curioustorvald/tsvm
1
0
Fork 0
mirror of https://github.com/curioustorvald/tsvm.git synced 2026-06-09 06:34:04 +09:00
Code Issues Packages Projects Releases Wiki Activity

ICtCp colour space impl

Browse Source
This commit is contained in:
minjaesong
2025-09-15 09:52:23 +09:00
parent 9c2aa96b73
commit 34cf5cb591
3 changed files with 39 additions and 31 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
assets/disk0/tvdos/bin/playtev.js
View File

@@ -589,7 +589,7 @@ let frameDuped = false
// Main decoding loop - simplified for performance // Main decoding loop - simplified for performance
try { try {
let t1 = sys.nanoTime() let t1 = sys.nanoTime()
while (!stopPlay && seqread.getReadCount() < FILE_LENGTH && trueFrameCount < totalFrames) { while (!stopPlay && seqread.getReadCount() < FILE_LENGTH /*&& trueFrameCount < totalFrames*/) {
// Handle interactive controls // Handle interactive controls
if (interactive) { if (interactive) {

28
tsvm_core/src/net/torvald/tsvm/GraphicsJSR223Delegate.kt
View File

@@ -2171,9 +2171,9 @@ class GraphicsJSR223Delegate(private val vm: VM) {
val Cp = (cp / 255.0) val Cp = (cp / 255.0)
// ICtCp -> L'M'S' (inverse matrix) // ICtCp -> L'M'S' (inverse matrix)
val Lp = (I + 0.015718580108730416 * Ct + 0.2095810681164055 * Cp).coerceIn(0.0, 1.0) val Lp = I + 0.015718580108730416 * Ct + 0.2095810681164055 * Cp
val Mp = (I - 0.015718580108730416 * Ct - 0.20958106811640548 * Cp).coerceIn(0.0, 1.0) val Mp = I - 0.015718580108730416 * Ct - 0.20958106811640548 * Cp
val Sp = (I + 1.0212710798422344 * Ct - 0.6052744909924316 * Cp).coerceIn(0.0, 1.0) val Sp = I + 1.0212710798422344 * Ct - 0.6052744909924316 * Cp
// HLG decode: L'M'S' -> linear LMS // HLG decode: L'M'S' -> linear LMS
val L = HLG_inverse_OETF(Lp) val L = HLG_inverse_OETF(Lp)
@@ -2181,9 +2181,9 @@ class GraphicsJSR223Delegate(private val vm: VM) {
val S = HLG_inverse_OETF(Sp) val S = HLG_inverse_OETF(Sp)
// LMS -> linear sRGB (inverse matrix) // LMS -> linear sRGB (inverse matrix)
val rLin = 29.601046511687 * L - 21.364325340529906 * M - 4.886500015143518 * S val rLin = 3.436606694333079 * L -2.5064521186562705 * M + 0.06984542432319149 * S
val gLin = -12.083229161592032 * L + 10.673933874098694 * M + 1.5369143265611211 * S val gLin = -0.7913295555989289 * L + 1.983600451792291 * M -0.192270896193362 * S
val bLin = 0.38562844776642574 * L - 0.6536244436141302 * M + 1.0968381245163787 * S val bLin = -0.025949899690592665 * L -0.09891371471172647 * M + 1.1248636144023192 * S
// Gamma encode to sRGB // Gamma encode to sRGB
val rSrgb = srgbUnlinearize(rLin) val rSrgb = srgbUnlinearize(rLin)
@@ -2648,7 +2648,9 @@ class GraphicsJSR223Delegate(private val vm: VM) {
// PASS 2: Apply proper knusperli boundary optimization (Google's algorithm) // PASS 2: Apply proper knusperli boundary optimization (Google's algorithm)
