Expand description
SIMD-dispatched color-conversion kernels for the FFmpeg AVPixelFormat
space, exposed through honest, independently-usable tiers.
§Tier 0: Convert — the golden one-call decode
Convert::from(&frame) decodes a validated source frame to one or more
output buffers with zero redundant parameters: the dimensions come from the
frame, the colorimetry from a single ColorSpec, and each per-format walk
knob is derived from that spec (overridable). Attach outputs —
rgb / rgba / luma /
hsv — plus an optional area or filtered
resize with infallible consuming setters;
Convert::run is the only fallible call. Failures surface as Error, an
alias of MixedSinkerError.
use pixon::{Convert, ColorMatrix, ColorSpec, DynamicRange, PixelFormat};
use pixon::frame::Yuv420pFrame;
let (w, h) = (8u32, 8u32);
let y = [16u8; 64];
let (u, v) = ([128u8; 16], [128u8; 16]);
let frame = Yuv420pFrame::new(&y, &u, &v, w, h, w, w / 2, w / 2);
let spec = ColorSpec::resolve(PixelFormat::Yuv420p, DynamicRange::Limited, ColorMatrix::Bt709);
let mut rgb = [0u8; 4 * 4 * 3];
let mut luma = [0u8; 4 * 4];
Convert::from(&frame)
.spec(spec)
.resize(4, 4)
.rgb(&mut rgb)
.luma(&mut luma)
.run()?;§Tier 1: Walker / MixedSinker / {fmt}_to — streaming & custom sinks
Every source pixel format also has a per-format walker (yuv420p_to,
nv12_to, bgr24_to, …) that walks the source row by row and hands each row
to a caller-supplied PixelSink. The Sink decides what to derive — luma
only, RGB only, HSV only, all three, or something custom — and writes into
whatever buffers it owns. Reach for this tier when Tier 0’s fixed shape is
too rigid: custom output geometry / strides, partial-frame feeding, a bespoke
sink, or a Bayer CFA source (whose mosaic pattern is frame-intrinsic rather
than spec-derivable, so it is Tier-1-only — see raw::bayer_to /
raw::bayer16_to).
The row the Sink receives (Self::Input<'_>) has a shape that reflects the
source format: source::Yuv420pRow carries Y / U / V slices plus matrix /
range metadata; packed‑RGB row types (e.g. source::Rgb24Row,
source::Bgr24Row) carry a single packed slice; etc. Each source family
declares a subtrait (Yuv420pSink: PixelSink<Input<'_> = Yuv420pRow<'_>>) so
kernel signatures stay sharp. For the common “give me RGB / Luma / HSV or any
subset” case the crate ships sinker::MixedSinker, configured via
with_rgb /
with_luma /
with_hsv to select which channels to
derive; the generic Walker plus per-format Options (YuvOptions, …)
give the same walk a uniform, spec-configurable shape.
As of 0.2 the per-row SIMD kernels (row::*) are crate-internal plumbing
consumed by the sinker and walkers — they are no longer public API.
§HDR → SDR for analysis: hdr
Both tiers are transfer-preserving, so a PQ or HLG picture decodes to
R'G'B' that still holds HDR codes. hdr::SdrMapping maps those codes
to BT.709 SDR R'G'B' for a consumer that reads ordinary display values.
It is one documented, analysis-grade mapping, not a renderer’s tone mapper.
The module also publishes the BT.2100 curves the mapping is built from.
