Add Export -> Batch Export Images..., which renders and saves one PNG
per angle for whichever of CH1 (RF), CH3 (DC-A), CH4 (DC-B), and
Velocity the user selects, for the currently open file. Each channel
gets its own fixed min/max colorbar range, entered once in the dialog
and held constant across every exported angle, so the resulting images
are directly comparable to each other instead of each auto-scaling to
its own data (today's live-view default).
BatchExportDialog (sras_viewer.py) collects the output folder, file
prefix, and per-channel checkbox + range, seeded from whatever's
already cached (session DC/FFT caches, or the file's own v7 cache
blocks) so opening the dialog never triggers a fresh compute. Export
runs on a background thread via a new BatchExportWorker
(sras_workers.py), which computes each angle's channel image with the
existing dc_image_mv/compute_rf_image helpers (reusing one FFT per
angle for both RF and Velocity) and renders it with a headless
matplotlib Agg canvas.
Also includes several small correctness/robustness fixes that were
already staged in the working tree: an int16-vs-float32 accumulator
mismatch in sras_average.py's row averaging, a DC4-mask cache reuse and
atomic sidecar write in sras_compute.py, a dc3/dc4 pairing guard in
sras_format.py's v7 cache writer, and matching updates to the
tools/ test fixtures.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Three defects found by testing the parallel paths end to end.
Nested parallelism was unbounded. DcPrecomputeWorker and the alignment
driver both parallelise over angles, and each angle's compute_dc_image
then parallelised over rows again: 14 x 14 threads, and — worse — every
angle sized its chunk against the *whole* memory budget, so worst-case
live buffers were ~13.9 GB against a 1 GB budget. Capping the inner
worker count alone does not fix it; the chunk is still sized against the
full budget. plan_angle_level now hands each angle both max_workers=1
and its share of the budget, which bounds the real file at 1002 MB.
Chunk planning could not produce concurrency at all. _plan_chunks sized
chunk_rows first, and _chunk_rows_for spends whatever budget it is given
— so a single chunk consumed the lot and budget//chunk_bytes came back
as one worker. It now picks the worker count first and sizes the chunk to
it, giving 7-16 workers on the real file instead of 1.
Cancellation was too coarse to shut down. stop() was only checked between
angles, and one angle of the real scan is ~40 s, so closing the window
sat through it and returned with threads still running. should_stop is
now polled per row chunk and closeEvent signals every worker before
waiting: close went from 4.0 s with two live threads to 1.04 s with both
finished. A cancelled ComputeWorker emits None so a half-filled image is
never cached.
Also: BatchCacheWorker no longer reports every file as failed when the
process pool itself dies (unguarded __main__, frozen build, sandbox) —
it retries those files in-process. Batches below 512 MB skip the pool
entirely, since interpreter startup would otherwise make small jobs
slower. cache_file rejects an unknown mode instead of silently treating
it as "fft".
Measured on the real 496 GB scan, angle 3, 32 rows:
DC 2.94s -> 0.75s RF 6.10s -> 1.99s peak RSS 1.40 -> 1.93 GB
Batch convert, 24 files / 1 GB: 4.03s -> 2.36s.
Memory budget defaults to 1024 MB (SRAS_MEM_BUDGET_MB), the measured knee.
Testing: tools/test_refactor.py (70 checks: cache round-trip and block
carry-forward, partial-cache handling, parallel-vs-serial identity,
no-mask path, ROI mask vs full-grid, v2/v3/v4 parsing, sras_average
round-trip incl. remainder handling) and tools/test_gui.py (46 checks
driving the real widgets and workers headlessly). check_equivalence.py
still reports all 167 outputs identical to ed0eba4.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Thomas Ales [M S E]