val (optimizedYBlocks, optimizedCoBlocks, optimizedCgBlocks) = applyKnusperliOptimization( val (optimizedYBlocks, optimizedCoBlocks, optimizedCgBlocks) = applyKnusperliOptimization(
yBlocks, coBlocks, cgBlocks, yBlocks, coBlocks, cgBlocks,
QUANT_TABLE_Y, QUANT_TABLE_C, QUANT_TABLE_C, if (tevVersion == 3) QUANT_TABLE_Y else QUANT_TABLE_Y,
if (tevVersion == 3) QUANT_TABLE_C else QUANT_TABLE_C,
if (tevVersion == 3) QUANT_TABLE_C else QUANT_TABLE_C,
qY, qCo, qCg, rateControlFactors, qY, qCo, qCg, rateControlFactors,
blocksX, blocksY blocksX, blocksY
) )
@@ -2869,9 +2871,9 @@ class GraphicsJSR223Delegate(private val vm: VM) {
readPtr += 768 readPtr += 768
// Perform hardware IDCT for each channel using fast algorithm // Perform hardware IDCT for each channel using fast algorithm
val yBlock = tevIdct16x16_fast(coeffShortArray.sliceArray(0 until 256), QUANT_TABLE_Y, qY, rateControlFactor) val yBlock = tevIdct16x16_fast(coeffShortArray.sliceArray(0 until 256), if (tevVersion == 3) QUANT_TABLE_Y else QUANT_TABLE_Y, qY, rateControlFactor)
val coBlock = tevIdct8x8_fast(coeffShortArray.sliceArray(256 until 320), QUANT_TABLE_C, true, qCo, rateControlFactor) val coBlock = tevIdct8x8_fast(coeffShortArray.sliceArray(256 until 320), if (tevVersion == 3) QUANT_TABLE_C else QUANT_TABLE_C, true, qCo, rateControlFactor)
val cgBlock = tevIdct8x8_fast(coeffShortArray.sliceArray(320 until 384), QUANT_TABLE_C, true, qCg, rateControlFactor) val cgBlock = tevIdct8x8_fast(coeffShortArray.sliceArray(320 until 384), if (tevVersion == 3) QUANT_TABLE_C else QUANT_TABLE_C, true, qCg, rateControlFactor)
// Convert to RGB (YCoCg-R for v2, XYB for v3) // Convert to RGB (YCoCg-R for v2, XYB for v3)
val rgbData = if (tevVersion == 3) { val rgbData = if (tevVersion == 3) {
@@ -2893,9 +2895,9 @@ class GraphicsJSR223Delegate(private val vm: VM) {
readPtr += 768 readPtr += 768
// Step 2: Decode residual DCT // Step 2: Decode residual DCT
val yResidual = tevIdct16x16_fast(coeffShortArray.sliceArray(0 until 256), QUANT_TABLE_Y, qY, rateControlFactor) val yResidual = tevIdct16x16_fast(coeffShortArray.sliceArray(0 until 256), if (tevVersion == 3) QUANT_TABLE_Y else QUANT_TABLE_Y, qY, rateControlFactor)
val coResidual = tevIdct8x8_fast(coeffShortArray.sliceArray(256 until 320), QUANT_TABLE_C, true, qCo, rateControlFactor) val coResidual = tevIdct8x8_fast(coeffShortArray.sliceArray(256 until 320), if (tevVersion == 3) QUANT_TABLE_C else QUANT_TABLE_C, true, qCo, rateControlFactor)
val cgResidual = tevIdct8x8_fast(coeffShortArray.sliceArray(320 until 384), QUANT_TABLE_C, true, qCg, rateControlFactor) val cgResidual = tevIdct8x8_fast(coeffShortArray.sliceArray(320 until 384), if (tevVersion == 3) QUANT_TABLE_C else QUANT_TABLE_C, true, qCg, rateControlFactor)
// Step 3: Build motion-compensated YCoCg-R block and add residuals // Step 3: Build motion-compensated YCoCg-R block and add residuals
val finalY = IntArray(256) val finalY = IntArray(256)

