wirtz
/
Teensy-reSID
mirror of https://github.com/FrankBoesing/Teensy-reSID.git
1
0
Fork 0
Code Issues Releases Wiki Activity

Reduce sampling-methods (won't work anyway)

Browse Source
master
Frank Bösing 9 years ago
parent 54b7f82887
commit 6b5d9e4609
4 changed files with 6 additions and 298 deletions
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Show Stats Download Patch File Download Diff File
  1. 1
      reSID/filter.cc
  2. 287
      reSID/sid.cc
  3. 13
      reSID/sid.h
  4. 3
      reSID/siddefs.h

1
reSID/filter.cc
Unescape View File

@ -148,6 +148,7 @@ Filter::Filter()
f0 = f0_6581;
f0_points = f0_points_6581;
f0_count = sizeof(f0_points_6581)/sizeof(*f0_points_6581);
//Serial.print(f0_count);
set_w0();
set_Q();
}

287
reSID/sid.cc
Unescape View File

@ -44,9 +44,6 @@ const int SID::FIXP_MASK = 0xffff;
// ----------------------------------------------------------------------------
SID::SID()
{
// Initialize pointers.
sample = 0;
fir = 0;
voice[0].set_sync_source(&voice[2]);
voice[1].set_sync_source(&voice[0]);
@ -66,8 +63,6 @@ SID::SID()
// ----------------------------------------------------------------------------
SID::~SID()
{
delete[] sample;
delete[] fir;
}
@ -429,6 +424,7 @@ void SID::enable_external_filter(bool enable)
// I0() computes the 0th order modified Bessel function of the first kind.
// This function is originally from resample-1.5/filterkit.c by J. O. Smith.
// ----------------------------------------------------------------------------
/*
float SID::I0(float x)
{
// Max error acceptable in I0.
@ -448,7 +444,7 @@ float SID::I0(float x)
return sum;
}
*/
// ----------------------------------------------------------------------------
// Setting of SID sampling parameters.
@ -476,14 +472,6 @@ bool SID::set_sampling_parameters(float clock_freq, sampling_method method,
float sample_freq, float pass_freq,
float filter_scale)
{
// Check resampling constraints.
if (method == SAMPLE_RESAMPLE_INTERPOLATE || method == SAMPLE_RESAMPLE_FAST)
{
// Check whether the sample ring buffer would overfill.
if (FIR_N*clock_freq/sample_freq >= RINGSIZE) {
return false;
}
}
// The default passband limit is 0.9*sample_freq/2 for sample
// frequencies below ~ 44.1kHz, and 20kHz for higher sample frequencies.
@ -515,89 +503,6 @@ bool SID::set_sampling_parameters(float clock_freq, sampling_method method,
sample_offset = 0;
sample_prev = 0;
// FIR initialization is only necessary for resampling.
if (method != SAMPLE_RESAMPLE_INTERPOLATE && method != SAMPLE_RESAMPLE_FAST)
{
delete[] sample;
delete[] fir;
sample = 0;
fir = 0;
return true;
}
const float pi = 3.1415926535897932385;
// 16 bits -> -96dB stopband attenuation.
const float A = -20.0*log10(1.0/(1 << 16));
// A fraction of the bandwidth is allocated to the transition band,
float dw = (1.0 - 2.0*pass_freq/sample_freq)*pi;
// The cutoff frequency is midway through the transition band.
float wc = (2.0*pass_freq/sample_freq + 1.0)*pi/2.0;
// For calculation of beta and N see the reference for the kaiserord
// function in the MATLAB Signal Processing Toolbox:
// http://www.mathworks.com/access/helpdesk/help/toolbox/signal/kaiserord.html
const float beta = 0.1102*(A - 8.7);
const float I0beta = I0(beta);
// The filter order will maximally be 124 with the current constraints.
// N >= (96.33 - 7.95)/(2.285*0.1*pi) -> N >= 123
// The filter order is equal to the number of zero crossings, i.e.
// it should be an even number (sinc is symmetric about x = 0).
int N = int((A - 7.95)/(2.285*dw) + 0.5);
N += N & 1;
float f_samples_per_cycle = sample_freq/clock_freq;
float f_cycles_per_sample = clock_freq/sample_freq;
// The filter length is equal to the filter order + 1.
// The filter length must be an odd number (sinc is symmetric about x = 0).
fir_N = int(N*f_cycles_per_sample) + 1;
fir_N |= 1;
// We clamp the filter table resolution to 2^n, making the fixpoint
// sample_offset a whole multiple of the filter table resolution.
int res = method == SAMPLE_RESAMPLE_INTERPOLATE ?
FIR_RES_INTERPOLATE : FIR_RES_FAST;
int n = (int)ceil(log(res/f_cycles_per_sample)/log(2));
fir_RES = 1 << n;
// Allocate memory for FIR tables.
delete[] fir;
fir = new short[fir_N*fir_RES];
// Calculate fir_RES FIR tables for linear interpolation.
for (int i = 0; i < fir_RES; i++) {
int fir_offset = i*fir_N + fir_N/2;
float j_offset = float(i)/fir_RES;
// Calculate FIR table. This is the sinc function, weighted by the
// Kaiser window.
for (int j = -fir_N/2; j <= fir_N/2; j++) {
float jx = j - j_offset;
float wt = wc*jx/f_cycles_per_sample;
float temp = jx/(fir_N/2);
float Kaiser =
fabs(temp) <= 1 ? I0(beta*sqrt(1 - temp*temp))/I0beta : 0;
float sincwt =
fabs(wt) >= 1e-6 ? sin(wt)/wt : 1;
float val =
(1 << FIR_SHIFT)*filter_scale*f_samples_per_cycle*wc/pi*sincwt*Kaiser;
fir[fir_offset + j] = short(val + 0.5);
}
}
// Allocate sample buffer.
if (!sample) {
sample = new short[RINGSIZE*2];
}
// Clear sample buffer.
for (int j = 0; j < RINGSIZE*2; j++) {
sample[j] = 0;
}
sample_index = 0;
return true;
}
@ -779,10 +684,6 @@ int SID::clock(cycle_count& delta_t, short* buf, int n, int interleave)
return clock_fast(delta_t, buf, n, interleave);
case SAMPLE_INTERPOLATE:
return clock_interpolate(delta_t, buf, n, interleave);
case SAMPLE_RESAMPLE_INTERPOLATE:
return clock_resample_interpolate(delta_t, buf, n, interleave);
case SAMPLE_RESAMPLE_FAST:
return clock_resample_fast(delta_t, buf, n, interleave);
}
}
@ -872,188 +773,4 @@ int SID::clock_interpolate(cycle_count& delta_t, short* buf, int n,
}
