Alango Parametric EQ Optimizer Mass Production Yield Problem

Alango Parametric EQ Optimizer

Mass Production Yield Problem and Solution PROBLEM Inefficient yield of mass-produced audio products due to excessive frequency response variation CONSEQUENCES Scrapped goods and reduced production efficiency CAUSE Speaker tolerance and batch deviation SOLUTION Parametric equalization using digital filters to correct frequency magnitude response variation of each individual production device Alango Technologies – Sound Equalization Tool 2

The Solution: Parametric Equalization Two types of digital filters can potentially be employed: • Finite Impulse Response (FIR) • Infinite Impulse Response (IIR) Parametric equalization using IIR requires calculation of these individual parameters for each filter: • Center frequency, f c • Gain, G • Q (bandwidth selectivity) The computational task of parametric equalization using IIR represents a multidimensional problem since many filters and their respective parameters (fc, G, Q) must be calculated to correct for frequency response deviation. Alango Technologies – Sound Equalization Tool 3

Choice of Digital Filter: FIR or IIR Infinite Impulse Response (IIR) Finite Impulse Response (FIR) Pros • Calculating the parameters of the optimal correction filter is relatively straightforward; easily automated with existing methods Pros • Easy to implement • Near zero latency • Low computational cost Cons • Computationally expensive (MIPS) • Increased latency • Low memory requirements • Same filter parameters can be reused for multiple sampling rates • Relatively high memory requirements • Different sets of parameters must be precalculated for different sampling rates • Dependence between the complexity / resolution of the filter’s magnitude response and the latency and computational cost Alango Technologies – Sound Equalization Tool Cons • Calculating the parameters Fc, G, Q of the set of the optimal correction filters is a non-trivial task and is difficult to automate 4

Alango Parametric EQ Optimizer “ALPEQO” • ALPEQO cascades several IIR filters, easily automating the process of calculating the optimal correction filter parameters • ALPEQO is a standalone tool, easily embedded into any MS Windows PCbased test and measurement system • ALPEQO takes measured magnitude response curve and derives filter parameters for optimal correction based on the target magnitude response limits • ALPEQO contains additional factors such as maximum positive gain limit and factory EQ correction Alango Technologies – Sound Equalization Tool 5

Deviant Sample • This graph depicts a real-life example of a speaker characterized by a deviant magnitude response • The target magnitude response is graphically depicted in d. B between 20 Hz and 15 k. Hz • A measurement of the magnitude response of a deviant sample (yellow curve) is clearly outside the desired limits and requires compensation Alango Technologies – Sound Equalization Tool 6

Explanation of ALPEQO Operation Plots of the individual filter responses used to correct the deviant sample, as appear prior to optimization (initial state). Filter ‘master’ gain. Plot of the combined filter response using the filters (initial state). Plot of the residuals. Green points indicate residuals within the target limits. Plot of the target (desired response). Target response limits. INITIAL STATE Alango Technologies – Sound Equalization Tool Red points indicate residuals outside the target limits. 7

Video of ALPEQO Operation The data is passed to ALPEQO and fitting of the optimal filter parameters begins… Fitting the filter parameters represents an iterative process and takes a few seconds CLICK HERE TO PLAY ALPEQO tool in the process of calculation Alango Technologies – Sound Equalization Tool 8

Explanation of ALPEQO Filter Parameters Once the fitting is done, the following set of filter parameters is returned from ALPEQO: Low-shelf filter Peak filter 2 Gain=0. 058 d. B Gain=1. 76 d. B Frequency=70 Hz Frequency=1078 Hz Q=0. 627538 Q=2. 120758 Peak filter 1 Peak filter 3 Gain=-3. 25 d. B Gain=-5. 75 d. B Frequency=267 Hz Frequency=3461 Hz Q=0. 709020 Q=1. 181551 Plot of the combined filter response. High-shelf filter Plot of the target response. Master gain = -3. 15 d. B Gain=-2. 19 d. B Frequency=5160 Hz Q=2. 839006 Alango Technologies – Sound Equalization Tool 9

Results of ALPEQO Operation Plot of the residuals. Green points indicate residuals within the target limits. Target response limits. FINAL STATE Alango Technologies – Sound Equalization Tool 10

Fitted Sample The newly calculated filter parameters are applied, and the corrected magnitude response is measured. ALPEQO Performance Benchmarks: Fit rate (time/sample): • 10 sec/sample typ. The magnitude response is now within the specified limits! Alango Technologies – Sound Equalization Tool alangotool Success rate (rescued samples): • 90% typ. 11