Struct YCbCrPlanarEncoder 

pub struct YCbCrPlanarEncoder { /* private fields */ }

Encoder for planar f32 YCbCr input.

Use when you have pre-converted YCbCr from video decoders, etc. Skips RGB->YCbCr conversion entirely.

Only valid with ColorMode::YCbCr. XYB mode requires RGB input.

YCbCr Value Range

Input values should be in the centered range:

• Y: 0.0 to 255.0 (luma)
• Cb, Cr: -128.0 to 127.0 (centered chroma)

This matches the output of standard RGB→YCbCr conversion with BT.601 coefficients.

Streaming

Data can be pushed in any row count - the encoder buffers partial strips internally and flushes when a complete strip is accumulated.

Implementations

impl YCbCrPlanarEncoder

pub fn push( &mut self, planes: &YCbCrPlanes<'_>, rows: usize, stop: impl Stop, ) -> Result<()>

Push full-resolution planes. Encoder subsamples chroma as needed.

All three planes must be at full luma resolution (width × rows). The encoder will perform chroma subsampling according to the configured ChromaSubsampling mode.

Data can be pushed in any amount - the encoder buffers partial strips internally and flushes when a complete strip is accumulated.

Arguments

• planes: Y, Cb, Cr plane data with per-plane strides
• rows: Number of luma rows to push
• stop: Cancellation token (use Unstoppable if not needed)

Value Range

• Y: 0.0 to 255.0
• Cb, Cr: -128.0 to 127.0

pub fn push_subsampled( &mut self, planes: &YCbCrPlanes<'_>, y_rows: usize, stop: impl Stop, ) -> Result<()>

Push with pre-subsampled chroma.

Use this when your chroma planes are already at the target subsampled resolution. Y plane is still at full resolution.

Note: Unlike push(), this method does not buffer partial strips. For best results, push complete strips (multiples of strip_height rows, which is 8 for 4:4:4 or 16 for 4:2:0).

Arguments

• planes: Y at full resolution, Cb/Cr at subsampled resolution
• y_rows: Number of luma rows to push
• stop: Cancellation token

Chroma Dimensions

The expected chroma dimensions depend on the subsampling mode:

• 4:4:4 (None): cb/cr at full width × full height
• 4:2:2 (HalfHorizontal): cb/cr at width/2 × full height
• 4:2:0 (Quarter): cb/cr at width/2 × height/2

pub fn width(&self) -> u32

Get image width.

pub fn height(&self) -> u32

Get image height.

pub fn rows_pushed(&self) -> u32

Get number of rows pushed so far (including buffered rows).

pub fn rows_remaining(&self) -> u32

Get number of rows remaining.

pub fn finish(self) -> Result<Vec<u8>>

Finish encoding, return JPEG bytes.

pub fn finish_to<W: Write>(self, output: W) -> Result<W>

Finish encoding to Write destination.

pub fn finish_to_vec(self, output: &mut Vec<u8>) -> Result<()>

Finish encoding, appending JPEG bytes to an existing Vec.

Useful for no_std environments or buffer reuse.