40
video_encoder/encoder_tev.c
View File

@@ -154,7 +154,6 @@ static const uint32_t QUANT_TABLE_C[HALF_BLOCK_SIZE_SQR] =
99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
99, 99, 99, 99, 99, 99, 99, 99}; 99, 99, 99, 99, 99, 99, 99, 99};
// Audio constants (reuse MP2 from existing system) // Audio constants (reuse MP2 from existing system)
#define MP2_SAMPLE_RATE 32000 #define MP2_SAMPLE_RATE 32000
#define MP2_DEFAULT_PACKET_SIZE 1728 #define MP2_DEFAULT_PACKET_SIZE 1728
@@ -353,19 +352,18 @@ static inline double HLG_inverse_OETF(double V) {
} }
// ---------------------- Matrices (doubles) ---------------------- // ---------------------- Matrices (doubles) ----------------------
// Combined linear sRGB -> LMS (single 3x3): product of sRGB->XYZ, XYZ->BT2020, BT2020->LMS // linear RGB -> LMS (technically we should convert sRGB to Rec.2100, but if encoder and decoder agrees on the same colourimetry, this utilises more bits
// Computed from standard matrices (double precision).
static const double M_RGB_TO_LMS[3][3] = { static const double M_RGB_TO_LMS[3][3] = {
{0.20502672199540622, 0.42945363228947586, 0.31165003516511786}, {1688.0/4096.0,2146.0/4096.0, 262.0/4096.0},
{0.2233144413317712, 0.5540422172466897, 0.21854692537153908}, { 683.0/4096.0,2951.0/4096.0, 462.0/4096.0},
{0.0609931761282002, 0.17917502499816504, 0.9323768661336348} { 99.0/4096.0, 309.0/4096.0,3688.0/4096.0}
}; };
// Inverse: LMS -> linear sRGB (inverse of above) // Inverse: LMS -> linear sRGB (inverse of above)
static const double M_LMS_TO_RGB[3][3] = { static const double M_LMS_TO_RGB[3][3] = {
{29.601046511687, -21.364325340529906, -4.886500015143518}, {3.436606694333079, -2.5064521186562705, 0.06984542432319149},
{-12.083229161592032, 10.673933874098694, 1.5369143265611211}, {-0.7913295555989289, 1.983600451792291, -0.192270896193362},
{0.38562844776642574, -0.6536244436141302, 1.0968381245163787} {-0.025949899690592665, -0.09891371471172647, 1.1248636144023192}
}; };
// ICtCp matrix (L' M' S' -> I Ct Cp). Values are the BT.2100 integer-derived /4096 constants. // ICtCp matrix (L' M' S' -> I Ct Cp). Values are the BT.2100 integer-derived /4096 constants.
@@ -1011,8 +1009,13 @@ static void convert_rgb_to_color_space_block(tev_encoder_t *enc, const uint8_t *
} }
// Average and store subsampled chroma, scale to signed 8-bit equivalent range // Average and store subsampled chroma, scale to signed 8-bit equivalent range
c2_workspace[cy * HALF_BLOCK_SIZE + cx] = (float)((sum_ct / 4.0) * 255.0); // Apply centering to ensure chroma is balanced around 0 (like YCoCg-R)
c3_workspace[cy * HALF_BLOCK_SIZE + cx] = (float)((sum_cp / 4.0) * 255.0); double avg_ct = sum_ct / 4.0;
double avg_cp = sum_cp / 4.0;
// Scale and clamp to [-256, 255] range like YCoCg-R
c2_workspace[cy * HALF_BLOCK_SIZE + cx] = (float)CLAMP(avg_ct * 255.0, -256, 255);
c3_workspace[cy * HALF_BLOCK_SIZE + cx] = (float)CLAMP(avg_cp * 255.0, -256, 255);
} }
} }
} else { } else {
@@ -1338,7 +1341,7 @@ static void encode_block(tev_encoder_t *enc, int block_x, int block_y, int is_ke
dct_16x16_fast(enc->y_workspace, enc->dct_workspace); dct_16x16_fast(enc->y_workspace, enc->dct_workspace);
// quantise Y coefficients (luma) using per-block rate control // quantise Y coefficients (luma) using per-block rate control
