Package com.tino1b2be.dtmf.internal

Class BlockSizer

java.lang.Object
com.tino1b2be.dtmf.internal.BlockSizer
public final class BlockSizer extends Object
Analysis-block sizing logic for the DTMF detection pipeline.

The detector picks a block length N such that the effective Goertzel bin width sampleRate / N lands in the closed range [40, 60] Hz, with a target of 50 Hz in the middle of the band. That range is wide enough that every one of the Supported_Sample_Rate values (8 kHz, 16 kHz, 44.1 kHz, 48 kHz) lands on an integer N producing exactly 50 Hz bin width (Requirement 3.5), and comfortably satisfies Requirement 3.4 for any integer sample rate in [4000, 192000] Hz.

The algorithm is a rounded divide plus a clamp: start from N = round(sampleRate / 50), then nudge N up or down until sampleRate / N lands back inside [40, 60]. At any positive integer sample rate in the supported range this terminates in at most two steps because the 1 / N spacing between candidate bin widths is strictly narrower than the 20 Hz tolerance band.

Although the type is public so DtmfConfig in the sibling package com.tino1b2be.dtmf can call it, the convention is that com.tino1b2be.dtmf.internal.* is not part of the published API: callers go through DtmfConfig.Advanced rather than constructing this type directly. There is no constructor to call; every entry point is a static method.

  • Method Details

    • blockSizeFor

      public static int blockSizeFor(int sampleRate)
      Compute the analysis block size for the given sample rate so that the effective Goertzel bin width lies in [40, 60] Hz.
      Parameters:
      sampleRate - sample rate in Hz; must be > 0
      Returns:
      a positive int N such that sampleRate / N is in [40.0, 60.0]
      Throws:
      IllegalArgumentException - if sampleRate <= 0