§Supported source formats
Shipped (4:1:0, 4:1:1, 4:2:0, 4:2:2, 4:4:0, and 4:4:4 subsampling, plus
non-subsampled RGB / gray / indexed-color sources). This table is
mechanically derived from the crate’s own Source-trait exhaustiveness
proof (src/convert/source_coverage.rs, a #[cfg(test)]-only internal
module) — every row below has a corresponding assert_source::<F>()
call there, gated by the same Cargo feature:
| Family | Bit depth | Subsampling | Packing | FFmpeg name | rgb_u16 depth |
|---|---|---|---|---|---|
Yuv411p | 8 | 4:1:1 | planar (DV-NTSC legacy) | yuv411p | — |
Yuv410p | 8 | 4:1:0 | planar (DV-NTSC / Cinepak-Sorenson legacy) | yuv410p | — |
Yuv420p | 8 | 4:2:0 | planar | yuv420p | — |
Yuv422p | 8 | 4:2:2 | planar | yuv422p | — |
Yuv440p | 8 | 4:4:0 | planar | yuv440p | — |
Yuv444p | 8 | 4:4:4 | planar | yuv444p | — |
Nv12 | 8 | 4:2:0 | semi-planar UV | nv12 | — |
Nv21 | 8 | 4:2:0 | semi-planar VU | nv21 | — |
Nv16 | 8 | 4:2:2 | semi-planar UV | nv16 | — |
Nv20 | 10 | 4:2:2 | semi-planar UV, low-packed | nv20le | 10 |
Nv24 | 8 | 4:4:4 | semi-planar UV | nv24 | — |
Nv42 | 8 | 4:4:4 | semi-planar VU | nv42 | — |
Yuv420p9 | 9 | 4:2:0 | planar, low-packed | yuv420p9le | 9 |
Yuv420p10 | 10 | 4:2:0 | planar, low-packed | yuv420p10le | 10 |
Yuv420p12 | 12 | 4:2:0 | planar, low-packed | yuv420p12le | 12 |
Yuv420p14 | 14 | 4:2:0 | planar, low-packed | yuv420p14le | 14 |
Yuv420p16 | 16 | 4:2:0 | planar | yuv420p16le | 16 |
Yuv422p9 | 9 | 4:2:2 | planar, low-packed | yuv422p9le | 9 |
Yuv422p10 | 10 | 4:2:2 | planar, low-packed | yuv422p10le | 10 |
Yuv422p12 | 12 | 4:2:2 | planar, low-packed | yuv422p12le | 12 |
Yuv422p14 | 14 | 4:2:2 | planar, low-packed | yuv422p14le | 14 |
Yuv422p16 | 16 | 4:2:2 | planar | yuv422p16le | 16 |
Yuv440p10 | 10 | 4:4:0 | planar, low-packed | yuv440p10le | 10 |
Yuv440p12 | 12 | 4:4:0 | planar, low-packed | yuv440p12le | 12 |
Yuv444p9 | 9 | 4:4:4 | planar, low-packed | yuv444p9le | 9 |
Yuv444p10 | 10 | 4:4:4 | planar, low-packed | yuv444p10le | 10 |
Yuv444p12 | 12 | 4:4:4 | planar, low-packed | yuv444p12le | 12 |
Yuv444p14 | 14 | 4:4:4 | planar, low-packed | yuv444p14le | 14 |
Yuv444p16 | 16 | 4:4:4 | planar | yuv444p16le | 16 |
Yuv444p10Msb | 10 | 4:4:4 | planar, MSB-packed | yuv444p10msble | 10 |
Yuv444p12Msb | 12 | 4:4:4 | planar, MSB-packed | yuv444p12msble | 12 |
P010 | 10 | 4:2:0 | semi-planar, high-packed | p010le | 10 |
P012 | 12 | 4:2:0 | semi-planar, high-packed | p012le | 12 |
P016 | 16 | 4:2:0 | semi-planar | p016le | 16 |
P210 | 10 | 4:2:2 | semi-planar, high-packed | p210le | 10 |
P212 | 12 | 4:2:2 | semi-planar, high-packed | p212le | 12 |
P216 | 16 | 4:2:2 | semi-planar | p216le | 16 |
