PDA

View Full Version : The way mv_x and _y work

Krash
9th March 2002, 12:29
After generating the mobile motionvector template (which has since been fiddle with and used to great effect by Rovastar), I've been thinking the way that the vectors work might be changed.
Currently, when the number of motion vectors is between 0 and 1, you have no vectors at all. Once the value hits 1, you get a single vector. Greater than one, and the vector moves to the left (or down, if you're talking about mv_y), until it gets to 2, and the second vector appears.
Using this method, it is impossible to get a single vector anywhere in quadrants 1, 2, or 4 (if you don't know what I mean by that, don't worry).
I suggest a different way of doing it.
mv_x = 0 - no vectors.
mv_x = 0.001 - one vector, on the extreme right (or left) of the screen
mv_x = 0.001 to 0.5 - vector moves across the screen, until it reaches the center at 0.5
mv_x = 0.501 to 1 - vector continues moving to get to other side of the screen.
mv_x = 1.001 - second vector appears at opposite side of screen
mv_x = 1.001 to 2 - two vectors converge on the center of the screen, until at mv_x = 2, it looks the same as mv_x = 0.5.
mv_x = 2.001 - the two vectors at the center become a single vector, and we get two more vectors at the edges, which converge.
mv_x = 3 - the two vectors have reached the center - looks like just one vector again.
mv_x = 3.001 - two more vectors appear at the edges, and begin converging. The single vector at the center separates, so now we have four vectors. At mv_x = 3.5, the vectors are evenly spaced, and at mv_x = 4, the two pairs on either side have reached each other, so it looks like just two, evenly spaced.

The pattern would continue on.
I hope what I've said makes sense - it would work, and I've followed the logic through pretty well, but I wouldn't want to try to code it.

- Krash

Rovastar
9th March 2002, 14:22
That all looks good esp for the below 1 part. But the problem I have is that all the rest of my preset 'might' look wrong or at least not as intended if all this was to change.. This would be espally true where the vectors have lots of movement.

I suggest having a flag for the motion vector mode. Old style and New style. And by default alway set this to the old style for the prevoius presets.

Also maybe have negative vector numbers where it does the opposite or mirror image of what the positive does. Like the mv_l variable where negative length goes in the opposite direction.

Krash
10th March 2002, 01:34
You could just as easily add 1 to all of those values, it would still work.
And all the code that I did and you modified for the moving vectors would still work - it's just that you could actually have the center of the vector "matrix" up on the top right side of the screen, instead of always in the lower left.

It might work - Ryan would have to try it. And as usual, he's a bit of an enigma between versions, so we don't really know if he's paying attention or not ;)

- Krash

Rovastar
10th March 2002, 14:50
Originally posted by Krash
And as usual, he's a bit of an enigma between versions, so we don't really know if he's paying attention or not ;)

True too true........

Ryan,.....Ryan,.....Ryan,...Ryan..
Echo,.....Echo,.....Echo,...Echo..

ryan
27th May 2002, 01:49
After generating the mobile motionvector template (which has since been fiddle with and used to great effect by Rovastar)

Can someone post the template for me ?

Krash
29th May 2002, 08:03
I don't have it anymore...
I can't even remember which thread it was in... Rova?

- Krash

Rovastar
29th May 2002, 11:52
:) I left this thread because I though you would know Krash.

ummmh........The code is still active in any of my Vectrip mixes of Ryan's presets.

Try something like surface and reduce the default mv_x and mv_y values to 1.25 (to centre it in v1.03b3)

ryan
29th May 2002, 22:22
Thanks.. If I cant find it i will go back through older threads and look throught there.