some new source

git-svn-id: svn://localhost/ardour2/branches/3.0@8523 d708f5d6-7413-0410-9779-e7cbd77b26cf

1 files changed, 311 insertions, 0 deletions

diff --git a/libs/ardour/panner_shell.cc b/libs/ardour/panner_shell.cc
new file mode 100644
index 0000000000..c6cbbe36ab
--- /dev/null
+++ b/libs/ardour/panner_shell.cc
@@ -0,0 +1,311 @@
+/*
+    Copyright (C) 2004-2011 Paul Davis
+
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+
+*/
+
+#include <inttypes.h>
+
+#include <cmath>
+#include <cerrno>
+#include <fstream>
+#include <cstdlib>
+#include <string>
+#include <cstdio>
+#include <locale.h>
+#include <unistd.h>
+#include <float.h>
+#include <iomanip>
+
+#include <glibmm.h>
+
+#include "pbd/cartesian.h"
+#include "pbd/convert.h"
+#include "pbd/error.h"
+#include "pbd/failed_constructor.h"
+#include "pbd/xml++.h"
+#include "pbd/enumwriter.h"
+
+#include "evoral/Curve.hpp"
+
+#include "ardour/audio_buffer.h"
+#include "ardour/audio_buffer.h"
+#include "ardour/automatable.h"
+#include "ardour/buffer_set.h"
+#include "ardour/pannable.h"
+#include "ardour/panner.h"
+#include "ardour/panner_manager.h"
+#include "ardour/panner_shell.h"
+#include "ardour/runtime_functions.h"
+#include "ardour/session.h"
+#include "ardour/utils.h"
+
+#include "i18n.h"
+
+#include "pbd/mathfix.h"
+
+using namespace std;
+using namespace ARDOUR;
+using namespace PBD;
+
+PannerShell::PannerShell (string name, Session& s, boost::shared_ptr<Pannable> p)
+	: SessionObject (s, name)
+        , _pannable (p)
+{
+	set_name (name);
+}
+
+PannerShell::~PannerShell ()
+{
+}
+
+void
+PannerShell::configure_io (ChanCount in, ChanCount out)
+{
+        uint32_t nouts = out.n_audio();
+        uint32_t nins = in.n_audio();
+
+	/* if new and old config don't need panning, or if
+	   the config hasn't changed, we're done.
+	*/
+
+	if (_panner && _panner->in().n_audio() == nins && _panner->out().n_audio() == nouts) {
+                return;
+	}
+
+	if (nouts < 2 || nins == 0) {
+		/* no need for panning with less than 2 outputs or no inputs */
+		if (_panner) {
+                        _panner.reset ();
+			Changed (); /* EMIT SIGNAL */
+		}
+		return;
+	}
+
+        PannerInfo* pi = PannerManager::instance().select_panner (in, out);
+
+        if (pi == 0) {
+                abort ();
+        }
+
+        _panner.reset (pi->descriptor.factory (_pannable, _session.get_speakers()));
+
+        Changed (); /* EMIT SIGNAL */
+}
+
+XMLNode&
+PannerShell::get_state (void)
+{
+	return state (true);
+}
+
+XMLNode&
+PannerShell::state (bool full)
+{
+	XMLNode* node = new XMLNode ("PannerShell");
+
+        if (_panner) {
+		node->add_child_nocopy (_panner->state (full));
+	}
+
+	return *node;
+}
+
+int
+PannerShell::set_state (const XMLNode& node, int version)
+{
+	XMLNodeList nlist = node.children ();
+	XMLNodeConstIterator niter;
+	const XMLProperty *prop;
+	LocaleGuard lg (X_("POSIX"));
+
+        _panner.reset ();
+
+	for (niter = nlist.begin(); niter != nlist.end(); ++niter) {
+
+		if ((*niter)->name() == X_("Panner")) {
+
+			if ((prop = (*niter)->property (X_("type")))) {
+
+                                list<PannerInfo*>::iterator p;
+                                PannerManager& pm (PannerManager::instance());
+
+				for (p = pm.panner_info.begin(); p != pm.panner_info.end(); ++p) {
+					if (prop->value() == (*p)->descriptor.name) {
+
+						/* note that we assume that all the stream panners
+						   are of the same type. pretty good
+						   assumption, but it's still an assumption.
+						*/
+                                                        
+                                                _panner.reset ((*p)->descriptor.factory (_pannable, _session.get_speakers ()));
+                                                
+						if (_panner->set_state (**niter, version) == 0) {
+                                                        return -1;
+                                                }
+
+						break;
+					}
+				}
+