P410 | 10 | 4:4:4 | semi-planar, high-packed | p410le | 10 |
P412 | 12 | 4:4:4 | semi-planar, high-packed | p412le | 12 |
P416 | 16 | 4:4:4 | semi-planar | p416le | 16 |
Yuva420p | 8 | 4:2:0 | planar Y/U/V/A (4 planes), source α | yuva420p | — |
Yuva420p9 | 9 | 4:2:0 | planar Y/U/V/A, low-packed, source α | yuva420p9le | 9 |
Yuva420p10 | 10 | 4:2:0 | planar Y/U/V/A, low-packed, source α | yuva420p10le | 10 |
Yuva420p12 | 12 | 4:2:0 | planar Y/U/V/A, low-packed, source α (no FFmpeg enum) | yuva420p12le | 12 |
Yuva420p16 | 16 | 4:2:0 | planar Y/U/V/A, source α | yuva420p16le | 16 |
Yuva422p | 8 | 4:2:2 | planar Y/U/V/A (4 planes), source α | yuva422p | — |
Yuva422p9 | 9 | 4:2:2 | planar Y/U/V/A, low-packed, source α | yuva422p9le | 9 |
Yuva422p10 | 10 | 4:2:2 | planar Y/U/V/A, low-packed, source α | yuva422p10le | 10 |
Yuva422p12 | 12 | 4:2:2 | planar Y/U/V/A, low-packed, source α | yuva422p12le | 12 |
Yuva422p16 | 16 | 4:2:2 | planar Y/U/V/A, source α | yuva422p16le | 16 |
Yuva444p | 8 | 4:4:4 | planar Y/U/V/A (4 planes), source α | yuva444p | — |
Yuva444p9 | 9 | 4:4:4 | planar Y/U/V/A, low-packed, source α | yuva444p9le | 9 |
Yuva444p10 | 10 | 4:4:4 | planar Y/U/V/A, low-packed, source α | yuva444p10le | 10 |
Yuva444p12 | 12 | 4:4:4 | planar Y/U/V/A, low-packed, source α | yuva444p12le | 12 |
Yuva444p14 | 14 | 4:4:4 | planar Y/U/V/A, low-packed, source α (no FFmpeg enum) | yuva444p14le | 14 |
Yuva444p16 | 16 | 4:4:4 | planar Y/U/V/A, source α | yuva444p16le | 16 |
Yuyv422 | 8 | 4:2:2 | packed byte quad (Y0,U0,Y1,V0) | yuyv422 | — |
Uyvy422 | 8 | 4:2:2 | packed byte quad (U0,Y0,V0,Y1) | uyvy422 | — |
Yvyu422 | 8 | 4:2:2 | packed byte quad (Y0,V0,Y1,U0) | yvyu422 | — |
Uyyvyy411 | 8 | 4:1:1 | packed, 6B/4px block (U,Y0,Y1,V,Y2,Y3) | uyyvyy411 | — |
V210 | 10 | 4:2:2 | packed (3 x 10-bit/u32) | v210 | 10 |
Y210 | 10 | 4:2:2 | packed, MSB-aligned u16 | y210le | 10 |
Y212 | 12 | 4:2:2 | packed, MSB-aligned u16 | y212le | 12 |
Y216 | 16 | 4:2:2 | packed, full-range u16 | y216le | 16 |
V410 | 10 | 4:4:4 | packed (one 32-bit word) | v410 | 10 |
V30X | 10 | 4:4:4 | packed (one 32-bit word) | v30xle | 10 |
Xv36 | 12 | 4:4:4 | packed u16 quadruple, α-as-padding | xv36le | 12 |
Xv48 | 16 | 4:4:4 | packed u16 quadruple, α-as-padding | xv48le | 16 |
Vuya | 8 | 4:4:4 | packed byte quadruple, source α | vuya | — |
Vuyx | 8 | 4:4:4 | packed byte quadruple, α-as-padding | vuyx | — |
Ayuv | 8 | 4:4:4 | packed byte quadruple (A,Y,U,V), source α | ayuv | — |
Uyva | 8 | 4:4:4 | packed byte quadruple (U,Y,V,A), source α | uyva | — |
Vyu444 | 8 | 4:4:4 | packed byte triple (V,Y,U), no alpha | vyu444 | — |
Ayuv64 | 16 | 4:4:4 | packed u16 quadruple, source α | ayuv64le | 16 |
