[PD] infinite zoomer feedback
Peter Todd
peter_todd82 at yahoo.co.uk
Wed Oct 20 12:53:45 CEST 2004
> ive tried putting the "separator" object. I think maybe you dont
> understood my question
> I know how to rotate or translate the image without affecting its
> structure using separator objects. What im trying to do is to affect
> the
> structure, but not the whole at the same time)all the balls). what i
> want to do is for example : when i change translate or rotate, begin
> affecting the balls that are far away no the ones that are near of the
> screen. So all the structure will change in a "progressive" way.
>
> many thanks
> siggmus
>
That may sound simple logically, but with the current structure of the
program, it doesn't really work so easily. It's important you realise that
all of the positions are calculated pretty much fresh every frame. In fact,
I can't stress enough how useful it is to thoroughly understand which
calculations are computed in which order, and with the results of which
other ones etc. The structure (roughly what I implemented) in the patch
you're working on is a fairly specific case, I think, and maybe not a good
one to work from.
Assuming you carry on using my algorithm, you could perhaps try something
nifty with tables... as the function iterates and the index increases, you
read different values for [rotate] / [traslate] for each iteration. Off the
top of my head, you should increment a write pointer once per frame, and use
the index for the read pointer. I think. That would take a bit of
programming, but could achieve the desired result, and by the time you
manage it you'd have a far greater understanding.
As for moving the camera, you need to use view messages to the gemwin,
(which involves computing what the messages should be).
p.s. you do realize that the 90 in [pd fmod 90] has no effect on the
program - it was misleading of me to post a patch where it looked like that
was an argument for substition. If you look inside, too, there is no $1, so
nowhere for that number to be plugged (not that this would happen in a
subpatch, anyway). Frank just posted on this, but I don't know if you read
that / clicked. This is why you had overlapping elements, rather than one
long line.
More information about the Pd-list
mailing list