// ----------------------------------------------------------------------------
// SID clocking with audio sampling - cycle based with audio resampling.
//
// This is the theoretically correct (and computationally intensive) audio
// sample generation. The samples are generated by resampling to the specified
// sampling frequency. The work rate is inversely proportional to the
// percentage of the bandwidth allocated to the filter transition band.
//
// This implementation is based on the paper "A Flexible Sampling-Rate
// Conversion Method", by J. O. Smith and P. Gosset, or rather on the
// expanded tutorial on the "Digital Audio Resampling Home Page":
// http://www-ccrma.stanford.edu/~jos/resample/
//
// By building shifted FIR tables with samples according to the
// sampling frequency, this implementation dramatically reduces the
// computational effort in the filter convolutions, without any loss
// of accuracy. The filter convolutions are also vectorizable on
// current hardware.
//
// Further possible optimizations are:
// * An equiripple filter design could yield a lower filter order, see
// http://www.mwrf.com/Articles/ArticleID/7229/7229.html
// * The Convolution Theorem could be used to bring the complexity of
// convolution down from O(n*n) to O(n*log(n)) using the Fast Fourier
// Transform, see http://en.wikipedia.org/wiki/Convolution_theorem
// * Simply resampling in two steps can also yield computational
// savings, since the transition band will be wider in the first step
// and the required filter order is thus lower in this step.
// Laurent Ganier has found the optimal intermediate sampling frequency
// to be (via derivation of sum of two steps):
// 2 * pass_freq + sqrt [ 2 * pass_freq * orig_sample_freq
// * (dest_sample_freq - 2 * pass_freq) / dest_sample_freq ]
//
// NB! the result of right shifting negative numbers is really
// implementation dependent in the C++ standard.
// ----------------------------------------------------------------------------
RESID_INLINE
int SID::clock_resample_interpolate(cycle_count& delta_t, short* buf, int n,
int interleave)
{
int s = 0;
for (;;) {
cycle_count next_sample_offset = sample_offset + cycles_per_sample;
cycle_count delta_t_sample = next_sample_offset >> FIXP_SHIFT;
if (delta_t_sample > delta_t) {
break;
}
if (s >= n) {
return s;
}
for (int i = 0; i < delta_t_sample; i++) {
clock();
sample[sample_index] = sample[sample_index + RINGSIZE] = output();
++sample_index;
sample_index &= 0x3fff;
}
delta_t -= delta_t_sample;
sample_offset = next_sample_offset & FIXP_MASK;
int fir_offset = sample_offset*fir_RES >> FIXP_SHIFT;
int fir_offset_rmd = sample_offset*fir_RES & FIXP_MASK;
short* fir_start = fir + fir_offset*fir_N;
short* sample_start = sample + sample_index - fir_N + RINGSIZE;
// Convolution with filter impulse response.
int v1 = 0;
for (int j = 0; j < fir_N; j++) {
v1 += sample_start[j]*fir_start[j];
}
// Use next FIR table, wrap around to first FIR table using
// previous sample.
if (++fir_offset == fir_RES) {
fir_offset = 0;
--sample_start;
}
fir_start = fir + fir_offset*fir_N;
// Convolution with filter impulse response.
int v2 = 0;
for (int j = 0; j < fir_N; j++) {
v2 += sample_start[j]*fir_start[j];
}
// Linear interpolation.
// fir_offset_rmd is equal for all samples, it can thus be factorized out:
// sum(v1 + rmd*(v2 - v1)) = sum(v1) + rmd*(sum(v2) - sum(v1))
int v = v1 + (fir_offset_rmd*(v2 - v1) >> FIXP_SHIFT);
v >>= FIR_SHIFT;
// Saturated arithmetics to guard against 16 bit sample overflow.
//(fb)
asm ("ssat %0, #16, %1" : "=r" (v) : "r" (v));
/*
const int half = 1 << 15;
if (v >= half) {
v = half - 1;
}
else if (v < -half) {
v = -half;
}
*/
buf[s++*interleave] = v;
}
for (int i = 0; i < delta_t; i++) {
clock();
sample[sample_index] = sample[sample_index + RINGSIZE] = output();
++sample_index;
sample_index &= 0x3fff;
}
sample_offset -= delta_t << FIXP_SHIFT;
delta_t = 0;
return s;
}
// ----------------------------------------------------------------------------
// SID clocking with audio sampling - cycle based with audio resampling.
// ----------------------------------------------------------------------------
RESID_INLINE
int SID::clock_resample_fast(cycle_count& delta_t, short* buf, int n,
int interleave)
{
int s = 0;
for (;;) {
cycle_count next_sample_offset = sample_offset + cycles_per_sample;
cycle_count delta_t_sample = next_sample_offset >> FIXP_SHIFT;
if (delta_t_sample > delta_t) {
break;
}
if (s >= n) {
return s;
}
for (int i = 0; i < delta_t_sample; i++) {
clock();
sample[sample_index] = sample[sample_index + RINGSIZE] = output();
++sample_index;
sample_index &= 0x3fff;
}
delta_t -= delta_t_sample;
sample_offset = next_sample_offset & FIXP_MASK;
int fir_offset = sample_offset*fir_RES >> FIXP_SHIFT;
short* fir_start = fir + fir_offset*fir_N;
short* sample_start = sample + sample_index - fir_N + RINGSIZE;
// Convolution with filter impulse response.
int v = 0;
for (int j = 0; j < fir_N; j++) {
v += sample_start[j]*fir_start[j];
}
v >>= FIR_SHIFT;
// Saturated arithmetics to guard against 16 bit sample overflow.
//(fb)
asm ("ssat %0, #16, %1" : "=r" (v) : "r" (v));
/*
const int half = 1 << 15;
if (v >= half) {
v = half - 1;
}
else if (v < -half) {
v = -half;
}
*/
buf[s++*interleave] = v;
}
for (int i = 0; i < delta_t; i++) {
clock();
sample[sample_index] = sample[sample_index + RINGSIZE] = output();
++sample_index;
sample_index &= 0x3fff;
}
sample_offset -= delta_t << FIXP_SHIFT;
delta_t = 0;
return s;
}
RESID_NAMESPACE_STOP

