Struct TrellisConfig 

pub struct TrellisConfig {
    pub enabled: bool,
    pub dc_enabled: bool,
    pub eob_opt: bool,
    pub use_lambda_weight_tbl: bool,
    pub use_scans_in_trellis: bool,
    pub q_opt: bool,
    pub lambda_log_scale1: f32,
    pub lambda_log_scale2: f32,
    pub freq_split: i32,
    pub num_loops: i32,
    pub delta_dc_weight: f32,
}

Configuration for trellis quantization.

Fields

enabled: bool

Enable trellis quantization for AC coefficients

dc_enabled: bool

Enable trellis quantization for DC coefficients

eob_opt: bool

Optimize for sequences of EOB

use_lambda_weight_tbl: bool

Use perceptual lambda weighting table

use_scans_in_trellis: bool

Consider scan order in trellis optimization

q_opt: bool

Optimize quantization table in trellis loop

lambda_log_scale1: f32

Lambda log scale parameter 1

lambda_log_scale2: f32

Lambda log scale parameter 2

freq_split: i32

Frequency split point for spectral selection

num_loops: i32

Number of trellis optimization loops

delta_dc_weight: f32

DC delta weight for vertical gradient consideration

Implementations

impl TrellisConfig

pub const fn disabled() -> TrellisConfig

Configuration with trellis disabled (fastest mode).

pub fn favor_size() -> TrellisConfig

Preset that favors smaller file sizes over quality.

Uses lower lambda values which makes the trellis algorithm more aggressive about zeroing coefficients, resulting in smaller files at the cost of some quality loss.

Lambda = 2^scale1 / (2^scale2 + norm). Lower lambda = more aggressive zeroing.

pub fn favor_quality() -> TrellisConfig

Preset that favors quality over file size.

Uses higher lambda values which makes the trellis algorithm more conservative, preserving more coefficients for better quality at the cost of larger files.

Lambda = 2^scale1 / (2^scale2 + norm). Higher lambda = more conservative.

pub fn lambda_scales(self, scale1: f32, scale2: f32) -> TrellisConfig

Set the lambda log scale parameters directly.

These control the rate-distortion tradeoff in trellis quantization:

• scale1: Controls rate penalty (higher = smaller files, default 14.75)
• scale2: Controls distortion sensitivity (higher = better quality, default 16.5)

The effective lambda is: 2^scale1 / (2^scale2 + block_norm)

pub fn rd_factor(self, factor: f32) -> TrellisConfig

Adjust rate-distortion balance with a simple factor.

• factor > 1.0: Favor quality (higher lambda, more conservative)
• factor < 1.0: Favor smaller files (lower lambda, more aggressive)
• factor = 1.0: Default behavior

The factor multiplies the effective lambda value logarithmically.

pub fn ac_trellis(self, enabled: bool) -> TrellisConfig

Enable or disable AC coefficient trellis optimization.

pub fn dc_trellis(self, enabled: bool) -> TrellisConfig

Enable or disable DC coefficient trellis optimization.

pub fn eob_optimization(self, enabled: bool) -> TrellisConfig

Enable or disable EOB run optimization.

Trait Implementations

impl Clone for TrellisConfig

fn clone(&self) -> TrellisConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TrellisConfig

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

Formats the value using the given formatter.

impl Default for TrellisConfig

fn default() -> TrellisConfig

Returns the “default value” for a type.

impl PartialEq for TrellisConfig

fn eq(&self, other: &TrellisConfig) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for TrellisConfig

impl StructuralPartialEq for TrellisConfig