void testInterpolatingDelayZero()
{
Jamoma::SampleVector input = {0,1,0,0, 0,0,0,0, 2,0,0,0, 0,0,3,0 };
Jamoma::SampleVector output;
Jamoma::DelayWithLinearInterpolation delay;
delay.size = 0.;
for (auto& in : input) {
Jamoma::Sample out = delay(in);
output.push_back(out);
}
Jamoma::Sample temp = 0.0;
Jamoma::Sample tempExpected = 0.0;
int badSampleCount = 0;
for (int i=0; i < input.size(); i++) {
temp = output[i];
tempExpected = input[i];
if (! mTest->compare(temp, tempExpected, true, 8) ) {
badSampleCount++;
std::cout << "sample " << i << " had a difference of " << std::fabs(temp - tempExpected) << std::endl;
}
}
mTest->TEST_ASSERT("delay.size of 0 produced correct output", badSampleCount == 0);
}
void testImpulseResponse()
{
Jamoma::Dcblock my_dcblock;
Jamoma::UnitImpulse impulse;
impulse.channelCount = 1;
impulse.frameCount = 64;
auto out_samples = my_dcblock( impulse() );
Jamoma::SampleVector expectedIR = generateReferenceImpulseResponse();
int badSampleCount = 0;
Jamoma::Sample temp = 0.0;
Jamoma::Sample tempExpected = 0.0;
for (int i=0; i < expectedIR.size(); i++) {
temp = out_samples[0][0][i];
tempExpected = expectedIR[i];
if ( ! mTest->compare(temp, tempExpected ) ) {
badSampleCount++;
mTest->log("sample %i had a difference of %.10f", i, std::fabs(temp - tempExpected));
}
}
mTest->TEST_ASSERT("Bad Sample Count", badSampleCount == 0);
if (badSampleCount)
mTest->log("the impulse response of my_dcblock has %i bad sample(s)", badSampleCount);
}
// Generate an impulse response from a set of coefficients
// TODO: factor this out into the core as it is generally useful
std::vector<double> generateReferenceImpulseResponse()
{
int N = 64; // size of the impulse response
Jamoma::SampleVector a = {1.0, -1.0}; // feedforward coefficients (numerator)
Jamoma::SampleVector b = {1.0, -0.9997}; // feedback coefficients (denominator)
Jamoma::SampleVector x(N); // input -- feedforward history
Jamoma::SampleVector y(N); // output -- feedback history
std::fill_n(x.begin(), N, 0.0);
std::fill_n(y.begin(), N, 0.0);
x[0] = 1.0;
for (int n=0; n<N; n++) {
int i;
for (i=0; i<a.size(); i++) {
Jamoma::Sample sample = x[n-i];
if (n-i < 0 )
y[n] += 0.0;
else
y[n] += ( a[i] * sample );
}
// ignore b[0] and assume it is normalized to 1.0
for (i=1; i<b.size(); i++) {
if (n-i < 0)
y[n] -= 0.0;
else
y[n] -= ( b[i] * y[n - i] );
}
}
return y;
}
void testDelaySingleSample()
{
Jamoma::SampleVector input = {0,1,0,0, 0,0,0,0, 2,0,0,0, 0,0,3,0 };
Jamoma::SampleVector output;
Jamoma::Delay delay;
delay.size = 3;
for (auto& in : input) {
Jamoma::Sample out = delay(in);
output.push_back(out);
}
mTest->TEST_ASSERT("First Vector Output Correct", output[0] == 0 && output[1] == 0 && output[2] == 0 && output[3] == 0
&& output[4] == 1 && output[5] == 0 && output[6] == 0 && output[7] == 0
&& output[8] == 0 && output[9] == 0 && output[10] == 0 && output[11] == 2
&& output[12] == 0 && output[13] == 0 && output[14] == 0 && output[15] == 0
);
}
void testImpulseResponse()
{
Jamoma::LowpassOnePole my_lowpass;
my_lowpass.coefficient = 0.5;
Jamoma::UnitImpulse impulse;
impulse.channelCount = 1;
impulse.frameCount = 64;
auto out_samples = my_lowpass( impulse() );
// The following impulse was based on the code from jamoma
// implemented in Processing by NW
Jamoma::SampleVector expectedIR = {
0.5,
0.25,
0.125,
0.0625,
0.03125,
0.015625,
0.0078125,
0.00390625,
0.001953125,
9.765625E-4,
4.8828125E-4,
2.44140625E-4,
1.220703125E-4,
6.103515625E-5,
3.0517578125E-5,
1.52587890625E-5,
7.62939453125E-6,
3.814697265625E-6,
1.9073486328125E-6,
9.5367431640625E-7,
