blender
/
blender
136
406
Fork 589
Code Issues 6.2k Pull Requests 760 Projects 19 Wiki Activity

ImBuf: Precalc subsamples to reduce branching. 

Micro improvement to store delta uv per subsample. This reduces
branching that might happen on the CPU, but also makes it possible
to add other sub-sampling filters as well.

No changes for the end-user.
This commit is contained in:
Jeroen Bakker 2023-01-27 08:39:26 +01:00
parent 2b4bafeac6
commit b3fd169259
Notes: blender-bot 2024-01-08 16:52:56 +01:00 
Referenced by pull request #116628, VSE: Fix various "off by half a pixel" issues in image transform
Referenced by commit 877d9c596a, VSE: Fix various "off by half a pixel" issues in image transform
1 changed files with 28 additions and 27 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

55
source/blender/imbuf/intern/transform.cc
View File

@ -12,6 +12,7 @@
#include "BLI_math_color_blend.h"
#include "BLI_math_vector.hh"
#include "BLI_rect.h"
#include "BLI_vector.hh"
#include "IMB_imbuf.h"
#include "IMB_imbuf_types.h"
@ -39,11 +40,10 @@ struct TransformUserData {
double2 add_y;
struct {
int num;
double2 offset_x;
double2 offset_y;
double2 add_x;
double2 add_y;
/**
* Contains per sub-sample a delta to be added to the uv of the source image buffer.
*/
Vector<double2, 9> delta_uvs;
} subsampling;
/**
@ -96,11 +96,19 @@ struct TransformUserData {
void init_subsampling(const int num_subsamples)
{
subsampling.num = max_ii(num_subsamples, 1);
subsampling.add_x = add_x / (subsampling.num);
subsampling.add_y = add_y / (subsampling.num);
subsampling.offset_x = -add_x * 0.5 + subsampling.add_x * 0.5;
subsampling.offset_y = -add_y * 0.5 + subsampling.add_y * 0.5;
double2 subsample_add_x = add_x / num_subsamples;
double2 subsample_add_y = add_y / num_subsamples;
double2 offset_x = -add_x * 0.5 + subsample_add_x * 0.5;
double2 offset_y = -add_y * 0.5 + subsample_add_y * 0.5;
for (int y : IndexRange(0, num_subsamples)) {
for (int x : IndexRange(0, num_subsamples)) {
double2 delta_uv = -offset_x - offset_y;
delta_uv += x * subsample_add_x;
delta_uv += y * subsample_add_y;
subsampling.delta_uvs.append(delta_uv);
}
}
}
};
@ -526,7 +534,7 @@ class ScanlineProcessor {
*/
void process(const TransformUserData *user_data, int scanline)
{
if (user_data->subsampling.num > 1) {
if (user_data->subsampling.delta_uvs.size() > 1) {
process_with_subsampling(user_data, scanline);
}
else {
@ -595,26 +603,19 @@ class ScanlineProcessor {
sample.clear();
int num_subsamples_added = 0;
double2 subsample_uv_y = uv + user_data->subsampling.offset_y;
for (int subsample_yi : IndexRange(user_data->subsampling.num)) {
UNUSED_VARS(subsample_yi);
double2 subsample_uv = subsample_uv_y + user_data->subsampling.offset_x;
for (int subsample_xi : IndexRange(user_data->subsampling.num)) {
UNUSED_VARS(subsample_xi);
if (!discarder.should_discard(*user_data, subsample_uv)) {
typename Sampler::SampleType sub_sample;
sampler.sample(user_data->src, subsample_uv, sub_sample);
sample.add_subsample(sub_sample, num_subsamples_added);
num_subsamples_added += 1;
}
subsample_uv += user_data->subsampling.add_x;
for (double2 delta_uv : user_data->subsampling.delta_uvs) {
double2 subsample_uv = uv + delta_uv;
if (!discarder.should_discard(*user_data, subsample_uv)) {
typename Sampler::SampleType sub_sample;
sampler.sample(user_data->src, subsample_uv, sub_sample);
sample.add_subsample(sub_sample, num_subsamples_added);
num_subsamples_added += 1;
}
subsample_uv_y += user_data->subsampling.add_y;
}
if (num_subsamples_added != 0) {
float mix_weight = float(num_subsamples_added) /
(user_data->subsampling.num * user_data->subsampling.num);
const float mix_weight = float(num_subsamples_added) /
user_data->subsampling.delta_uvs.size();
channel_converter.mix_and_store(sample, output, mix_weight);
}
uv += user_data->add_x;