13
reSID/sid.h
Unescape View File

@ -89,17 +89,13 @@ public:
int output(int bits);
protected:
static float I0(float x);
RESID_INLINE int clock_fast(cycle_count& delta_t, short* buf, int n,
int interleave);
RESID_INLINE int clock_interpolate(cycle_count& delta_t, short* buf, int n,
int interleave);
RESID_INLINE int clock_resample_interpolate(cycle_count& delta_t, short* buf,
int n, int interleave);
RESID_INLINE int clock_resample_fast(cycle_count& delta_t, short* buf,
int n, int interleave);
Voice voice[3];
Voice voice[3];
Filter filter;
ExternalFilter extfilt;
Potentiometer potx;
@ -133,11 +129,6 @@ protected:
int fir_N;
int fir_RES;
// Ring buffer with overflow for contiguous storage of RINGSIZE samples.
short* sample;
// FIR_RES filter tables (FIR_N*FIR_RES).
short* fir;
};
RESID_NAMESPACE_STOP

3
reSID/siddefs.h
Unescape View File

@ -68,8 +68,7 @@ typedef sound_sample fc_point[2];
//enum chip_model { MOS6581, MOS8580 };
enum sampling_method { SAMPLE_FAST, SAMPLE_INTERPOLATE,
SAMPLE_RESAMPLE_INTERPOLATE, SAMPLE_RESAMPLE_FAST };
enum sampling_method { SAMPLE_FAST, SAMPLE_INTERPOLATE};
extern "C"
{

Write Preview
Loading…
Cancel
Save