fix(isomp4): fix M4A moov-at-end probe failures and streaming fallback (#757)

* fix(stream): defer M4A probe until moov tail or fast-start prefix

Ranged moov-at-end M4A started Symphonia format probe as soon as ~384 KiB
linear data arrived, before the parallel tail prefetch filled the moov
atom — probe hit end of stream and skipped the track. Wait for tail_ready
or detect fast-start moov in the prefix; do not arm playback from linear
bytes alone when tail prefetch is active.

* fix(isomp4): skip EOF-spanning mdat after moov-at-end is parsed

Second-pass header scan with moov already loaded still tried to read
through mdat→EOF on RangedHttpSource holes, causing format probe
end of stream. Re-check play generation after moov wait.

* fix(isomp4): fix AtomIterator overread after seek in patched demuxer

`AtomIterator::new_root(reader, len)` treats `len` as bytes available
from the current `reader.pos()`, not the absolute file length. After
`mss.seek(resume_at)` we were passing the absolute `total_len`, so the
iterator thought there were `total_len` bytes left and tried to read
past EOF on the next iteration, returning "end of stream".

Fix: pass `total_len.map(|tl| tl.saturating_sub(resume_at))` (remaining
bytes from the new position) in both branches:
- `moov.is_none()` (moov-at-end layout, seek to moov offset)
- `moov.is_some()` (fast-start layout, skip bounded mdat body)

This caused Symphonia to fail probing moov-at-end M4A files read from
local disk (hot-cache) and from in-memory buffers — every decode attempt
returned "end of stream", analysis fell back to `byte_envelope_no_ebu`
(no EBU R128 loudness), and rodio produced distorted audio.

Also in this commit:
- `resolve_playback_format_hint()` helper to resolve hint from URL,
  stream suffix, Content-Disposition, or byte sniff
- ISO-BMFF diagnostic helpers (`isobmff_buffer_looks_complete`,
  `log_isobmff_buffer_diagnostic`, `mp4_suspect_zero_holes`)
- Probe-fallback path for ranged-stream failures now uses these helpers
  to decide whether to refetch or wait for the in-flight download

* chore(audio): remove redundant hint recomputation and add missing blank line

`bytes_hint_for_wait` in the ranged-stream fallback path was an exact
duplicate of `effective_hint` already in scope — reuse the existing binding.
Also add missing blank line after `wait_for_ranged_mp4_probe_ready` in mod.rs.

* chore(release): CHANGELOG and credits for PR #757

Add Fixed entry for M4A moov-at-end probe fix and credits line in
settingsCredits.ts under existing cucadmuh contributions.

* fix(audio): extract BuildSourceArgs to fix clippy::too_many_arguments

`build_playback_source_with_probe_fallback` had 12 parameters, exceeding
the clippy limit of 7. Group url/gen/hints/fade/hi-res/duration into
`BuildSourceArgs` so the function signature stays at 4 arguments.

CHANGELOG.md

@@ -663,6 +663,14 @@ Foundational work: faster reviews, narrower diffs, and a safety net under the pa
 
 * Selecting an empty library no longer leaves these pages as a fully blank canvas. A shared `common.libraryEmpty` message ("Your library is empty.") added across all nine locales is shown in place of the empty rails/grids. Pages that already had a dedicated empty-state (Artists, Genres, Composers, Playlists, Favorites, Most Played, Lossless Albums, Label Albums, Internet Radio) keep their per-page wording. On Albums and New Releases, an active genre / year / starred / compilation filter still shows the regular filtered-results behaviour rather than the library-empty message.
 
+### M4A playback — probe failures and distorted audio on moov-at-end files
+
+**By [@cucadmuh](https://github.com/cucadmuh), PR [#757](https://github.com/Psychotoxical/psysonic/pull/757)**
+
+* Fixed a bug in the patched `symphonia-format-isomp4` demuxer that caused Symphonia to fail probing every M4A file read from the hot-cache (in-memory bytes) or from a local file after playing it once. `AtomIterator::new_root` treats `len` as bytes available from the current reader position, not the absolute file length; after `seek(offset)` the code was passing `total_len` (absolute) instead of `total_len − offset` (remaining), so the iterator overread past EOF and returned `end of stream`. Loudness analysis fell back to `byte_envelope_no_ebu` (no EBU R128 gain applied) and the audio was decoded with distortion. Both affected branches — moov-at-end layout and fast-start mdat skip — now receive the correct remaining length.
+* Ranged streaming of moov-at-end M4A files no longer races Symphonia's probe against the tail prefetch: a `RangedMp4ProbeGate` holds the probe thread until either the moov tail fetch completes or moov is confirmed in the already-downloaded prefix (fast-start). Previously the probe could start on a partial buffer and immediately return `format probe failed: end of stream`, falling through to the sequential-download fallback even when the track would have played fine.
+* The full-download fallback path now runs ISO-BMFF diagnostic checks on the completed ranged buffer before handing it to Symphonia: zero-hole detection (`mp4_suspect_zero_holes`) and completeness gating (`isobmff_buffer_looks_complete`) trigger an automatic HTTP refetch when the buffer looks sparse or structurally incomplete.
+
 ## [1.45.0] - 2026-05-04
 
 ## Added