AEC Debugging Started with Channel Mapping, Not DSP Tuning

Before tuning echo cancellation I verified that the algorithm was receiving the microphone and far-end reference channels I thought it was.

By Reaz Romen✓ Site verified

Lab period February 2026 Published 2023-01-16 1 min read

Lab Note. Reconstructed in September 2026 from current engineering work and lab notes. The archive date indicates the period covered; the current site publication date is shown separately.

Echo cancellation can waste days if the signal plumbing is wrong. The algorithm may run, CPU usage may look normal, and the result may still be useless because the reference channel does not contain the far-end signal.

I moved channel verification to the start of the process. Capture the individual I2S channels, identify which one is microphone audio, identify the playback reference, and verify their timing. Only then does it make sense to tune filter length, adaptation or suppression.

This also caught assumptions introduced by codec routing. A channel index in the application is not proof of what the codec placed on that slot. Hardware mapping, TDM configuration and firmware channel order all have to agree.

Once the reference was correct, echo behavior changed in a way that tuning could actually improve. That was a useful reminder that DSP problems are often ordinary integration problems first.