I am seriously interested in upgrading to V3, primarily for the 3-D camera tracker. Apparently, I must upgrade to the current version by 15 April to get V3 at the discounted price ($149). However, there is no information I can find that illustrates how the camera tracker works and what its advantages are over a conventional 3-D point tracker. I need to know more before I invest. Is there something I am missing?
You will be able to see example videos on our site, www.imagineersystems.com (Boris FX | Videos), in the next two weeks, as we plan to release the software soon. I don’t know if they will have downloadable examples though. We’ll see!
>From using the solver for the past several months, I do not think mocha’s camera solver has limited capabilities so much as different capabilities. What I was trying to say is that comparing it to a conventional camera tracker is like comparing our planar tracker to a point tracker, it is just not the same thing. We are meant to do different tasks and have different outputs, we are not trying to completely replace high-end camera tracking tools in your toolbox so much as augment your toolbox with fast, easy ways to plug accurate 3D objects into your scene that is based on our planar data. Some shots will still need a conventional camera tracker. But 70% or 80% should be able to be solved in five or ten minutes with mocha.
Now, that being said, you can use our mocha planar data to get a better camera track with Bouju or SynthEyes on a difficult shot that neither of those programs can track well on their own by inserting new planes with better detail OVER bad detail in your shot with mocha and then using your conventional camera tracker to solve the shot for the exact camera with the new data. We have people using this method right now even without our 3D data, just by inserting 2D corner pins into the scene.
Because our tracker does not care about blur and feature detail and hangs on through a great many tracking problems, we can lock on where nothing else can. This gives mocha a HUGE advantage over other trackers.
I would say that if you are unsure that you might wait until mocha V3 comes out, download a trial, and decide for yourself. It takes a little bit of a curve to get the hang of the tracker, but we should have some instructional videos up by the time the tracker is released.
I think y’all will find it interesting!
See also: First look: Imagineer’s mocha v3 – fxguide
This is a new article that goes over what is new in V3. You might find it helpful.
mocha’s 3D Camera Solver is not like a conventional camera tracker at all, and that’s why it is interesting. It does not solve for the camera so much as it solves for our planar tracking data in 3D space, essentially converting 2.5D planes into 3D planes. It does not generate a point cloud, but instead relies on our planar tracking technology in order to solve one of many camera solves within the scene.
You can use it to track data quickly and easily where other camera trackers would fail, as long as you find good planes to track you should be fine (one plane is needed for pan/tilt/zoom solves or more for small and large parallax change solves).
It is important to note that we are not generating a point cloud. We are solving for the planes mocha has been able to track for years, and you attach 3D objects into the scene by aligning them to said planes.
Much like our un-conventional planar tracker is unlike point or feature trackers, mocha’s camera solver is unlike other camera solvers out there as well.
>From the early review on the web I understand that the 3D solver will have limited capabilities, but as Mary said it could be used where the conventional point tracking would fail (for example when tracking planar surface or not much paralax going on in the scene).
In any case it will be interesting to see an example.
Is there any downloadable output files (for AE or FBX) to look at?