I am not sure if this is a right place to post, if not point me to the correct place please
it appears that default GFX library function drawChar updates one pixel at the time, setting window for each pixel individulally in drawPixel routine.
this causes a lot of overhead, as it is setting "window" 48 times instead of 1 for each character (size 1)
Set Window takes 7 SPI writes for each pixel, that we can avoid
Overhead is even greater for the font sizes 2 or larger.
I think it would be more efficient to set window for whole character once (6x8 size) and then push all pixels directly writing spidata
Hardware will take care of shifting to the next pixel and setting next row in this case.
But this will require to read all five bytes of each character before displaying and as a result 6 bytes of RAM instead of 1 for the "line" variable.
I did not write whole new routine, just made quick test for font size 1 and it works much faster, making text appearance almost instant.
Any comments or ideas?