4.76837158203125E-7,
2.384185791015625E-7,
1.1920928955078125E-7,
5.9604644775390625E-8,
2.9802322387695312E-8,
1.4901161193847656E-8,
7.450580596923828E-9,
3.725290298461914E-9,
1.862645149230957E-9,
9.313225746154785E-10,
4.6566128730773926E-10,
2.3283064365386963E-10,
1.1641532182693481E-10,
5.820766091346741E-11,
2.9103830456733704E-11,
1.4551915228366852E-11,
7.275957614183426E-12,
3.637978807091713E-12,
1.8189894035458565E-12,
9.094947017729282E-13,
4.547473508864641E-13,
2.2737367544323206E-13,
1.1368683772161603E-13,
5.6843418860808015E-14,
2.8421709430404007E-14,
1.4210854715202004E-14,
7.105427357601002E-15,
3.552713678800501E-15,
1.7763568394002505E-15,
8.881784197001252E-16,
4.440892098500626E-16,
2.220446049250313E-16,
1.1102230246251565E-16,
5.551115123125783E-17,
2.7755575615628914E-17,
1.3877787807814457E-17,
6.938893903907228E-18,
3.469446951953614E-18,
1.734723475976807E-18,
8.673617379884035E-19,
4.3368086899420177E-19,
2.1684043449710089E-19,
1.0842021724855044E-19,
5.421010862427522E-20
};
int badSampleCount = 0;
Jamoma::Sample temp = 0.0;
Jamoma::Sample tempExpected = 0.0;
for (int i = 0; i < expectedIR.size(); i++)
{
temp = out_samples[0][0][i];
tempExpected = expectedIR[i];
if (! mTest->compare(temp, tempExpected) ) {
badSampleCount++;
std::cout << "sample " << i << " had a difference of " << std::fabs(temp - tempExpected) << std::endl;
}
}
std::cout << "the impulse response of my_lowpass (1 pole) has " << badSampleCount << " bad samples" << std::endl;
mTest->TEST_ASSERT("Bad Sample Count", badSampleCount == 0);
// testing range limitations
my_lowpass.coefficient = 0.7;
mTest->TEST_ASSERT("coefficient setting", my_lowpass.coefficient == 0.7);
my_lowpass.coefficient = -1.0;
mTest->TEST_ASSERT("low coefficient limiting", my_lowpass.coefficient == 0.0);
my_lowpass.coefficient = 1.2;
mTest->TEST_ASSERT("high coefficient limiting", my_lowpass.coefficient == 1.0);
// Testing coefficient setting via frequency
// The following Max 7 patcher can be convenient as an interactive reference for the
// relationship between frequency response and coefficient
/*
<pre><code>
----------begin_max5_patcher----------
2115.3oc6asrjiZCEcc2eEJplEcRY6fdAlYURVjJUMS9BRkpKYiraxfAG.mz
8L0ju8nm.1FroaaHdQVLF8BoiNb0U2G87k6uCtH6YQAD7dvuAt6tub+c2oaR
0vc152A2vedYBuPOL3FQQAes.NwzWo34Rc6XOOWawQ5VxV7GSCpZrn7kDgtc
WKY6JSDkkurUXVeHD761tR2sINU1odIw1F2xKW9Tb55GyEKKMuBwmMyaB.6o
9kfzkIy7plms4hBQZIuLNKcuWiV+ZteZt3FnoWcjpwud+8pelzSNJU72xM+Q
Tj34s4.zLvTfrzCSwx076.jYHJhIK7tUHv2CnTjm2rusEpze9kSkntoRJSSD
gdFR7Xt7xokSH5H2ysskoCpziOogzCcdekd7I9CpzypjL4jzFcPdMzwD.bAO
c8aVR.gYlGdGQLqxx2v0i1uZZx4aDkh7GEo7EFr4cZZjxLLNB0CdDeEE2Prt
D2PCp3FK3MItwlqE2PgCj3VmzTmmJGVc5Lq.2qkln3A8T41jrx+oE1fE5Zju
KJNak5bPQ7m06ejGlNwMW4qiSs2zp2TyqgKqITuBTH0vcgZ4MeGSx1mO1lEm
ZlIFxMW4REFhD9BQhd+8q70owk6hDfGh9op6jhx1vkvnZT+bt3O2IRW9B3ge
4yUi5.Fuw5JIx81.E6VT0lxJjpO.x9Vlkjka5vaFch6m.l7.Q81QI19T7xOk
JLerTZTbcn2lOVKollk5j0k8lDmJZzopJOutaIbWH0o00auJNQpxS2AuLKZQ
cOFNR9h7b82DuFHx1mHUKDgwdrV5sAlxVez7Jkl9jH2xXtY.XtGUqqvUXtqP
fqfuq.yUf5JPbEvtBHWA2L6lX275lV2r5lT2b5lRbCj4.lCWNX4PkCTNL4fj
CQN.YwiENVzXAiEKVnXQhEHVbXggEEVPXwfEBVDz3ZP45aW9ZiklaWayR6NO
OwdJjzPaF5P8R0eQ0GmbeP0CSZVDrBuFyqqnQH9CvFeuk89AXi4TeLtV3aJh
faHgY50J94cTGceZvz+dxd5keJRq5cJVaU8Ti4QSMptmZXjoAlgL2LjPyPPd
ysuucdT37n8QSpwrcUTgcQgxmvp0FJeBqf.T9DVgDn7IrBPP80XNbAkOgUvC
JeBqPIT9DVCVnp.rFzJ.XvCwAHBt4Gir7Xol+Ge9P1119K11MMatOpSMfAJo
G7b0ujv+WEXkX3oOO8exIiSJF+1EO.+9UzDOZGV3wHCpEdzflV30au1olypi
tgv3tXI7vxR3SyRCnilLuwzQSlwLU1E5l4k6v3xrMajxcGI.rpx.23TvSetE
BidhXyzW1v52igMr1ffQc9A26JJgKWKTa6J+AU.2YJmwsXBsu5ABI5wyFY0.