Gbrp | 8 | 4:4:4 | planar GBR (3 planes) | gbrp | — |
Gbrp9 | 9 | 4:4:4 | planar GBR (3 planes), low-packed | gbrp9le | 9 |
Gbrp10 | 10 | 4:4:4 | planar GBR (3 planes), low-packed | gbrp10le | 10 |
Gbrp12 | 12 | 4:4:4 | planar GBR (3 planes), low-packed | gbrp12le | 12 |
Gbrp14 | 14 | 4:4:4 | planar GBR (3 planes), low-packed | gbrp14le | 14 |
Gbrp16 | 16 | 4:4:4 | planar GBR (3 planes) | gbrp16le | 16 |
Gbrp10Msb | 10 | 4:4:4 | planar GBR (3 planes), MSB-packed | gbrp10msble | 10 |
Gbrp12Msb | 12 | 4:4:4 | planar GBR (3 planes), MSB-packed | gbrp12msble | 12 |
Gbrap | 8 | 4:4:4 | planar GBR + A (4 planes, source α) | gbrap | — |
Gbrap10 | 10 | 4:4:4 | planar GBR+A (4 planes), low-packed, source α | gbrap10le | 10 |
Gbrap12 | 12 | 4:4:4 | planar GBR+A (4 planes), low-packed, source α | gbrap12le | 12 |
Gbrap14 | 14 | 4:4:4 | planar GBR+A (4 planes), low-packed, source α | gbrap14le | 14 |
Gbrap16 | 16 | 4:4:4 | planar GBR+A (4 planes), source α | gbrap16le | 16 |
Gbrap32 | 32 | 4:4:4 | planar GBR+A (4 planes), full u32, source α | gbrap32le | 16 (full scale) |
Gbrpf16 | 16f | 4:4:4 | planar GBR (3 planes), half-float | gbrpf16le | 16 (full scale) |
Gbrpf32 | 32f | 4:4:4 | planar GBR (3 planes), f32 | gbrpf32le | 16 (full scale) |
Gbrapf16 | 16f | 4:4:4 | planar GBR+A (4 planes), half-float, source α | gbrapf16le | 16 (full scale) |
Gbrapf32 | 32f | 4:4:4 | planar GBR+A (4 planes), f32, source α | gbrapf32le | 16 (full scale) |
Rgb24 | 8 | 4:4:4 | packed (R,G,B) | rgb24 | — |
Bgr24 | 8 | 4:4:4 | packed (B,G,R) | bgr24 | — |
Rgba | 8 | 4:4:4 | packed (R,G,B,A), source α | rgba | — |
Bgra | 8 | 4:4:4 | packed (B,G,R,A), source α | bgra | — |
Argb | 8 | 4:4:4 | packed (A,R,G,B), source α | argb | — |
Abgr | 8 | 4:4:4 | packed (A,B,G,R), source α | abgr | — |
Xrgb | 8 | 4:4:4 | packed (X,R,G,B), α-as-padding | 0rgb | — |
Rgbx | 8 | 4:4:4 | packed (R,G,B,X), α-as-padding | rgb0 | — |
Xbgr | 8 | 4:4:4 | packed (X,B,G,R), α-as-padding | 0bgr | — |
Bgrx | 8 | 4:4:4 | packed (B,G,R,X), α-as-padding | bgr0 | — |
Rgb48 | 16 | 4:4:4 | packed u16 triple (R,G,B) | rgb48le | 16 |
Bgr48 | 16 | 4:4:4 | packed u16 triple (B,G,R) | bgr48le | 16 |
Rgba64 | 16 | 4:4:4 | packed u16 quadruple, source α | rgba64le | 16 |
Bgra64 | 16 | 4:4:4 | packed u16 quadruple, source α | bgra64le | 16 |
Rgb96 | 32 | 4:4:4 | packed u32 triple (R,G,B) | rgb96le | 16 (full scale) |
Rgba128 | 32 | 4:4:4 | packed u32 quadruple, source α | rgba128le | 16 (full scale) |
X2Rgb10 | 10 | 4:4:4 | packed u32, 2-bit pad (R,G,B) | x2rgb10le | 10 |
X2Bgr10 | 10 | 4:4:4 | packed u32, 2-bit pad (B,G,R) | x2bgr10le | 10 |
Rgb444 | 4 | 4:4:4 | packed u16, 4-bit pad (R,G,B) | rgb444le | 4 |
Bgr444 | 4 | 4:4:4 | packed u16, 4-bit pad (B,G,R) | bgr444le | 4 |
