Garrett

Need a spot of help with some opengl. Trying to demonstrate a sphere tracking an object in a video on top of that video. The sphere is set up like so:

reset, glmatrixmode projection, glloadidentity, glortho -160. 160. 120. -120. -100. 1., glviewport 0 0 320 240, glmatrixmode modelview, glloadidentity, gltranslate $1 $2 $3, sphere 100

Depth data I have coming in runs from 150 to 1900, 150 is the closest to the camera and 1900 is the furthest away so I'm simply dividing by 15 and multiplying by -1 to get it moving in and out at the moment.

The issue is trying to get the sphere to recede more realistically along with the item it's tracking in the video. I presume it's an issue with the depth in glortho, that I need to feed it a formula that speeds up the closer it gets to -100 but I'm also wondering should I be affecting the size of the sphere or am I missing something important like gluperspective? Suggestions greatly appreciated.

Need a spot of help with some opengl.  Trying to demonstrate a sphere tracking an object in a video on top of that video.  The sphere is set up like so:

The issue is trying to get the sphere to recede more realistically along with the item it's tracking in the video.  I presume it's an issue with the depth in glortho, that I need to feed it a formula that speeds up the closer it gets to -100 but I'm also wondering should I be affecting the size of the sphere or am I missing something important like gluperspective?  Suggestions greatly appreciated.

glortho-depth

might help with the speeding up when closer to the camera, with a bit of jigging it might be useful

hi, thanks but could you elaborate? nothing in that patch is opengl based, all the calculation is done in Javascript, several functions in there and not very well commented so I'm not too sure which I should even be looking at.

i was thinking that you would put the depth data in to the tween, so that the tween would accelerate the depth data when it is near the closest

Ok managing to make some sense of it and the Quadratic formula I pulled out of Masato's patch seems to be close to working (the basic guts of it are below) but I'm having to add offsets to scale it up or down (in my case down as I'm working with ints). I'm just a bit uncomfortable that some of it seems to be guess work - surely there is some formula where if the perspective used and size of object is known (or something like that) then distance and rate of recession can be accurately calculated.

I'm not a mathematician, pages I'm coming across on search engines are daunting but surely someone has done something like this before in max a rigorous way. Anyone?

Ok managing to make some sense of it and the Quadratic formula I pulled out of Masato's patch seems to be close to working (the basic guts of it are below) but I'm having to add offsets to scale it up or down (in my case down as I'm working with ints).  I'm just a bit uncomfortable that some of it seems to be guess work - surely there is some formula where if the perspective used and size of object is known (or something like that) then distance and rate of recession can be accurately calculated.

I'm not a mathematician, pages I'm coming across on search engines are daunting but surely someone has done something like this before in max a rigorous way.  Anyone?

glortho depth