Re-integrate export-optimization branch.

Export now happens directly to file (unless normalizing is required), and can be easily optimized even further.
The Session process connection is still broken during export (as it was before this commit also).


git-svn-id: svn://localhost/ardour2/branches/3.0@6401 d708f5d6-7413-0410-9779-e7cbd77b26cf

1 files changed, 92 insertions, 0 deletions

diff --git a/libs/audiographer/src/gdither/gdither.h b/libs/audiographer/src/gdither/gdither.h
new file mode 100644
index 0000000000..d2b2657f32
--- /dev/null
+++ b/libs/audiographer/src/gdither/gdither.h
@@ -0,0 +1,92 @@
+/*
+ *  Copyright (C) 2002 Steve Harris <steve@plugin.org.uk>
+ *
+ *  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.
+ *
+ */
+
+#ifndef GDITHER_H
+#define GDITHER_H
+
+#ifdef __cplusplus
+extern "C" {
+#endif
+
+#include "gdither_types.h"
+
+/* Create and initialise a state structure, takes a dither type, a number of
+ * channels and a bit depth as input
+ *
+ * The Dither type is one of
+ *
+ *   GDitherNone - straight nearest neighbour rounding. Theres no pressing
+ *   reason to do this at 8 or 16 bit, but you might want to at 24, for some
+ *   reason. At the lest it will save you writing int->float conversion code,
+ *   which is arder than it sounds.
+ *
+ *   GDitherRect - mathematically most accurate, lowest noise floor, but not
+ *   that good for audio. It is the fastest though.
+ *
+ *   GDitherTri - a happy medium between Rectangular and Shaped, reasonable
+ *   noise floor, not too obvious, quite fast.
+ *
+ *   GDitherShaped - should have the least audible impact, but has the highest
+ *   noise floor, fairly CPU intensive. Not advisible if your going to apply
+ *   any frequency manipulation afterwards.
+ *
+ * channels, sets the number of channels in the output data, output data will
+ * be written interleaved into the area given to gdither_run(). Set to 1
+ * if you are not working with interleaved buffers.
+ *
+ * bit depth, sets the bit width of the output sample data, it can be one of:
+ *
+ *   GDither8bit   - 8 bit unsiged
+ *   GDither16bit  - 16 bit signed
+ *   GDither32bit  - 24+bits in upper bits of a 32 bit word
+ *   GDitherFloat  - IEEE floating point (32bits)
+ *   GDitherDouble - Double precision IEEE floating point (64bits)
+ *
+ * dither_depth, set the number of bits before the signal will be truncated to,
+ * eg. 16 will produce an output stream with 16bits-worth of signal. Setting to
+ * zero or greater than the width of the output format will dither to the
+ * maximum precision allowed by the output format.
+ */
+GDither gdither_new(GDitherType type, uint32_t channels,
+
+                    GDitherSize bit_depth, int dither_depth);
+
+/* Frees memory used by gdither_new.
+ */
+void gdither_free(GDither s);
+
+/* Applies dithering to the supplied signal.
+ *
+ * channel is the channel number you are processing (0 - channles-1), length is
+ * the length of the input, in samples, x is the input samples (float), y is
+ * where the output samples will be written, it should have the approaprate
+ * type for the chosen bit depth
+ */
+void gdither_runf(GDither s, uint32_t channel, uint32_t length,
+		   float const *x, void *y);
+
+/* see gdither_runf, vut input argument is double format */
+void gdither_run(GDither s, uint32_t channel, uint32_t length,
+		   double const *x, void *y);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif