Interpolate the second half of a two-data-point frame from dp0 to dp1 when
the object has no previous matrix; a single data point keeps the hard cut.
Rebuild the passthrough frame as the 8-plane core order L, R, C, LFE, Ls,
Rs, (Lrs, Rrs) written to output slots 0..7, all delayed by 1217 samples,
and keep that 8-plane history updated on every frame. Frames without JOC
metadata always take this passthrough path, and a core channel count
change clears the LFE ring while keeping the clipgain ramp state.
Reject unsupported downmix configurations with a structured variant error
in the metadata scan, and score ID14 payloads by the frame status instead
of accepting any lenient parse.
Update the math notes and README for the two-data-point interpolation
without a previous state and the 8-plane passthrough output.
Decode clipgain multiplicatively as 2^(x-4)*(1 + y/32), keep the raw
5-bit offset_ts values and reject frames whose offsets leave 0..23 or are
not strictly increasing, and read the EMDF sample offset as
variable_bits(11). Report parse errors through the frame status: 2
unsupported downmix configuration, 3 object count clamped, 4 parameter
band index, 5 invalid offset_ts, 6 trailing data after byte padding.
Correct the 15/9/5/3 subband-to-parameter-band rows.
Apply the clipgain ramp to the LFE samples before the 1217-sample delay,
stop fading objects in from zero when they have no previous matrix, and
keep disappeared objects published with an all-zero matrix until they
reappear.
Render each frame through the seq state machine: hold the previous
matrices while the decoder is unsynchronized, hard cut on seq 0 or a core
channel count change, and pass the delayed core PCM through when no JOC
metadata is present. Keep the 16-channel output as LFE + 15 objects so
CBA bed channels stay on the fixed-position object path that the ADM,
speaker, and binaural backends already consume.
Update the math notes and README for the multiplicative clipgain, the
no-previous-state interpolation, the LFE gain ramp, and the frame-level
decoding behavior. State that E-AC-3 input without JOC metadata is
outside the supported scope.
- Render 5.1-family and 4.0 layouts through expanded surround geometry before folding.
- Keep trim, balance and automatic position compensation tied to the final target layout.
- Preserve height channels, LFE passthrough and per-object trim bypass behavior.
- Update the bilingual README and mathematical documentation.
Read warp mode from OAMD instead of always scaling the floor Y coordinate.
For eligible layouts, mode 1 uses clamp(2Y, 0, 1), while mode 0 leaves Y
unchanged. Height coordinates are not scaled. Modes 2/3 remain explicitly
unsupported, and the CLI provides no warp override.
Keep separate coordinates for floor panning, height panning and trim
evaluation, with W denoting warp and B denoting front-back balance:
floor Y = B(W(Y))
height Y = B(Y)
trim Y = B(Y)
Apply explicit trims as a position-dependent gain for each object before
accumulation, replacing fixed gains applied to output channels after mixing.
Automatic, disabled and explicit trim modes are mutually exclusive.
Per-object bypass disables balance and trim while retaining warp.
Update controls at OAMD metadata events and retain the previous controls
when a payload has no trim element. Keep source trajectories and binaural
rendering unchanged.
Write the OAMD warp code into dbmd segment 9 while preserving the other
packed fields. Update the bilingual mathematical notes, README files and
code comments to describe parameters and rendering behavior.
Checked: warp 0/1, Y clipping, height coordinates, balance/trim ordering,
per-object bypass, metadata updates, bed positions and LFE passthrough.
Select the trim configuration that applies to the output layout from its
surround and height speaker counts:
cfg = kappa_f(surrounds) + 3 * kappa_h(heights)
Trim and balance apply to seven layouts: 2.0, 3.1, 4.0, 5.1, 5.1.2, 5.1.4
and 7.1. Wider layouts are left untrimmed.
An explicit configuration contributes a static level for the centre,
surround and height groups, which multiplies the panning gains per output
channel. Automatic and disabled configurations contribute no static
level: the attenuation an automatic configuration describes is already
applied per object by the layout position compensation.
An explicit configuration also contributes the two front-back balance
values. Writing n = round(16 |b|) and s = n / 18, each balance
reparameterises the OAMD front-back axis in [0, 1]:
y <= 0.05 or y >= 0.95 -> y
y < 0.5, b > 0 -> (1 + s) y - 0.05 s
y >= 0.5, b > 0 -> (1 - s) (y - 0.5) + 0.5 + 0.45 s
y < 0.5, b < 0 -> (1 - s) y + 0.05 s
y >= 0.5, b < 0 -> (1 + s) (y - 0.5) + 0.5 - 0.45 s
The map is monotonic onto [0.05, 0.95], keeps both walls fixed, and needs
no clamping. Height blends the listener-plane and overhead values before
panning; the horizontal and vertical coordinates are untouched.
Warp is not read here: it selects a downmix mode and does not enter the
speaker path.
The runtime renders two inputs now: the Rosella .personalized_headphone model and
an already compiled .jochrtf field. SOFA import, the full-HRIR direct renderer
and the SOFA -> .jochrtf fitting compiler leave this checkout; they continue as
an offline tool that only produces caches this repository loads.
The .jochrtf renderer goes back to the implementation that predates the direct
SOFA work: 64-QMF analysis, 77 hybrid bands, per-object hybrid path delays and
the project room (six first-order images plus the shared FDN tail) computed in
code. It is restored verbatim under the .jochrtf vocabulary
(SofaBinauralBackend -> JochrtfBackend) and gains one switch:
--jochrtf-room on|off, off by default. With the room off the renderer outputs
only the field's direct binaural sound, and in either mode it reads no audio
assets.
Because the room is computed in code again, the generated sofa-room*.npz assets
and the modules only the SOFA path used are removed (sofa_canonical, sofa_hrir,
sofa_direct_backend, sofa_room, hybrid_paths, partitioned_convolution,
point_source, compile_hrtf, generate_sofa_room). data/ keeps the decode tables
and the Rosella filterbanks.
The CLI drops --sofa-hrtf, --hrtf-radius-m, --sofa-interpolation,
--hrtf-cache-policy, --hrtf-cache-dir, --sofa-room, --sofa-room-preset and
--source-reference-distance-m. A .jochrtf is still explicit through
--compiled-hrtf-cache, and the default binaural input is the personalized model.
Rosella parsing, filterbank, room and timing are untouched and its output stays
bit-identical. The .jochrtf path reproduces the pre-SOFA renderer sample for
sample, with the room on and with it off.
SOFA and explicit .jochrtf inputs now render direct sound by default.
--sofa-room enables the project image-source/FDN room, with the compact
preset for Near/Mid and the original large room for Far.
- preserve native HRIR timing and apply Data.Delay once
- remove external reference configuration, radial head synthesis and the
artificial centre response
- ship both generated float64 room assets and their generator
- retain .jochrtf as an explicit lightweight input and offline export
- keep personalized headphone processing unchanged and update bilingual
documentation
Checked: independent direct/reflection convolution, FDN FIR/recursive
agreement, fractional delay and metadata timing, chunk invariance,
8 bit-exact personalized regressions, CLI smoke tests, and full-song
D1/D2 renders of two tracks.
SOFA inputs now use delay-aligned local interpolation and partitioned
convolution instead of compiling a compressed directional field.
- preserve full FIR support, complex phase and explicit delay handling
- add log-magnitude/phase and local magnitude-corrected interpolation
- support Near/Mid/Far with a shared calibration anchor
- retain .jochrtf as an explicit lightweight input and offline export
- keep personalized headphone rendering unchanged
- update the CLI, reports and bilingual documentation
Checked: 21 automated tests, 8 bit-exact personalized regressions,
and full-song D2 renders in Near/Mid/Far.
The Rosella path now chooses its filterbank. Bank 1 is the public bank
the reference implementation has always used (13 taps / 6 slots,
961 samples); bank 2 is the captured bank the reference headphone
renderer actually runs (11 taps / 3 slots, 448 samples).
- src/rosella_captured_filterbank.py implements the four half-steps in
float64/complex128; the captured prototype, rotation and hybrid
coefficients live in data/rosella_captured_kernels.npz
- src/rosella_bank.py owns the choice: both banks expose the same four
stages, each with its own coefficients, history and latency
- RosellaBinauralRenderer takes filterbank= and derives the startup
skip, the flush capacity and the report fields from it
- --rosella-filterbank 1|2 selects the bank, default 1; it is rejected
for the SOFA and .jochrtf inputs instead of being ignored
- document both banks in docs/rosella_filterbank.{md,en.md}, and point
the binaural docs, the data README and the options tables at it
Checked: the captured stages reproduce the reference vectors the C++
core checks (within 1e-10), an impulse returns at 961 and 448 samples,
and chunking, reset and bank independence hold for both.