in my program i’m using invisible (alpha=0) dummies for all my physics/colision test. Now it’s a PNG image with alpha 0, would it be better for CPU/GPU power to use display.newRect or display.newCircle?
Since they are created as vector objects, don’t they use more CPU/GPU power? [import]uid: 44010 topic_id: 7979 reply_id: 307979[/import]
How many vectors are you displaying on the screen at the time?
Corona is OpenGL based. So the GPU is used more than the CPU.
But for performance? The overhead of a push to the OGL stack with four vector points far out weights a push to the OGL stack with 75,000 vectors.
C. [import]uid: 24 topic_id: 7979 reply_id: 28413[/import]
Well, let’s say i have 100 invisibles rectangles on my scene for testing physics.
Now they are an image object, all from the same image.png
So, would it be better for performance and memory usage to convert them to a square object? [import]uid: 44010 topic_id: 7979 reply_id: 28472[/import]
for 100 elements? don’t worry about it.
In Corona, the image is being re-used in memory. we don’t load 100 of them into memory. we load one. And a move x,y is just as fast since the image is already in texture buffer memory.
c. [import]uid: 24 topic_id: 7979 reply_id: 28678[/import]