YCgCo refactor

app/src/main.rs

@@ -33,7 +33,16 @@ use image::{ColorType, DynamicImage, EncodableLayout, GenericImageView, ImageRea
 use std::fs::File;
 use std::io::Read;
 use std::time::Instant;
-use yuvutils_rs::{ar30_to_rgb8, i010_to_rgb10, i010_to_rgb_f16, i012_to_rgb12, i014_to_rgb14, i016_to_rgb16, i210_to_rgb10, i210_to_rgb_f16, i210_to_rgba_f16, i214_to_rgb14, i214_to_rgb_f16, i214_to_rgba14, i216_to_rgb16, i410_to_rgb10, i410_to_rgb_f16, i410_to_rgba10, i414_to_rgb14, i414_to_rgb_f16, i416_to_rgb16, p210_to_ar30, p212_to_ar30, rgb10_to_i010, rgb10_to_i210, rgb10_to_i410, rgb10_to_p210, rgb12_to_i012, rgb12_to_p212, rgb14_to_i014, rgb14_to_i214, rgb14_to_i414, rgb16_to_i016, rgb16_to_i216, rgb16_to_i416, rgba14_to_i214, Rgb30ByteOrder, YuvBiPlanarImageMut, YuvChromaSubsampling, YuvPlanarImageMut, YuvRange, YuvStandardMatrix};
+use yuvutils_rs::{
+    ar30_to_rgb8, i010_to_rgb10, i010_to_rgb_f16, i012_to_rgb12, i014_to_rgb14, i016_to_rgb16,
+    i210_to_rgb10, i210_to_rgb_f16, i210_to_rgba_f16, i214_to_rgb14, i214_to_rgb_f16,
+    i214_to_rgba14, i216_to_rgb16, i410_to_rgb10, i410_to_rgb_f16, i410_to_rgba10, i414_to_rgb14,
+    i414_to_rgb_f16, i416_to_rgb16, p210_to_ar30, p212_to_ar30, rgb10_to_i010, rgb10_to_i210,
+    rgb10_to_i410, rgb10_to_p210, rgb12_to_i012, rgb12_to_p212, rgb14_to_i014, rgb14_to_i214,
+    rgb14_to_i414, rgb16_to_i016, rgb16_to_i216, rgb16_to_i416, rgb_to_ycgco444, rgba14_to_i214,
+    ycgco444_to_rgb, ycgco444_to_rgba, Rgb30ByteOrder, YuvBiPlanarImageMut, YuvChromaSubsampling,
+    YuvPlanarImageMut, YuvRange, YuvStandardMatrix,
+};
 
 fn read_file_bytes(file_path: &str) -> Result<Vec<u8>, String> {
     // Open the file
@@ -94,20 +103,19 @@ fn main() {
     );
 
     let mut planar_image =
-        YuvPlanarImageMut::<u16>::alloc(width as u32, height as u32, YuvChromaSubsampling::Yuv444);
+        YuvPlanarImageMut::<u8>::alloc(width as u32, height as u32, YuvChromaSubsampling::Yuv444);
     //
     let mut bytes_16: Vec<u16> = src_bytes
         .iter()
         .map(|&x| ((x as u16) << 4) | ((x as u16) >> 4))
         .collect();
 
     let start_time = Instant::now();
-    rgb12_to_p212(
-        &mut bi_planar_image,
-        &bytes_16,
+    rgb_to_ycgco444(
+        &mut planar_image,
+        &src_bytes,
         rgba_stride as u32,
         YuvRange::Limited,
-        YuvStandardMatrix::Bt709,
     )
     .unwrap();
 
@@ -200,18 +208,14 @@ fn main() {
 
     let mut j_rgba = vec![0u8; dimensions.0 as usize * dimensions.1 as usize * 4];
 
-    p212_to_ar30(
-        &fixed_biplanar,
-        &mut j_rgba,
-        dimensions.0 as u32 * 4,
-        Rgb30ByteOrder::Host,
+    ycgco444_to_rgb(
+        &fixed_planar,
+        &mut rgba,
+        dimensions.0 as u32 * 3,
         YuvRange::Limited,
-        YuvStandardMatrix::Bt709,
     )
     .unwrap();
 