Rgb555 | 5 | 4:4:4 | packed u16, 1-bit pad (R,G,B) | rgb555le | 5 |
Bgr555 | 5 | 4:4:4 | packed u16, 1-bit pad (B,G,R) | bgr555le | 5 |
Rgb565 | 5/6/5 | 4:4:4 | packed u16 (R,G,B) | rgb565le | 5/6/5 |
Bgr565 | 5/6/5 | 4:4:4 | packed u16 (B,G,R) | bgr565le | 5/6/5 |
Rgb8 | 3/3/2 | 4:4:4 | packed byte (R,G,B) | rgb8 | 3/3/2 |
Bgr8 | 2/3/3 | 4:4:4 | packed byte (B,G,R) | bgr8 | 3/3/2 |
Rgb4Byte | 1/2/1 | 4:4:4 | packed byte, low nibble (R,G,B) | rgb4_byte | 1/2/1 |
Bgr4Byte | 1/2/1 | 4:4:4 | packed byte, low nibble (B,G,R) | bgr4_byte | 1/2/1 |
Rgb4 | 1/2/1 | 4:4:4 | bitstream, 2px/byte (R,G,B) | rgb4 | 1/2/1 |
Bgr4 | 1/2/1 | 4:4:4 | bitstream, 2px/byte (B,G,R) | bgr4 | 1/2/1 |
Rgbf16 | 16f | 4:4:4 | packed half-float triple (R,G,B) | rgbf16 | 16 (full scale) |
Rgbf32 | 32f | 4:4:4 | packed f32 triple (R,G,B) | rgbf32 | 16 (full scale) |
Rgbaf16 | 16f | 4:4:4 | packed half-float quadruple, source α | rgbaf16le | 16 (full scale) |
Rgbaf32 | 32f | 4:4:4 | packed f32 quadruple, source α | rgbaf32le | 16 (full scale) |
Gray8 | 8 | N/A | single luma plane | gray | — |
Gray9 | 9 | N/A | single luma plane, low-packed u16 | gray9le | 9 |
Gray10 | 10 | N/A | single luma plane, low-packed u16 | gray10le | 10 |
Gray12 | 12 | N/A | single luma plane, low-packed u16 | gray12le | 12 |
Gray14 | 14 | N/A | single luma plane, low-packed u16 | gray14le | 14 |
Gray16 | 16 | N/A | single luma plane | gray16le | 16 |
Gray32 | 32 | N/A | single luma plane, u32 | gray32le | 16 (full scale) |
Grayf16 | 16f | N/A | single luma plane, half-float | grayf16le | 16 (full scale) |
Grayf32 | 32f | N/A | single luma plane, f32 | grayf32le | 16 (full scale) |
Ya8 | 8 | N/A | packed luma+alpha byte pairs, source α | ya8 | 8 |
Ya16 | 16 | N/A | packed luma+alpha u16 pairs, source α | ya16le | 16 |
Yaf16 | 16f | N/A | packed luma+alpha half-float pairs, source α | yaf16le | 16 (full scale) |
Yaf32 | 32f | N/A | packed luma+alpha f32 pairs, source α | yaf32le | 16 (full scale) |
Monoblack | 1 | N/A | 1bpp bitmap, MSB-first, 0=black | monoblack | 8 |
Monowhite | 1 | N/A | 1bpp bitmap, MSB-first, 0=white | monowhite | 8 |
Pal8 | 8 | N/A | single index plane + 256-entry palette | pal8 | 16 (full scale) |
Xyz12 | 12 | 4:4:4 | packed CIE XYZ (3 x u16, high-bit-packed: bits [15:4]) | xyz12le / xyz12be | 16 (full scale) |
The last column is the depth of the source’s packed u16 RGB face — what
Convert::rgb_u16 and the sink tier’s with_rgb_u16 write — where it has
one (RgbU16Source::RGB_U16_DEPTHS). An integer source keeps its native
depth, low-packed; a legacy packed RGB source keeps each channel’s own,
given R/G/B where they differ (RgbDepths). “Full scale” is a face the