const uint32_t *y_quant = QUANT_TABLE_Y; const uint32_t *y_quant = enc->ictcp_mode ? QUANT_TABLE_Y : QUANT_TABLE_Y;
const float qmult_y = jpeg_quality_to_mult(enc->qualityY * block->rate_control_factor); const float qmult_y = jpeg_quality_to_mult(enc->qualityY * block->rate_control_factor);
for (int i = 0; i < BLOCK_SIZE_SQR; i++) { for (int i = 0; i < BLOCK_SIZE_SQR; i++) {
// Apply rate control factor to quantization table (like decoder does) // Apply rate control factor to quantization table (like decoder does)
@@ -1350,7 +1353,7 @@ static void encode_block(tev_encoder_t *enc, int block_x, int block_y, int is_ke
dct_8x8_fast(enc->co_workspace, enc->dct_workspace); dct_8x8_fast(enc->co_workspace, enc->dct_workspace);
// quantise Co coefficients (chroma - orange-blue) using per-block rate control // quantise Co coefficients (chroma - orange-blue) using per-block rate control
const uint32_t *co_quant = QUANT_TABLE_C; const uint32_t *co_quant = enc->ictcp_mode ? QUANT_TABLE_C : QUANT_TABLE_C;
const float qmult_co = jpeg_quality_to_mult(enc->qualityCo * block->rate_control_factor); const float qmult_co = jpeg_quality_to_mult(enc->qualityCo * block->rate_control_factor);
for (int i = 0; i < HALF_BLOCK_SIZE_SQR; i++) { for (int i = 0; i < HALF_BLOCK_SIZE_SQR; i++) {
// Apply rate control factor to quantization table (like decoder does) // Apply rate control factor to quantization table (like decoder does)
@@ -1362,7 +1365,8 @@ static void encode_block(tev_encoder_t *enc, int block_x, int block_y, int is_ke
dct_8x8_fast(enc->cg_workspace, enc->dct_workspace); dct_8x8_fast(enc->cg_workspace, enc->dct_workspace);
// quantise Cg coefficients (chroma - green-magenta, qmult_cg is more aggressive like NTSC Q) using per-block rate control // quantise Cg coefficients (chroma - green-magenta, qmult_cg is more aggressive like NTSC Q) using per-block rate control
const uint32_t *cg_quant = QUANT_TABLE_C; // In ICtCp mode, Cg becomes Cp (chroma-red) which needs special quantization table
const uint32_t *cg_quant = enc->ictcp_mode ? QUANT_TABLE_C : QUANT_TABLE_C;
const float qmult_cg = jpeg_quality_to_mult(enc->qualityCg * block->rate_control_factor); const float qmult_cg = jpeg_quality_to_mult(enc->qualityCg * block->rate_control_factor);
for (int i = 0; i < HALF_BLOCK_SIZE_SQR; i++) { for (int i = 0; i < HALF_BLOCK_SIZE_SQR; i++) {
// Apply rate control factor to quantization table (like decoder does) // Apply rate control factor to quantization table (like decoder does)
@@ -2779,9 +2783,11 @@ int main(int argc, char *argv[]) {
} }
if (enc->ictcp_mode) { if (enc->ictcp_mode) {
int qc = (enc->qualityCo + enc->qualityCg) / 2; // ICtCp: Ct and Cp have different characteristics than YCoCg Co/Cg
enc->qualityCo = qc; // Cp channel now uses specialized quantization table, so moderate quality is fine
enc->qualityCg = qc; int base_chroma_quality = (enc->qualityCo + enc->qualityCg) >> 1;
enc->qualityCo = base_chroma_quality; // Ct channel: keep original Co quantization
enc->qualityCg = base_chroma_quality; // Cp channel: same quality since Q_Cp_8 handles detail preservation
} }
if (!test_mode && (!enc->input_file || !enc->output_file)) { if (!test_mode && (!enc->input_file || !enc->output_file)) {