curioustorvald/tsvm
1
0
Fork 0
mirror of https://github.com/curioustorvald/tsvm.git synced 2026-06-12 15:44:05 +09:00
Code Issues Packages Projects Releases Wiki Activity

TAV: allowing multi-title if video is larger than default size

Browse Source
This commit is contained in:
minjaesong
2025-10-01 09:32:34 +09:00
parent 70dfc7bf13
commit 3e40b048a7
4 changed files with 88 additions and 110 deletions
Download Patch File Download Diff File Expand all files Collapse all files

35
terranmon.txt
View File

@@ -987,13 +987,15 @@ transmission capability, and region-of-interest coding.
uint8 Quantiser override Y (uses exponential numeric system; stored with index bias of 1 (127->252, 255->4032); use 0 to disable overriding; shared with A channel)
uint8 Quantiser override Co (uses exponential numeric system; stored with index bias of 1 (127->252, 255->4032); use 0 to disable overriding)
uint8 Quantiser override Cg (uses exponential numeric system; stored with index bias of 1 (127->252, 255->4032); use 0 to disable overriding)
- note: quantiser overrides are always present regardless of the channel layout
## Coefficient Storage Format (Significance Map Compression)
- note: quantiser overrides are always present regardless of the channel layout
* Tile data (one compressed payload per tile)
### Coefficient Storage Format (Significance Map Compression)
Starting with encoder version 2025-09-29, DWT coefficients are stored using
significance map compression with concatenated maps layout for optimal efficiency:
### Concatenated Maps Format (Current)
#### Concatenated Maps Format
All channels are processed together to maximize Zstd compression:
uint8 Y Significance Map[(coeff_count + 7) / 8] // 1 bit per Y coefficient
@@ -1005,44 +1007,27 @@ transmission capability, and region-of-interest coding.
int16 Cg Non-zero Values[variable length] // Only non-zero Cg coefficients
int16 A Non-zero Values[variable length] // Only non-zero A coefficients (if alpha present)
### Significance Map Encoding
#### Significance Map Encoding
Each significance map uses 1 bit per coefficient position:
- Bit = 1: coefficient is non-zero, read value from corresponding Non-zero Values array
- Bit = 0: coefficient is zero
### Compression Benefits
#### Compression Benefits
- **Sparsity exploitation**: Typically 85-95% zeros in quantized DWT coefficients
- **Cross-channel patterns**: Concatenated maps allow Zstd to find patterns across similar significance maps
- **Overall improvement**: 16-18% compression improvement before Zstd compression
### Legacy Separate Format (2025-09-29 initial)
Early significance map implementation processed channels separately:
For each channel (Y, Co, Cg, optional A):
uint8 Significance Map[(coeff_count + 7) / 8] // 1 bit per coefficient
int16 Non-zero Values[variable length] // Only non-zero coefficients
### DWT Coefficient Structure (per tile)
## Legacy Format (for reference)
int16 Y channel DWT coefficients[width * height + 4]
int16 Co channel DWT coefficients[width * height + 4]
int16 Cg channel DWT coefficients[width * height + 4]
int16 A channel DWT coefficients[width * height + 4] (only when the video has alpha)
<for legacy non-monoblock format>
int16 Y channel DWT coefficients[tile width * tile height + 4]
int16 Co channel DWT coefficients[tile width * tile height + 4]
int16 Cg channel DWT coefficients[tile width * tile height + 4]
... (repeated per tile)
### DWT Coefficient Structure (per tile)
For each decomposition level L (from highest to lowest):
uint16 LL_size: size of LL subband coefficients
uint16 LH_size: size of LH subband coefficients
uint16 LH_size: size of LH subband coefficients
uint16 HL_size: size of HL subband coefficients
uint16 HH_size: size of HH subband coefficients
int16[] LL_coeffs: quantized LL subband (low-low frequencies)
int16[] LH_coeffs: quantized LH subband (low-high frequencies)
int16[] HL_coeffs: quantized HL subband (high-low frequencies)
int16[] HL_coeffs: quantized HL subband (high-low frequencies)
int16[] HH_coeffs: quantized HH subband (high-high frequencies)
## DWT Implementation Details