An annoyingly sleepless night. I went to bed at 21:30 but woke from 03:30 to about 07:00, then slept until nearly 10, the worst sort of night!
I started with Radioactive programming, studying Andrew's feedback from his latest version. There were two problems. Firstly, far too many screenmodes were being polled. The Enumerator routine was, for some reason, continuing to process after a CANCEL request, so I did things differently and only stored data when screenmodes were under the maximum. Secondly, the actual setting up of the 3D device was wrong. This is because this ancient game was written in DirectX5 or 6, and the older code wasn't updated correctly. My code said:
if (lpD3->CreateDevice(IID_D3DDevice, lpDDSbackbuffer, &lpD3Ddevice)==D3D_OK)
But the IID_D3DDevice define was an error, it should have been either IID_IDirect3DTnLHalDevice, IID_IDirect3DHALDevice, or IID_IDirect3DRGBDevice. This is obvious from the documentation BUT the documentation isn't online anywhere. Due to my meticulous filing, I remembered that I have the original DirectX7 documentation as a chm file, and discovered this error when checking the Enumeration routines. Anyway, this was quickly fixed and I was confident enough to update the game before Andrew could test it. I was right, and v1.15 is now error free, perhaps for the first time.
After that, work on colour studies for Meals of Warm Spring. This is difficult because almost any colour will suffice regarding 'reality', the pictorial reality of the mind, yet colour is crucial for conveying the mood via a sense of place as well as the inherent feeling of chromanance.
I started with a classical approach; colouring the flesh warm and the background cool, blues or greens. The result was rather pretty, warm, springlike, and something like the colours in my painting Romeo and Juliet.
Too happy though, too twee, too wrong! I painted an alternative with a red, foreboding sky and it looked instantly more powerful. I was hesitant to do this first for fear of overwhelming the whole image with reds and oranges; but there are always points of contrast. The lines in the sky (teeth, stictches, ribbons...) the light ones are light green (not evident from the insult of the digital imagery you can see, only in real life are the colours of any oil painting fully visible).
The area behind the crucifixion is heavenly. This is Jesus, Caravaggio's Jesus. The moon, originally yellow is now a light blue, brilliantly icy against the turbulent sky. The left will feature more green. There is vastly more drama in the second image, yet the only difference is hue. This is why colour studies are important.
You'll note the flesh is the same. I simply painted the new parts on a new piece of cardboard.
Sketchy, loose, and rough this may be but it still took a few hours. I'm ready to begin the painting, though I still feel somewhat sniffly, congested and sapped of energy. I'm not ready to start tomorrow, but will probably do so on Monday. I'm unsure whether to use Michael Harding's quick drying white. That might give me chance at a second layer before the deadline at the end of the month, but it probably yellows faster than the other white, and Michael Harding's paints yellow more than most anyway. When his colours are exhausted I'll be using Blockx and Winsor and Newton exclusively.