source widens or narrows to the whole u16 range (a palette, a 32-bit or
float source, XYZ12). — has no u16 face, and Convert::rgb_u16 on it
does not compile. A source whose sink tier has the face and whose walker!
row does not state it does not build.
Xyz12 is the DCP / digital-cinema source format. Decoding
it requires a SMPTE ST 428-1 §8 inverse OETF, a 3x3 matrix to one
of three target gamuts (DcpTargetGamut::DciP3 /
DcpTargetGamut::Rec709 / DcpTargetGamut::Rec2020), then a
sRGB-shape forward OETF and integer narrow. Every backend is
native SIMD; the OETFs run scalar per lane to preserve the 0-ULP
scalar↔SIMD parity contract.
§RAW (Bayer) sources
raw::Bayer (8-bit) and raw::Bayer16<BITS> (10/12/14/16-bit
low-packed u16, range [0, (1 << BITS) - 1]) feed bilinear
demosaic + white balance + 3x3
color-correction in a single per-row kernel. Caller supplies
raw::BayerPattern (BGGR / RGGB / GRBG / GBRG),
raw::WhiteBalance gains, and a raw::ColorCorrectionMatrix.
See raw for the full design and parameter docs.
Scope: pixon covers demosaic onwards. Producing the Bayer
plane itself is the upstream pipeline’s job — vendor-SDK
camera-RAW decoders (R3D / BRAW / NRAW) for compressed
camera bitstreams, or FFmpeg’s AV_PIX_FMT_BAYER_* pixel
formats / bayer_* decoders for already-uncompressed Bayer
sources. Once you have a BayerFrame / BayerFrame16, hand it
to raw::bayer_to / raw::bayer16_to with your sink of
choice.
§YUVA sources
Every FFmpeg yuva* pixel format (4:2:0 / 4:2:2 / 4:4:4, 8- through
16-bit) plus the two non-FFmpeg extension depths (yuva420p12le,
yuva444p14le, shipped for symmetry with their non-alpha siblings)
ships as a native Source format — Yuva420p and its
high-bit / 4:2:2 / 4:4:4 siblings, gated behind the yuva feature
(which auto-enables yuv-planar). See the roster table above for
the full 16-row list; it is mechanically derived from the same
source_coverage.rs proof these markers satisfy, rather than
hand-copied, so it cannot silently drift out of sync the way the list
that used to live in this section did.
RGBA pass-through — preserving the source’s real alpha plane into the
output, rather than a constant fill — is shipped:
with_rgba /
with_rgba_u16 on
sinker::MixedSinker attach an 8-bit or native-depth RGBA buffer
that carries the YUVA source’s actual alpha channel;
rgba rides the same path.
A caller who only wants RGB / Luma output and would rather not enable
the yuva feature can still reach for alpha-drop: hand the
Y / U / V slices from a 4-plane YUVA buffer to the matching non-alpha
Yuv*p*Frame constructor (available under yuv-planar alone) and
ignore the alpha plane. This remains a valid lighter-weight
alternative, not the only mechanism.