dyBaiiXCJGgoM4Hb3LV+3H7bbMoNpjDqMRZXiiGx3n6qmjPFIO+QmjZ8z1vd
oqyqlNIoAKr2emzLq1t+j9ZtkQdWk5pfW84Gjc6Z7EiL.2h10sFRyjAayRDf
jr+dqbv.qoxGSDjvK+1CpKv8l8q01g46caZZYUfolOyavuUAAdWax4jfgUio
8PsQbmv56sJAFqKwCV.D6hldGBzVlyHCrNyfyPSCn4kD7nXdoSR.2HlL2l1W
JDQK3K+DXYlX0p3kwxAAdXwin1RCHw6xUVfCZxLsYqYWGRrWhPl2iCIdWuqP
9lofVuBAGNdWg32URlGtqPVykNbrmPA+Qu1DJPjqfPgMRkcKTb4ed6Tkv3nQ
vY7yMiFgtRpDoxyqqediNA+DtWhivAmKwQTT64MhuYahNuQckxnB0HDEmNWQ
CQdgXcFTzkw4KS5Nrnlvw8VBK59cz+nhROcHQOHBjsEUTZi61sGtajAZiZMS
JIX8JuDl3cCTlqXeOHxjUh57UXVGn5VJ65Ck14ehHv5MC2U7WIrKK0Dx2W8q
Nl45hX0OHcQjd+xrWhM0Tdpt8olwnP1IChqF5.ndbP2KqphTUsYpvTT0QU6X
cEroBQWg.fibHg6RyCE+ehlGeS3+5slmNSY81m3EBvC47nXdZQm4rNUonOQ5
MTDnJ5d2RZiFnTz7V0EgtTUQVSEQUmBIUrj9Ln+DW1szI25nbacFsQUGpf1o
UVlBsSrrrOzN0xxpj2gLuK5jIMcnxLjKcoFUnSsoI0jyTSdTOVabmYBUq4wl
3QiZHaJLo5J1LX5qqXye4b.rZ0UYyDre1TGcUQcXyuQrMRTxiSZyhwpvHnBr
QpzHMcW+XdLO4UZL4Q+6sZXIkpeD1VpLZD7ERGAew+5QdQwKKmsU4SYwxmDa
3uG3nyh30o.45Jx4kY4uGDIRyjaRSkenT9ccsTREf.31Hc+qBoqPmzwRd9Kc
P0jyS017hgCeabcS2zzuo9j6A+kPqWdU66+AnHaW9R2dw8mlKnF.RRtTwnxs
XyAQ1aPOEGIo9ltSDEWn7vHpamq5KdB75AdXdiFdX8AOGPhCId7o2V7yAhFs
iG73AGzMF8biINqt5473Ib7NdguwNtStsvC0+l53kJkw2RvgdaAmaKcgz9n6
gLd3AG1C7PGuyVj9nKTY1H.ON3g1W7fFG7zGcyGPhCp7SefC8lBN3w6lTbut
oHXbszfdtOW9mAOahipibl5uoHlJ7Kz.cjBBo5+TPz0pWnq6N.cN8U3W2N.E
FpBVDMP67muD5LasytCL95w2t8ujd4ZQgF7R2p+CSP7BlnqFmZppc6GlK9qX
230nBxyktlVJ8KcWtwc2ms+2UDtIKRjmtK15moj1t25V7A9WV4.cIeqwYYIA
shuKobeVsoy1RJ6iRm0AeLd8SprYZ1aSreTxERR5.GvgajSXriHzAo4q2+u.
80vVA
-----------end_max5_patcher-----------
</code></pre>
*/
my_lowpass.sampleRate = 96000;
my_lowpass.frequency = 1000.0; // hz
mTest->TEST_ASSERT("coefficient for 1K cutoff @ 96K fs", mTest->compare((double)my_lowpass.coefficient, 0.06335217076965427));
my_lowpass.frequency = 4000.0; // hz
mTest->TEST_ASSERT("coefficient for 1K cutoff @ 96K fs", mTest->compare((double)my_lowpass.coefficient, 0.23032864479520065));
my_lowpass.sampleRate = 44100;
my_lowpass.frequency = 1000.0; // hz
mTest->TEST_ASSERT("coefficient for 1K cutoff @ 44.1K fs", mTest->compare((double)my_lowpass.coefficient, 0.132787865213287));
my_lowpass.frequency = 4000.0; // hz
mTest->TEST_ASSERT("coefficient for 1K cutoff @ 44.1K fs", mTest->compare((double)my_lowpass.coefficient, 0.43441043913502342));
}