Recently, I\'ve been trying to set up a CIColorCube on a CIImage to create a custom effect. Here\'s what I have now:
uint8_t color_cube_data[8*4] = {
0,
The accepted answer is incorrect. While the cube data is indeed supposed to be scaled to [0 .. 1]
, it's supposed to be float
, not int
.
float color_cube_data[8*4] = {
0.0, 0.0, 0.0, 1.0,
1.0, 0.0, 0.0, 1.0,
0.0, 1.0, 0.0, 1.0,
1.0, 1.0, 0.0, 1.0,
0.0, 0.0, 1.0, 1.0,
1.0, 0.0, 1.0, 1.0,
0.0, 1.0, 1.0, 1.0,
1.0, 1.0, 1.0, 1.0
};
(Technically, you don't have to put the ".0" on each number, the compiler knows how to handle it.)
I found the issue... I have updated my question if anyone has the same problem!
The input float array had to be pre-divided out of 255.
The original used 255:
uint8_t color_cube_data[8*4] = {
0, 0, 0, 1,
255, 0, 0, 1,
0, 255, 0, 1,
255, 255, 0, 1,
0, 0, 255, 1,
255, 0, 255, 1,
0, 255, 255, 1,
255, 255, 255, 1
};
It should look like this instead:
uint8_t color_cube_data[8*4] = {
0, 0, 0, 1,
1, 0, 0, 1,
0, 1, 0, 1,
1, 1, 0, 1,
0, 0, 1, 1,
1, 0, 1, 1,
0, 1, 1, 1,
1, 1, 1, 1
};
Your problem is that you are using value 1(which is next to zero) for alpha channel, max for uint8_t is 255
See example below:
CIFilter *cubeHeatmapLookupFilter = [CIFilter filterWithName:@"CIColorCube"];
int dimension = 4; // Must be power of 2, max of 128
int cubeDataSize = 4 * dimension * dimension * dimension;
unsigned char cubeDataBytes[cubeDataSize];
//cubeDataBytes[cubeDataSize]
unsigned char cubeDataBytes[4*4*4*4] = {
0, 0, 0, 0,
255, 0, 0, 170,
255, 250, 0, 200,
255, 255, 255, 255
};
NSData *cube_data = [NSData dataWithBytes:cubeDataBytes length:(cubeDataSize*sizeof(char))];
//applying
[cubeHeatmapLookupFilter setValue:myImage forKey:@"inputImage"];
[cubeHeatmapLookupFilter setValue:cube_data forKey:@"inputCubeData"];
[cubeHeatmapLookupFilter setValue:@(dimension) forKey:@"inputCubeDimension"];
This is link to full project https://github.com/knerush/heatMap