mirror/lmms
1
0 Fork 0
mirror of https://github.com/LMMS/lmms.git synced 2026-10-21 06:14:47 -04:00
Code Issues Packages Projects Releases Wiki Activity

Revert "Revert "Extensions for peak controller. Fulfills request 144.""

Browse Source 
Now that we solved the problem with wrong default values in newly
introduced models when loading older projects we can safely implement
the new PeakController features.

This reverts commit 9cb52ccf0d.
This commit is contained in:
Tobias Doerffel committed 2014-01-20 21:28:10 +01:00
1 parent 44d0e7a48f
commit 7644d2c559
6 files changed
+76 -11
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

plugins/peak_controller_effect/peak_controller_effect.cpp
+31 -6
View File
@@ -64,6 +64,7 @@ PeakControllerEffect::PeakControllerEffect(
m_peakControls( this ),
m_lastSample( 0 ),
m_lastRMS( -1 ),
m_lastRMSavail(false),
m_autoController( NULL )
{
m_autoController = new PeakController( engine::getSong(), this );
@@ -83,13 +84,19 @@ PeakControllerEffect::~PeakControllerEffect()
}
}
//! returns 1.0f if val > 0.0f, -1.0 else
inline float my_sign(float val) { return -1.0f + 2.0f * (val > 0.0f); }
//! if val >= 0.0f, returns sqrtf(val), else: -sqrtf(-val)
inline float sqrt_neg(float val) {
return sqrtf(fabs(val)) * my_sign(val);
}
bool PeakControllerEffect::processAudioBuffer( sampleFrame * _buf,
const fpp_t _frames )
{
PeakControllerEffectControls & c = m_peakControls;
// This appears to be used for determining whether or not to continue processing
// audio with this effect
if( !isEnabled() || !isRunning() )
@@ -101,7 +108,12 @@ bool PeakControllerEffect::processAudioBuffer( sampleFrame * _buf,
double sum = 0;
for( int i = 0; i < _frames; ++i )
{
sum += _buf[i][0]*_buf[i][0] + _buf[i][1]*_buf[i][1];
float sign_0 = (c.m_absModel.value())
? 1.0f : my_sign(_buf[i][0]);
float sign_1 = (c.m_absModel.value())
? 1.0f : my_sign(_buf[i][1]);
sum += _buf[i][0]*_buf[i][0]*sign_0
+ _buf[i][1]*_buf[i][1]*sign_1;
}
if( c.m_muteModel.value() )
@@ -112,19 +124,22 @@ bool PeakControllerEffect::processAudioBuffer( sampleFrame * _buf,
}
}
float curRMS = sqrtf( sum / _frames );
float curRMS = sqrt_neg( sum / _frames );
const float origRMS = curRMS;
if( m_lastRMS < 0 )
if( !m_lastRMSavail )
{
m_lastRMSavail = true;
m_lastRMS = curRMS;
}
const float v = ( curRMS >= m_lastRMS ) ?
c.m_attackModel.value() :
c.m_decayModel.value();
const float a = sqrtf( sqrtf( v ) );
const float a = sqrt_neg( sqrt_neg( v ) );
curRMS = (1-a)*curRMS + a*m_lastRMS;
m_lastSample = c.m_baseModel.value() + c.m_amountModel.value()*curRMS;
const float amount = c.m_amountModel.value() * c.m_amountMultModel.value();
m_lastSample = c.m_baseModel.value() + amount*curRMS;
m_lastRMS = curRMS;
// on greater buffer sizes our LP is updated less frequently, therfore
@@ -138,6 +153,16 @@ bool PeakControllerEffect::processAudioBuffer( sampleFrame * _buf,
//checkGate( out_sum / _frames );
// finally, mute the output if wanted
// TODO: avoid clips?
if( c.m_muteOutputModel.value() )
{
for( int i = 0; i < _frames; ++i )
{
_buf[i][0] = _buf[i][1] = 0.0f;
}
}
return isRunning();
}