saa-py: feed_audio sample_rate resampling behavior not documented for non-16kHz input

[MunawarJohar]

Package

attenlabs-saa (Python)

Version

0.6.1

What happened?

The README states that feed_audio accepts a sample_rate= parameter and "it'll resample" when audio isn't already 16kHz, but it doesn't document:

• What resampling method/quality is used (linear, polyphase, etc.)
• Whether resampling adds noticeable latency for real-time use cases (e.g. an 8kHz Twilio call)
• Whether sample_rate needs to be passed on every feed_audio() call or only once at setup

Expected: the README or docstring would clarify resampling method, latency impact, and call-frequency requirements.
Actual: only one line mentions resampling exists, with no further detail.

Steps to reproduce

1. Read packages/saa-py/README.md, "Feeding external audio" section
2. Note the line: "pass sample_rate= if your audio isn't already 16 kHz and it'll resample"
3. No further detail on resampling method, latency, or per-call vs one-time usage is given anywhere in the README or linked docs

Environment

N/A — documentation issue, not runtime-specific

Logs / errors

[MunawarJohar]

I'd like to work on this one — planning to submit a docs PR clarifying the resampling method, latency impact, and call-frequency (per-call vs one-time) for feed_audio's sample_rate parameter, once confirmed with a maintainer.