Add SILK Burg LPC analysis

[thomas-vilte]

Description

adds burgModifiedFLP — LPC coefficient estimation via Burg's method, stacked over multiple subframes and capped at a max prediction gain (minInvGain). Wraps it in findLPC, which converts the result straight to NLSFs via a2nlsfFLP. the NLSF-interpolation search (picking the best interpolation index between the previous frame's NLSFs and this one) is deferred — the interpolation index stays fixed at 4 (no interpolation) for now

ported from burg_modified_FLP.c

caught one real bug while reviewing this locally, worth mentioning since it's not visible in the diff: I'd hand-copied the file instead of extracting it cleanly and introduced a few divergences from the reference (a dropped accumulation, a missing multiply, and a whole per-subframe update loop that went missing) — TestBurgModifiedFLPRecoversAR caught it immediately once I actually ran the tests. re-extracted the file straight from François's branch instead of retyping, so what's here now is his original, unmodified

Reference issue

part of the SILK encoder series (#158)

[codecov]

Codecov Report

❌ Patch coverage is 82.02247% with 16 lines in your changes missing coverage. Please review.
✅ Project coverage is 88.50%. Comparing base (ad89c91) to head (fe0e821).

Files with missing lines	Patch %	Lines
internal/silk/lpc_analysis.go	82.02%	13 Missing and 3 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #159      +/-   ##
==========================================
- Coverage   88.57%   88.50%   -0.07%     
==========================================
  Files          38       39       +1     
  Lines        8252     8341      +89     
==========================================
+ Hits         7309     7382      +73     
- Misses        712      725      +13     
- Partials      231      234       +3     

Flag	Coverage Δ
go	88.50% <82.02%> (-0.07%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.