+				if (p == pm.panner_info.end()) {
+					error << string_compose (_("Unknown panner plugin \"%1\" found in pan state - ignored"),
+							  prop->value())
+					      << endmsg;
+				}
+
+			} else {
+				error << _("panner plugin node has no type information!")
+				      << endmsg;
+				return -1;
+			}
+		}
+	}
+
+	return 0;
+}
+
+
+void
+PannerShell::distribute_no_automation (BufferSet& inbufs, BufferSet& outbufs, pframes_t nframes, gain_t gain_coeff)
+{
+	if (outbufs.count().n_audio() == 0) {
+		// Don't want to lose audio...
+		assert(inbufs.count().n_audio() == 0);
+		return;
+	}
+
+	if (outbufs.count().n_audio() == 1) {
+
+                /* just one output: no real panning going on */
+
+		AudioBuffer& dst = outbufs.get_audio(0);
+
+		if (gain_coeff == 0.0f) {
+
+			/* gain was zero, so make it silent */
+
+			dst.silence (nframes);
+
+		} else if (gain_coeff == 1.0f){
+
+			/* mix all input buffers into the output */
+
+			// copy the first
+			dst.read_from(inbufs.get_audio(0), nframes);
+
+			// accumulate starting with the second
+			if (inbufs.count().n_audio() > 0) {
+				BufferSet::audio_iterator i = inbufs.audio_begin();
+				for (++i; i != inbufs.audio_end(); ++i) {
+					dst.merge_from(*i, nframes);
+				}
+			}
+
+		} else {
+
+			/* mix all buffers into the output, scaling them all by the gain */
+
+			// copy the first
+			dst.read_from(inbufs.get_audio(0), nframes);
+
+			// accumulate (with gain) starting with the second
+			if (inbufs.count().n_audio() > 0) {
+				BufferSet::audio_iterator i = inbufs.audio_begin();
+				for (++i; i != inbufs.audio_end(); ++i) {
+					dst.accumulate_with_gain_from(*i, nframes, gain_coeff);
+				}
+			}
+
+		}
+
+		return;
+	}
+
+        /* multiple outputs ... we must have a panner */
+
+        assert (_panner);
+
+	/* setup silent buffers so that we can mix into the outbuffers (slightly suboptimal -
+           better to copy the first set of data then mix after that, but hey, its 2011)
+         */
+
+	for (BufferSet::audio_iterator b = outbufs.audio_begin(); b != outbufs.audio_end(); ++b) {
+		(*b).silence (nframes);
+	}
+
+        _panner->distribute (inbufs, outbufs, gain_coeff, nframes);
+}
+
+void
+PannerShell::run (BufferSet& inbufs, BufferSet& outbufs, framepos_t start_frame, framepos_t end_frame, pframes_t nframes)
+{
+	if (outbufs.count().n_audio() == 0) {
+		// Failing to deliver audio we were asked to deliver is a bug
+		assert(inbufs.count().n_audio() == 0);
+		return;
+	}
+
+	if (outbufs.count().n_audio() == 1) {
+
+                /* one output only: no panner */
+
+		AudioBuffer& dst = outbufs.get_audio(0);
+
+		// FIXME: apply gain automation?
+
+		// copy the first
+		dst.read_from (inbufs.get_audio(0), nframes);
+
+		// accumulate starting with the second
+		BufferSet::audio_iterator i = inbufs.audio_begin();
+		for (++i; i != inbufs.audio_end(); ++i) {
+			dst.merge_from (*i, nframes);
+		}
+
+		return;
+	}
+
+	// More than 1 output
+
+        AutoState as = _panner->automation_state ();
+
+	// If we shouldn't play automation defer to distribute_no_automation
+
+	if (!(as & Play || ((as & Touch) && !_panner->touching()))) {
+
+		// Speed quietning
+		gain_t gain_coeff = 1.0;
+                
+		if (fabsf(_session.transport_speed()) > 1.5f && Config->get_quieten_at_speed ()) {
+			gain_coeff = speed_quietning;
+		}
+
+		distribute_no_automation (inbufs, outbufs, nframes, gain_coeff);
+
+	} else {
+
+                /* setup the terrible silence so that we can mix into the outbuffers (slightly suboptimal -
+                   better to copy the first set of data then mix after that, but hey, its 2011)
+                */
+                for (BufferSet::audio_iterator i = outbufs.audio_begin(); i != outbufs.audio_end(); ++i) {
+                        i->silence(nframes);
+                }
+                
+                _panner->distribute_automated (inbufs, outbufs, start_frame, end_frame, nframes, _session.pan_automation_buffer());
+        }
+}
+