family: adx
equalitygroup:
bitdepth: 10ui
+ description: |
+ Film Scan, using the 10-bit Academy Density Encoding
+ isdata: false
+ allocation: uniform
+ allocationvars: [0, 1]
+ to_reference: !<GroupTransform>
+ children:
+ - !<FileTransform> {src: adx_adx10_to_cdd.spimtx}
+ - !<FileTransform> {src: adx_cdd_to_cid.spimtx}
+ - !<FileTransform> {src: adx_cid_to_rle.spi1d, interpolation: linear}
+ - !<LogTransform> {base: 10, direction: inverse}
+ - !<FileTransform> {src: adx_exp_to_aces.spimtx}
+
+ - !<ColorSpace>
+ name: adx16
+ family: adx
+ equalitygroup:
+ bitdepth: 16ui
+ description: |
+ Film Scan, using the 16-bit Academy Density Encoding
isdata: false
allocation: uniform
allocationvars: [0, 1]
to_reference: !<GroupTransform>
children:
- - !<FileTransform> {src: channel_dependent_offset.spimtx, interpolation: linear}
- - !<FileTransform> {src: ten_bit_scale.spimtx, interpolation: linear}
- - !<FileTransform> {src: channel_dependent_scalar.spimtx, interpolation: linear}
- - !<FileTransform> {src: cdd_to_cid.spimtx, interpolation: linear}
- - !<FileTransform> {src: rel_log_exp_scalar.spimtx, interpolation: linear}
- - !<FileTransform> {src: refpt.spimtx, interpolation: linear}
+ - !<FileTransform> {src: adx_adx16_to_cdd.spimtx}
+ - !<FileTransform> {src: adx_cdd_to_cid.spimtx}
+ - !<FileTransform> {src: adx_cid_to_rle.spi1d, interpolation: linear}
- !<LogTransform> {base: 10, direction: inverse}
- - !<FileTransform> {src: exp_to_aces.spimtx, interpolation: linear}
+ - !<FileTransform> {src: adx_exp_to_aces.spimtx}
- !<ColorSpace>
- name: slog10
+ name: slog
family: slog
equalitygroup:
bitdepth: 10ui
--- /dev/null
+#!/usr/bin/env python
+
+import math, numpy
+
+"""
+
+const float REF_PT = (7120.0 - 1520.0) / 8000.0 * (100.0 / 55.0) - log10(0.18);
+
+const float LUT_1D[11][2] = {
+ {-0.190000000000000, -6.000000000000000},
+ { 0.010000000000000, -2.721718645000000},
+ { 0.028000000000000, -2.521718645000000},
+ { 0.054000000000000, -2.321718645000000},
+ { 0.095000000000000, -2.121718645000000},
+ { 0.145000000000000, -1.921718645000000},
+ { 0.220000000000000, -1.721718645000000},
+ { 0.300000000000000, -1.521718645000000},
+ { 0.400000000000000, -1.321718645000000},
+ { 0.500000000000000, -1.121718645000000},
+ { 0.600000000000000, -0.926545676714876}
+};
+
+ // Convert Channel Independent Density values to Relative Log Exposure values
+ float logE[3];
+ if ( cid[0] <= 0.6) logE[0] = interpolate1D( LUT_1D, cid[0]);
+ if ( cid[1] <= 0.6) logE[1] = interpolate1D( LUT_1D, cid[1]);
+ if ( cid[2] <= 0.6) logE[2] = interpolate1D( LUT_1D, cid[2]);
+
+ if ( cid[0] > 0.6) logE[0] = ( 100.0 / 55.0) * cid[0] - REF_PT;
+ if ( cid[1] > 0.6) logE[1] = ( 100.0 / 55.0) * cid[1] - REF_PT;
+ if ( cid[2] > 0.6) logE[2] = ( 100.0 / 55.0) * cid[2] - REF_PT;
+"""
+
+
+def interpolate1D(x, xp, fp):
+ return numpy.interp(x, xp, fp)
+
+LUT_1D_xp = [-0.190000000000000,
+ 0.010000000000000,
+ 0.028000000000000,
+ 0.054000000000000,
+ 0.095000000000000,
+ 0.145000000000000,
+ 0.220000000000000,
+ 0.300000000000000,
+ 0.400000000000000,
+ 0.500000000000000,
+ 0.600000000000000]
+
+LUT_1D_fp = [-6.000000000000000,
+ -2.721718645000000,
+ -2.521718645000000,
+ -2.321718645000000,
+ -2.121718645000000,
+ -1.921718645000000,
+ -1.721718645000000,
+ -1.521718645000000,
+ -1.321718645000000,
+ -1.121718645000000,
+ -0.926545676714876]
+
+REF_PT = (7120.0 - 1520.0) / 8000.0 * (100.0 / 55.0) - math.log(0.18, 10.0)
+
+def cid_to_rle(x):
+ if x <= 0.6:
+ return interpolate1D(x, LUT_1D_xp, LUT_1D_fp)
+ return (100.0 / 55.0) * x - REF_PT
+
+def WriteSPI1D(filename, fromMin, fromMax, data):
+ f = file(filename,'w')
+ f.write("Version 1\n")
+ f.write("From %s %s\n" % (fromMin, fromMax))
+ f.write("Length %d\n" % len(data))
+ f.write("Components 1\n")
+ f.write("{\n")
+ for value in data:
+ f.write(" %s\n" % value)
+ f.write("}\n")
+ f.close()
+
+def Fit(value, fromMin, fromMax, toMin, toMax):
+ if fromMin == fromMax:
+ raise ValueError("fromMin == fromMax")
+ return (value - fromMin) / (fromMax - fromMin) * (toMax - toMin) + toMin
+
+NUM_SAMPLES = 2**12
+RANGE = (-0.19, 3.0)
+data = []
+for i in xrange(NUM_SAMPLES):
+ x = i/(NUM_SAMPLES-1.0)
+ x = Fit(x, 0.0, 1.0, RANGE[0], RANGE[1])
+ data.append(cid_to_rle(x))
+
+WriteSPI1D('adx_cid_to_rle.spi1d', RANGE[0], RANGE[1], data)
+
projects / OpenColorIO-Configs.git / commitdiff