I started coding some basic coding on Fez Spider. I get some data out of SPI / UART module and transmit it through UDP. Alongside, I would like to have a visualization of data I am transmitting on the 4.3" Touch+LCD module. I initially coded the prototype using WPF and quickly realized that the performance was nowhere near where it needed to be.
I coded a quick prototype using RLP and demonstrated that I can fill memory buffer in the size needed quick enough to keep up with the data capture. However, the moment I push the byte array back out to managed code, turn it into Bitmap and flush it, the perf is completely blown again.
So, my logical conclusion it that I need to directly fill the LCM pixel buffer from my RLP code. There are hints elsewhere on the boards that this has been done successfully. But I cannot find documentation that shows the memory map to Fez Spider so I can tell where in the memory to dump the bits. Can someone point me to the doc?
My assumption is that once I have the starting address, rendering is a simple matter of writing a 16bit 5-6-5 coded pixel data onto the starting address + x*y offset to set pixel color.
If there is a better way to gain higher perf, I am open for suggestion.