Audio: SRC: Add conversions from 8 - 192 kHz to 64 kHz

[singalsu]

No description provided.

[lgirdwood]

LGTM

[singalsu]

@lgirdwood I added new coefficients into the latest push. To minimize the changes I had to use another computer to generate with Matlab the coefficients since they were previously generated with it. With Octave with other FIR tool in practice all coefficients would have changed.

[singalsu]

Test results for new conversions with testbench and src_test(32, 32, [8000 11025 12000 16000 22050 24000 32000 44100 48000 88200 96000 176400 192000], 64000, 1, 1);:

SRC test result: Gain dB
in \ out,     64.0
     8.0,    -0.00
    11.0,     0.00
    12.0,    -0.00
    16.0,     0.00
    22.1,    -0.00
    24.0,    -0.00
    32.0,    -0.00
    44.1,     0.00
    48.0,     0.00
    88.2,    -0.00
    96.0,     0.00
   176.4,    -0.00
   192.0,    -0.00


SRC test result: Frequency response +/- X.XX dB (YY.Y kHz)
in \ out,         64.0
     8.0,  0.00 ( 3.6)
    11.0,  0.00 ( 5.0)
    12.0,  0.00 ( 5.4)
    16.0,  0.00 ( 7.2)
    22.1,  0.00 ( 9.9)
    24.0,  0.00 (10.8)
    32.0,  0.00 (14.4)
    44.1,  0.00 (19.8)
    48.0,  0.00 (21.6)
    88.2,  0.00 (28.7)
    96.0,  0.00 (28.7)
   176.4,  0.00 (28.7)
   192.0,  0.00 (28.7)


SRC test result: Frequency response -3 dB 0 - X kHz
in \ out,     64.0
     8.0,     3.73
    11.0,     5.14
    12.0,     5.60
    16.0,     7.46
    22.1,    10.28
    24.0,    11.19
    32.0,    14.92
    44.1,    20.56
    48.0,    22.39
    88.2,    29.85
    96.0,    29.85
   176.4,    29.85
   192.0,    29.85


SRC test result: Worst-case THD+N vs. frequency
in \ out,     64.0
     8.0,   -92.30
    11.0,   -96.61
    12.0,   -89.72
    16.0,   -94.83
    22.1,   -98.71
    24.0,   -90.51
    32.0,   -94.05
    44.1,   -93.25
    48.0,   -98.83
    88.2,   -93.97
    96.0,  -109.18
   176.4,   -93.48
   192.0,  -171.58


SRC test result: Dynamic range dB (CCIR-RMS)
in \ out,     64.0
     8.0,   169.07
    11.0,   167.85
    12.0,   170.18
    16.0,   184.30
    22.1,   165.68
    24.0,   153.32
    32.0,   186.03
    44.1,   149.15
    48.0,   178.81
    88.2,   153.37
    96.0,   189.28
   176.4,   154.33
   192.0,   190.68


SRC test result: Attenuation of alias products dB
in \ out,     64.0
     8.0,        -
    11.0,        -
    12.0,        -
    16.0,        -
    22.1,        -
    24.0,        -
    32.0,        -
    44.1,        -
    48.0,        -
    88.2,   -82.71
    96.0,   -63.47
   176.4,   -81.58
   192.0,   -69.72


SRC test result: Attenuation of image products dB
in \ out,     64.0
     8.0,   -92.92
    11.0,   -97.73
    12.0,   -92.23
    16.0,  -100.60
    22.1,   -94.81
    24.0,   -93.83
    32.0,  -100.93
    44.1,  -105.36
    48.0,   -99.17
    88.2,        -
    96.0,        -
   176.4,        -
   192.0,        -


SRC test result: chirp/gain/FR/THD+N/DR/AAP/AIP
in \ out,           64.0
     8.0,  0/0/0/0/0/x/0
    11.0,  0/0/0/0/0/x/0
    12.0,  0/0/0/0/0/x/0
    16.0,  0/0/0/0/0/x/0
    22.1,  0/0/0/0/0/x/0
    24.0,  0/0/0/0/0/x/0
    32.0,  0/0/0/0/0/x/0
    44.1,  0/0/0/0/0/x/0
    48.0,  0/0/0/0/0/x/0
    88.2,  0/0/0/0/0/0/x
    96.0,  0/0/0/0/0/0/x
   176.4,  0/0/0/0/0/0/x
   192.0,  0/0/0/0/0/0/x


Number of passed tests = 78
Number of failed tests = 0
Number of non-applicable tests = 13
Number of skipped tests = 0

Test passed.

I'll change this to non-draft.

[golowanow]

this change resolves a problem we've observed at 64 kHz rate set on a PCM

[lgirdwood]

SOFCI TEST

[singalsu]

this change resolves a problem we've observed at 64 kHz rate set on a PCM

Excellent, thanks for testing it!