D627: Memory usage optimization for the compositor.

The compostor used a fixed size of 4 floats to hold pixel data. this
patch will select size of a pixel based on its type.
It uses 1 float for Value, 3 float for vector and 4 floats for color
data types.

When benchmarking on shots (opening shot of caminandes) we get a
reduction of memory of 30% and a tiny speedup as less data
transformations needs to take place (but these are negligable.

More information of the patch can be found on
https://developer.blender.org/D627 and
http://wiki.blender.org/index.php/Dev:Ref/Proposals/Compositor2014_p1.1_TD

Developers: jbakker & mdewanchand
Thanks for Sergey for his indept review.

1 files changed, 2 insertions, 2 deletions

diff --git a/source/blender/compositor/operations/COM_GaussianAlphaYBlurOperation.cpp b/source/blender/compositor/operations/COM_GaussianAlphaYBlurOperation.cpp
index ab97c8b0d13..bb5b3c044af 100644
--- a/source/blender/compositor/operations/COM_GaussianAlphaYBlurOperation.cpp
+++ b/source/blender/compositor/operations/COM_GaussianAlphaYBlurOperation.cpp
@@ -108,11 +108,11 @@ void GaussianAlphaYBlurOperation::executePixel(float output[4], int x, int y, vo
 	float multiplier_accum = 0.0f;
 
 	/* dilate */
-	float value_max = finv_test(buffer[(x * 4) + (y * 4 * bufferwidth)], do_invert); /* init with the current color to avoid unneeded lookups */
+	float value_max = finv_test(buffer[(x) + (y * bufferwidth)], do_invert); /* init with the current color to avoid unneeded lookups */
 	float distfacinv_max = 1.0f; /* 0 to 1 */
 
 	for (int ny = ymin; ny < ymax; ny += step) {
-		int bufferindex = ((xmin - bufferstartx) * 4) + ((ny - bufferstarty) * 4 * bufferwidth);
+		int bufferindex = ((xmin - bufferstartx)) + ((ny - bufferstarty) * bufferwidth);
 
 		const int index = (ny - y) + this->m_filtersize;
 		float value = finv_test(buffer[bufferindex], do_invert);