Video help...

[Anonymous]

....How can i make a video "montage" (sorry about my english) using my video camera...or renting an helicopter to film the area..of the new building and them introduce my 3d model in the video???...What softwares do i need to use??

can you help me? I would appreciate very much.

thanks and regards.

[Anonymous]

Retirado wrote:
....using my video camera...or renting an helicopter to film the area..of the new building and them introduce my 3d model in the video???...What softwares do i need to use??

Like Matchmover? Good luck, this technology is way out there in cost and complication.

http://www.realviz.com/gallery/list.php?product=mpro



When you finish let's see the result.


psst hey, wanna see my video? Look here in my pocket:
http://www.ehomeupgrade.com/archives/000312.php

[Karl Ottenstein]

Matchmover sounds interesting, Mark. Thanks for the post. I'm skeptical about the ability to extract the exact camera sequence just because we know how unreliable "align view" is ... same idea.

In one of the Lord of the Rings special edition DVD's (after 20 hours of footage, can I possibly remember which one?!), they walked through how they blended a helicopter shot of a hill with a rendered town (can't remember which/where). Even with all of their considerable skills, the buildings were vibrating a little bit around the hillside and so they ended up modeling the entire hill to cover it up. It makes you wonder how much of the surroundings of a building might have to be modeled to create a smooth video montage...

Retirado, it sounds like with this MatchMover you then need a rendering package that can import a camera path etc ... and a video editing program to composite the two resulting videos. Most upper end ($500+) video editing software includes masking ... like "blue masking" in the movies (of course you can use any color). Place the original video in the A track, the computer rendered video - with solid color background on the B track, pick up the mask color, apply filters to adjust exposure for proper blending, and render.

I'd stick with still images! 😉 But, hey, if you've got the budget, this would be a heck of a lot of fun! 😉

Karl

[Anonymous]

we have on hower studio MatchMover and Boujou

but for can work wit that type of softwer you need to trace a camera wit noncompresed videos wit DIGITAL BETACAM camera grab Videos wit VIDEO TOSTER is not easy wit DIGITAL PAL V1 is INPOSIBLE all pixel moove you have the fields what you can't get off and that is the problem

other way is if you have video from scan that is other HI END thing on that type of quality of VIDEO material MatchMover and Boujou work fine

[stefan]

If you have max, a camera-matching utility is included.

And there was a free (for non-commercial use) tracker called Icarus, which is now commercial...

And there is Syntheyes which is a lowcost tracking software.

I'm not saying it's easy, but it's not impossible either.

Remember that slight movements of the footage is not too bad, when you're not the FX-team from Weta.

And for video-editing, tools like Combustion or After Effects will help a lot, but it's possible to do a lot of this with simple tools as well. It does help when you have some "real" 3D-animation software available, though.

---

Check this for an example:
http://caad.asro.kuleuven.ac.be/downloads/icarus.avi
When this link fails, download the ZIP-file in attachment. It's an AVI using the DivX compression.

How is this done? DV-footage (like the one from a consumer DV-cam) is brought in and in three steps, the software does some nice work, provided you help a bit: you point out straight lines, so distortion can be reverse-engineered. Then the movement is tracked and (optionally) you can add geometry that receives the textures.

The camera movement was exported as a Maxscript file, that generates an animated camera in 3ds max. Tracking points were also exported.

In 3ds max, the additional objects were added.

To have the CG-objects cast shadows on the video footage, some "shadow-catching" objects have been made, with a material that is invisible for a render, but does catch shadows.
Another solution was adding catching geometry and mapping it with the video footage, so it looks identical to the video, but is able to catch shadows and obscure stuff that lies after them.

Not a regular architectural visualisation job 😉