-    ar30_to_rgb8(&j_rgba,dimensions.0 as u32 * 4, Rgb30ByteOrder::Host, &mut rgba, dimensions.0 as u32 * 3, dimensions.0, dimensions.1 ).unwrap();
-
     // let a_plane = vec![1023u16; width as usize * height as usize];
     // let planar_with_alpha = YuvPlanarImageWithAlpha {
     //     y_plane: planar_image.y_plane.borrow(),

src/avx2/avx2_utils.rs

@@ -336,15 +336,6 @@ pub(crate) unsafe fn avx2_deinterleave_rgb(
 //     _mm256_permute4x64_epi64::<MASK>(v)
 // }
 
-#[inline(always)]
-pub(crate) unsafe fn avx2_pairwise_widen_avg(v: __m256i) -> __m256i {
-    let sums = _mm256_maddubs_epi16(v, _mm256_set1_epi8(1));
-    let shifted = _mm256_srli_epi16::<1>(_mm256_add_epi16(sums, _mm256_set1_epi16(1)));
-    let packed_lo = _mm256_packus_epi16(shifted, shifted);
-    const MASK: i32 = shuffle(3, 1, 2, 0);
-    _mm256_permute4x64_epi64::<MASK>(packed_lo)
-}
-
 #[inline(always)]
 pub(crate) unsafe fn avx_pairwise_avg_epi16(a: __m256i, b: __m256i) -> __m256i {
     let sums = _mm256_hadd_epi16(a, b);

src/avx2/mod.rs

@@ -28,7 +28,6 @@
  */
 #![deny(unreachable_code, unreachable_pub)]
 mod avx2_utils;
-mod avx2_ycgco;
 #[cfg(feature = "nightly_f16")]
 mod f16_converter;
 mod gbr_to_rgb;
@@ -39,7 +38,6 @@ mod rgb_to_nv420_prof;
 #[cfg(feature = "professional_mode")]
 mod rgb_to_nv_prof;
 mod rgb_to_y;
-mod rgb_to_ycgco;
 #[cfg(feature = "professional_mode")]
 mod rgb_to_yuv420_prof;
 mod rgb_to_yuv_p16;
@@ -61,7 +59,6 @@ mod rgba_to_yuv_fast420;
 mod shuffle;
 mod y_to_rgba;
 mod y_to_rgba_alpha;
-mod ycgco_to_rgb;
 mod ycgco_to_rgba_alpha;
 mod yuv_nv_to_rgba;
 mod yuv_nv_to_rgba420;
@@ -101,7 +98,6 @@ pub(crate) use rgb_to_nv420_prof::avx2_rgba_to_nv420_prof;
 #[cfg(feature = "professional_mode")]
 pub(crate) use rgb_to_nv_prof::avx2_rgba_to_nv_prof;
 pub(crate) use rgb_to_y::avx2_rgb_to_y_row;
-pub(crate) use rgb_to_ycgco::avx2_rgb_to_ycgco_row;
 #[cfg(feature = "professional_mode")]
 pub(crate) use rgb_to_yuv420_prof::avx2_rgba_to_yuv420_prof;
 pub(crate) use rgb_to_yuv_p16::avx_rgba_to_yuv_p16;
@@ -123,7 +119,6 @@ pub(crate) use rgba_to_yuv_fast420::avx2_rgba_to_yuv_dot_rgba420;
 pub(crate) use shuffle::{ShuffleConverterAvx2, ShuffleQTableConverterAvx2};
 pub(crate) use y_to_rgba::avx2_y_to_rgba_row;
 pub(crate) use y_to_rgba_alpha::avx2_y_to_rgba_alpha_row;
-pub(crate) use ycgco_to_rgb::avx2_ycgco_to_rgb_row;
 pub(crate) use ycgco_to_rgba_alpha::avx2_ycgco_to_rgba_alpha;
 pub(crate) use yuv_nv_to_rgba::avx2_yuv_nv_to_rgba_row;
 pub(crate) use yuv_nv_to_rgba420::avx2_yuv_nv_to_rgba_row420;