Multiple Instances of Motion Tracking???

[Steve Roberts]

I wouldn’t do it unless I was paid by the hour.

Think about it:
– how many times over a minute is a track likely to go off and require manual repositioning?
– how many students are in the class?
– how long is the class?

Now, the students aren’t likely to move that much if it’s a lecture (ZZZ), so the tracks might stick.

But I don’t know of any way to track multiple points at the same time, other than using perspective corner pin tracking to do four at once. Then using expressions, you could tie four dots to the tracks.
I don’t see any limit to the length of the tracks, but tracking for 30 minutes is just silly.

Since it sounds as if the prof is thinking out loud, tell him it’ll be a real PITA, and he’ll probably be okay with it.

Anybody else?

[Zygotesoup]

I tried telling him how much of a nightmare this is going to be and this was his response:

“Nathan,

Thanks for your input. I want to follow up a bit on this, though

[Aaron Zander]

well, heres what you would do, if you were A) suicidal B) getting payed enough

1) go to workspaces->motion tracking
2) select the footage
3) add your first motion tracker, and track
4) once done, apply track to a null or solid, or some way to save the data in a more organized fashion.
5) go to the tracker controls and add a ‘new’ track. you can do this is several ways, by using the tracker drop down menu on the panel, or simply hitting new.
6) repeat step 3-4 and repeat 5-6 as many times as needed

this will take a LONG time

[Aaron Zander]

[zygotesoup] “Not to mention, this is not a class of college students sitting in their desks, it is a class full of kintergardners running around all over the place.
“

alright, I’m so against wanting to do this, I might not even post, so you wont do it….

no seriously if he is willing to shell out doughfor something like mocha (or the cheaper mocha ae)you could set up all your tracks at once and do this like that, expect about 4-5 hours of track time minimum. (probably alot more)

when we say loose the track, if it’s ‘lost’ it’s rare it will come back into tracking.

when we say lost we mean this- when tracking a tracker is trying to repeatedly find a set of pixels that match a defined range of luma/chroma/ values in a similar pattern to the frame before. with fast moving objects tracks are lost much easier as the subject is subjected to motion blur (excuse the pun) as well as you than have to bump the defined search range up to a much higher size creating more lag and slow down as the program trys to keep up with searching from a set of maybe 64 pixels, to 256 pixels.

hope this makes sense

-Zander

[Darby Edelen]

[zygotesoup] “I

[Zygotesoup]

Thanks to everyone for their input. I forwarded most of your responses to my manager and CIO and they are contacting the professor to pull me away from this project (thank god!… i mean thank creative cow!!!)

Thanks again for all the input guys!

[Steve Roberts]

I think the scary part is the “once done” part. 🙂 When we’re talking about tracks going off, it could be over frames, seconds. It all depends on how the kids move around.

You should know that this is definitely not a job where the CPU crunches away and you go off and have a coffee or paint the house. You have to sit and watch the whole time, for each track, to see if it goes off. In **slower** than real time. Got it?

Really, if I saw 30 kids tracked on a prime time spot, I’d say “wow, they’ve got budget”.

If the prof has budget, you should look into Mocha or Mocha AE. Ask the Imagineer folks the same things you’ve asked us, as they know the software.

But yep, it’s a good idea to track a few kids, mild and wild, and use that to extrapolate how long it’ll take.

Here’s another solution: get a Wacom tablet, and follow each kid’s head with the pen using the write-on effect in real time or a little slower. Maybe it’ll be faster. Who knows?

[Aaron Zander]

don’t get me wrong there are a lot of troublesome jobs that just don’t seem like doing.

thats not to say this one was better or worse

but this is an exemplar case of something that is doable but just would be so intensive, it wouldn’t be worth doing and wasting hours upon hours of time on.

[Aaron Zander]

yea mocha ae is like 300 bucks, and would do this job like it was cake…it would still take a huuuuge amount of time don’t get me wrong, but it wouldn’t be as horrid. I would say what steve said, about tracking, that you can’t go and let it track, doesn’t work like that.

I’ve left mocha tracking and got lunch, thats how much i trust it…

[Rhett Robinson]

Okay, now I’m interested, although it looks like you escaped from having to do this. Is this related to a 1-time study of children’s movements? If they were willing to have dots on the kid’s foreheads, what about “special hats” that would be easier to track when the kids turn their heads?

Although I agree that a motion track that’s good enough to replace each child’s face would be a ridiculous task, one that is simply good enough to show their location or movement in a room would be pretty easy. Yes, you’d have to watch and adjust each track separately, but I get the feeling that this is not the kind of job that require commercial level precision, so you could burn through that pretty quickly, probably just a few minutes for each kid, depending on camera angle and where they are able to hide in the room.

Actually, I’m going to test this weekend, if I get a chance – I get the idea that AE’s built-in tracker is plenty adequate for this, just with a wider search range…