Fix SSD1322 NHD initialization and encode each pixel as 4bit+4bit identical nibbles

[georgeharker]

Each pixel should be encoded as two adjacent nibbles (2 nibbles per pixel with identical values) corresponding with the 2 SEG + 1 COM design of the NHD based panel, which uses the two SEG per pixel to drive the charge (hence needing two adjacent pixel values in GDRAM to drive them.

Closes #353, #199

[georgeharker]

Ah. Clearly there’s some tests I need to update

[rm-hull]

Code changes look reasonable, but the regression test data needs regenerating

[thijstriemstra]

the regression test data needs regenerating

@georgeharker this is the code that @rm-hull is referring to: https://github.com/georgeharker/luma.oled/blob/master/tests/test_ssd1322_nhd.py#L103

[georgeharker]

the regression test data needs regenerating

@georgeharker this is the code that @rm-hull is referring to: https://github.com/georgeharker/luma.oled/blob/master/tests/test_ssd1322_nhd.py#L103

My read after looking thru the tests was that I needed to uncomment the regen line, run the regen and comment it back out and submit the test data. The PR now has the regenerated test data. But perhaps I missed something. It didn’t look like the test code itself needed changing as it passed after a regen of the data.

George

[thijstriemstra]

👍🏻