§Kernel families
- Q15 i32 family — 8-bit kernels (
yuv_420_to_rgb_row,yuv_444_to_rgb_row,nv12_to_rgb_row,nv24_to_rgb_rowetc.) and 10/12/14-bit kernels (yuv_420p_n_to_rgb_*<BITS>,yuv_444p_n_to_rgb_*<BITS>,p_n_to_rgb_*<BITS>). Native SIMD on every backend (NEON / SSE4.1 / AVX2 / AVX-512 / wasm simd128).Yuv422p(and theYuv422p10/Yuv422p12/Yuv422p14family) reusesYuv420p’s per-row kernels (4:2:2 differs only in the vertical walker); same forNv16↔Nv12.Yuv444pandYuv444p10/Yuv444p12/Yuv444p14use a dedicated 4:4:4 kernel family (no horizontal chroma duplication step);Nv24andNv42share a 4:4:4 kernel family via aSWAP_UVconst generic. - 16-bit family — dedicated
yuv_420p16_to_rgb_*,yuv444p16_to_rgb_*,p16_to_rgb_*.Yuv422p16reuses the 4:2:0 16-bit kernels by shape equivalence. The u8-output kernels stay on i32 (output-range scaling keepscoeff x u_dwithin i32). The u16-output kernels widen the chroma matrix multiply-add to i64 to avoid the ~2.31·10⁹ chroma-channel sum overflowing i32 atBITS == 16; the Y path also widens to i64 to handle limited-range unclamped samples.
§SIMD coverage
Every format above has a native SIMD backend for each supported
target (NEON on aarch64; SSE4.1 / AVX2 / AVX-512 on x86_64; wasm
simd128). Every u8-output and u16-output path has a native
implementation on every backend — including the 16-bit u16-output
paths for Yuv420p16, P016, and Yuv444p16, which use the
backend-native i64 arithmetic (native _mm512_srai_epi64 on
AVX-512 and i64x2_shr on wasm; srai64_15 bias trick on SSE4.1
and AVX2 because those ISAs lack native i64 arithmetic right
shift).
§Not yet shipped (follow-up)
- Bayer SIMD backends — Tier 14 currently dispatches to the scalar reference path on every target; NEON / SSE4.1 / AVX2 / AVX-512 / wasm simd128 follow-ups will land per the established backend-symmetry pattern.
- Cinema-camera RAW source formats — vendor-decoded sensor RGB
in camera-native log + gamut (LogC4 / S-Log3 / REDLog3G10 /
Canon Log 2/3 / BMD Film Gen 5 / V-Log / F-Log) → working-space
conversion via inverse-OETF + 3x3 matrix + sRGB OETF. Roadmap
tracked in
docs/superpowers/plans/2026-05-07-be-rollout-tracking.mdunder “Cinema Camera RAW Support Roadmap”. Mirrors the Tier 12 (source::Xyz12) shape: per-vendor source format, fullMixedSinkeroutput coverage, polynomial OETF, 5 SIMD backends. - Higher-quality Bayer demosaic — current scalar Bayer kernel does bilinear demosaic; AHD / Malvar / DCB are quality levers for cinema-grade proxies (CinemaDNG / DJI Inspire workflows).
- 3D LUT (
.cube) row kernel — for OCIO-style color management in cinema pipelines.
See source for the per-format module-level breakdown and
frame for the validated frame types plus the BITS const
generic on the high-bit-depth families (Yuv420pFrame16<BITS>
and PnFrame<BITS>).
