-
Notifications
You must be signed in to change notification settings - Fork 14
Description
Hi again,
I'm working on bringing a big library of APIs on top of your work (to allow developing on PCs for faster speed and debugging).
I've already done the work with a fork of your lib that uses SDL, but it's an old fork of your work that is lacking serial support plus a lot of code you added, and I'd rather have is depend on X11 instead of SDL.
The SDL version does work well and fast, though, it writes to an emulated neopixel 1D array that gets rendered into a 2D SDL matrix: https://github.com/MarcFork/FastLEDonPc
My concern at this point is that all my code relies on a framebuffer that is stored in memory and synced to the hardware backend when matrix->show() is called.
The only way I knew how to implement this is by calling XWindow::drawPoint_RGB for every dot, in the matrix, but it is very slow (0.2sec for only 160x120 resolution). See:
https://github.com/marcmerlin/FastLED_TFTWrapper_GFX/blob/2a2de242054f227f603156139b5bbeed6197193a/FastLED_TFTWrapper_GFX.cpp#L28
Is there a better way to give a precomputed square matrix in RGB888 to Xwindow without calling the slow draw pixel, pixel by pixel?
Thanks.