Struct FrameHeader 

pub struct FrameHeader {

    pub frame_type: FrameType,
    pub encoding: Encoding,
    pub xyb_encoded: bool,
    pub flags: u64,
    pub do_ycbcr: bool,
    pub jpeg_upsampling: [u8; 3],
    pub upsampling: u32,
    pub ec_upsampling: Vec<u32>,
    pub group_size_shift: u32,
    pub x_qm_scale: u32,
    pub b_qm_scale: u32,
    pub num_passes: u32,
    pub x0: i32,
    pub y0: i32,
    pub width: u32,
    pub height: u32,
    pub blend_mode: BlendMode,
    pub ec_blend_modes: Vec<BlendMode>,
    pub blend_source: u32,
    pub alpha_blend_channel: u32,
    pub save_as_reference: u32,
    pub save_before_ct: bool,
    pub name: String,
    pub have_animation: bool,
    pub have_timecodes: bool,
    pub duration: u32,
    pub timecode: u32,
    pub is_last: bool,
    pub gaborish: bool,
    pub epf_iters: u32,
}

Frame header structure.

Used by both VarDCT and Modular encoding paths. Fields are parameterized to cover both modes. Use lossy() or lossless() constructors for defaults.

Fields

frame_type: FrameType

Frame type.

encoding: Encoding

Encoding method.

xyb_encoded: bool

Whether the image metadata has xyb_encoded=true. Controls whether do_ycbcr is written (only when false).

flags: u64

Frame flags (e.g., SKIP_ADAPTIVE_LF_SMOOTHING=0x80, ENABLE_NOISE=0x01).

do_ycbcr: bool

Whether the frame uses YCbCr color transform (only written when !xyb_encoded).

jpeg_upsampling: [u8; 3]

JPEG upsampling mode for chroma (only for VarDCT + YCbCr).

upsampling: u32

Upsampling factor (1, 2, 4, or 8).

ec_upsampling: Vec<u32>

Extra channel upsampling factors.

group_size_shift: u32

Group size shift (Modular only: 0=128, 1=256, 2=512, 3=1024).

x_qm_scale: u32

X channel quant matrix scale (VarDCT only, 3 bits, range 0-7).

b_qm_scale: u32

B channel quant matrix scale (VarDCT only, 3 bits, range 0-7).

num_passes: u32

Number of passes (1-10).

x0: i32

X offset for cropped frames.

y0: i32

Y offset for cropped frames.

width: u32

Frame width (0 = full image width).

height: u32

Frame height (0 = full image height).

blend_mode: BlendMode

Blending information for the main frame.

ec_blend_modes: Vec<BlendMode>

Per-extra-channel blending modes.

blend_source: u32

Source reference frame for blending (0-3).

alpha_blend_channel: u32

Alpha channel to use for blending.

save_as_reference: u32

Whether frame is saved for reference.

save_before_ct: bool

Whether to save before color transform.

name: String

Frame name.

have_animation: bool

Whether the file header signals animation (have_animation=true). When true, duration/timecode fields are written for normal frames.

have_timecodes: bool

Whether the file header signals have_timecodes.

duration: u32

Duration in ticks (for animation).

timecode: u32

Timecode (if have_timecodes).

is_last: bool

Whether this is the last frame.

gaborish: bool

Enable gaborish (Gabor-like blur in decoder loop filter).

epf_iters: u32

Number of EPF (Edge-Preserving Filter) iterations (0-3).

Implementations

impl FrameHeader

pub fn lossy() -> Self

Creates a frame header for a lossy VarDCT frame with default parameters.

Defaults: xyb_encoded=true, flags=SKIP_ADAPTIVE_LF_SMOOTHING (0x80), gaborish=true, epf_iters=2.

pub fn lossless() -> Self

Creates a frame header for a lossless Modular frame.

Defaults: xyb_encoded=false, do_ycbcr=false, flags=0, group_size_shift=1 (256), gaborish=false, epf_iters=0.

pub fn write(&self, writer: &mut BitWriter) -> Result<()>

Writes the frame header to the bitstream.

Follows the JXL codestream specification (ISO 18181-1) Table A.2.

Trait Implementations

impl Clone for FrameHeader

fn clone(&self) -> FrameHeader

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for FrameHeader

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for FrameHeader

fn default() -> Self

Returns the “default value” for a type.