Re-exports§
pub use walker::BayerOptions;bayerpub use walker::ColorSpec;pub use walker::Walker;pub use walker::YuvOptions;pub use convert::Convert;allocorstdpub use convert::FromSpec;allocorstdpub use convert::RgbDepths;allocorstdpub use convert::RgbU16Source;allocorstdpub use convert::Source;allocorstdpub use sinker::MixedSinkerError as Error;allocorstd
Modules§
- convert
allocorstd - Tier-0
Convert— the golden one-call decode entry point. - frame
- Frame primitives + the typed source-format
*Frame<'a, BE>borrow types. - hdr
allocorstd - Analysis-grade HDR → SDR: the ITU-R BT.2100 transfer curves, and one
documented mapping from a PQ or HLG picture’s
R'G'B'to the BT.709 SDRR'G'B'an analysis consumer reads. - raw
- RAW (Bayer) source kernels.
- resample
allocorstd - Resampling strategies for
MixedSinker. - sinker
PixelSinkimplementations shipped with the crate.- source
- Source pixel-format kernels — marker ZSTs, per-row borrow types
(
*Row<'a>), per-formatSinksubtraits, and walker fns that iteratecrate::frame::*Frame<'a, BE>row-by-row dispatching to aPixelSink. - walker
- A uniform
Walkerlayer over the per-format frame walkers.
Structs§
- Ayuv64
yuv-444-packed - Zero-sized marker for the packed AYUV64 source format
(
AV_PIX_FMT_AYUV64{LE,BE}).<const BE: bool>defaults tofalse(LE); the aliasAyuv64resolves toAyuv64<false>. - Ayuv64
Frame yuv-444-packed - Validated wrapper around a packed YUV 4:4:4 16-bit
AYUV64plane. - Ayuv64
Row yuv-444-packed - One row of an
Ayuv64source —width × 4u16 elements (4 channels per pixel: A, Y, U, V; the A slot is real source alpha). - Color
Info - Bundled color metadata that rides on every video frame.
- Unsupported
Kernel Gamut Error - The error the
DcpTargetGamut→KernelGamutconversion returns: this gamut has no defined XYZ → RGB matrix or luma basis. - Unsupported
Kernel Matrix Error - The error the
Matrix→KernelMatrixconversion returns: this matrix names no coefficient set a conversion kernel can use.
Enums§
- Ayuv64
Frame Error yuv-444-packed - Errors returned by
Ayuv64Frame::try_new. - Chroma
Location - Chroma sample location (for subsampled YUV formats).
- Color
Matrix - Color matrix coefficients per ITU-T H.273 MatrixCoefficients (Table 4) / ISO/IEC 23001-8.
- DcpTarget
Gamut - Target RGB gamut for the XYZ → RGB matrix step in the
source::Xyz12source pipeline (xyz12_to). - Dynamic
Range - Sample range — limited (TV / studio swing) vs. full (PC).
- Kernel
Gamut - The closed set of DCP target gamuts an XYZ → RGB kernel has a
matrix for —
KernelMatrix’s twin on the gamut axis. - Kernel
Matrix - The closed set of colour matrices a conversion kernel has
coefficients for — the
Copyselector the row walkers carry. - Pixel
Format - Pixel format identifier covering FFmpeg + Bayer + cinema-RAW.
- Primaries
- Color primaries per ITU-T H.273 ColourPrimaries (Table 2) / ISO/IEC 23001-8.
- Transfer
- Transfer characteristics per ITU-T H.273 (Table 3).
Traits§
- Ayuv64
Sink yuv-444-packed - Sinks that consume rows of this source format. The
<const BE>parameter encodes the source byte-order — sinkers typically impl for one specificBEmatching their storedMixedSinker<Marker<BE>>monomorphization. The Row type does not carryBE; the BE-aware kernel dispatch happens insideprocessvia the sinker’s own<const BE>parameter. - Pixel
Sink - Base trait for sinks that consume rows of a source pixel format.
- Source
Format - Sealed marker trait identifying a source pixel format.
Functions§
- ayuv64_
to yuv-444-packed - LE-only back-compat wrapper preserving the pre-Phase-4
packedwalker signature. Forwards to the const-generic helper withBE = false.