have comments on this, I’d like to hear them.

Thanks for the tip about using actual size is QT.

Lars

[Lars Bunch]

Hi,

Instead of working in 3D space, would it be possible to set up multiple versions of each image and trim the background version of one image to end just as a foreground version comes in? This would avoid using the processor intensive 3D space and sounds like it might do exactly what you need.

If you had an image that was appearing from under one layer while covering another layer that was actually in front of the first layer, then you could use masks to control what part of each layer was visible.

It also might be helpful to create null objects to use as position controls; one for all the instances of each image.

You would certainly end up with a lot of layers and it might get to be a bit of a headache, but I suspect it would ultimately be easier to deal with than using 3D space.

I hope this helps,

Lars

[Lars Bunch]

Hi,

I’ve never used AE’s tracking function much so I can’t speak for it. In Shake, it is possible to offset a tracker point’s pattern search area so that you can track a point that is onscreen to locate the point that is going off screen. AE may have an offset function like this.

In any case I have used Shake to track panning shots of signs and TV screens as they are panned on or off screen. You set you four tracker points and then as the object approaches the edge of the frame, you offset the points that are going off screen and continue the track. I will usually set the off screen search pattern to the still on screen track points until all points are off screen. As that last points pan off screen, I usually have to manually adjust exactly where they fall.

Another thing to consider. If the actor holds the dummy card with tracking points at an angle that approaches perpendicular to the camera so that the points become harder to distinguish, you may run into serious problems trying to track them. If you can the the card reasonably flat to the camera, then the track should be much easier.

Also if you can shoot in a progressive video format, that will make your life easier. Interlaced video is no friend to trackers.

Hope this helps.

Lars

[Lars Bunch]

Hi,

Have you used Zaxwerks Invigorator? It is a good start of basic 3D in AE.

Alternatively you might want to look into Blender (www.blender.org) since it is a reasonably sophisticated and free 3D animation program.

Maya is very sophisticated, but very hard to learn.

Lightwave seems more intuitive on the surface.

You should go to the various vendor’s web sites and download trial versions to get a feel for how they work.

Lars