summaryrefslogtreecommitdiffstats
path: root/skia/ext/convolver.h
Commit message (Collapse)AuthorAgeFilesLines
* Integration of most changes from the GoogleTV project around the ↵brettw@chromium.org2011-01-301-1/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | convolver/scaler. This contains the following improvements: - Adding a few extra convolution filters on top of the existing LANCZOS3 (used internally in Chrome), and BOX (used in unit tests): - LANCZOS2: a variation of LANCZOS3 except that the windowed function is limited to the [-2:2] range. - HAMMING1: this uses a Hamming window using the [-1:-1] range. If we define the zoom down factor to z, and w the size of the window, the actual cost of each filter (CPU wise) is proportional to (w * 2 * z + 1). So, if we look at what happens when you zoom down by a factor of 4 (as often found when creating thumbnails), the cost would be 25 for LANCZOS3, 17 for LANCZOS2, and 9 for HAMMING. As a result, HAMMING1 can end up be roughly three times as fast as the typical LANCZOS3. In terms of visual quality, HAMMING1 will be obviously worse than filters that have a larger window. The motivation of this change is that not all processors are equally equipped, and while LANCZOS3 does provide good quality, it will be completely inadequate in speed on slower processors (as found on Google TV), and it would be worth trading some visual quality for speed. Because the definitions of what is acceptable from one platform to another will differ, this change adds generic enums describing various trade offs between quality and speed. And depending on the platform, these would then be mapped to different filters. This change does not contain the other changes made to the all the call sites to transform LANCZOS3 to the appropriate enum. Another CL will have to be checked in for the policy definition. - Improvements in speed by around 10% (the actual speed up depends on the parameters of the scale (scale ratios, sizes of images), as well as the actual processor on which this is run on. The 10% was measured on scale down of 1920x1080 images to 1920/4x1080/4 using the LANCZOS3 filter on a 32bit Atom based using the image_operations_bench. Actual numbers for a 64bit processor are discussed below. This optimization attempts to basically eliminate all zeroes on each side of the filter_size, since it is very likely that the calculated window will go one fraction of a pixel outside of the window where the function is actuall not zero. In many cases, this means it gets rid the convolution by one point. So, using the math above, (w * 2 * z + 1) will have 1 subtracted. The code though is generic and will get rid of more points if possible. - To measure speed, a small utility image_operations_bench was added. Its purpose is to simply measure speed of the actual speed of the convolution without any regards to the actual data. Run with --help for a list of options. The actual measured number is in MB/s (source MB + dest MB / time). The following numbers were found on a 64 bit Release build on a z600: | zero optimization | Filter | no | yes | Hamming1 | 459 | 495 | Lanczos2 | 276 | 294 | Lanczos3 | 202 | 207 | The command line was: for i in HAMMING1 LANCZOS2 LANCZOS3 ; do echo $i; out/Release/image_operations_bench -source 1920x1080 -destination 480x270 -m $i -iter 50 ; done The actual improvements for the zero optimization mentioned above are much more prevalent on a 32bit Atom. - Commented that there is half-pixel error inside the code in image_operations. Because this would effectively changes the results of many scales that are used in win_layout tests, this would effectively break them. As a result, the change here only adds comments about what needs to be changed, but does not fix the issue itself. A subsequent change will remove the comments and enable the fix, and also adds the corrected reference images used for the test. See bug 69999: http://code.google.com/p/chromium/issues/detail?id=69999 - Enhanced the convolver to support arbitrary strides, instead of the hard coded 4 * width. This value is correct on most platforms, but is not on GoogleTV since buffers allocated need to be 32 pixel multiples to exploit HW capabilities. - Added numerous unit tests to cover the new filters as well as adding other ones that are more rigourous than the existing ones. Such a test is the reason, we have found the half pixel error mentioned above. TEST=This was tested against the existing unit tests, and the added unit tests on a 64 bit Linux platform. The tests were then ran under valgrind to check for possible memory leaks/ and errors. The tests do come out clean (except the preexisting file descriptor 'leaks' coming from other tests that are linked with test_shell_tests Actual credit to most of the actual changes go to various contributors of the Google TV team. Note that there are two types of optimizations that are possible beyond these changes that are not done here: 1/ Use the fact that the filter coefficients will be periodic to reduce the cost of calculating the coefficients (though typically in the noise), but rather when the convolution is done to decrease cache misses on the coefficients. Experiments showed that on an Atom, this can yield 5 % improvement. 2/ This code is the prime target for the use of SIMD instructions. BUG=47447, 62820, 69999 Patch by evannier@google.com Original review http://codereview.chromium.org/5575010/ git-svn-id: svn://svn.chromium.org/chrome/trunk/src@73110 0039d316-1c4b-4281-b951-d872f2087c98
* Start sorting methods in class declarations.erg@google.com2011-01-071-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | A lot of our headers are a mess and aren't organized. Impose the following order on files in the base/ directory: class Blah { each public/protected/private section: typedefs; enums; static constants; ctors; dtors; methods; overridden virtual methods; data members; }; BUG=68682 TEST=compiles Review URL: http://codereview.chromium.org/6081007 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@70749 0039d316-1c4b-4281-b951-d872f2087c98
* FBTF: Moves code to the headers.erg@google.com2010-09-281-2/+2
| | | | | | | | | | | | | One of the big things is starting to move/declare ctors/dtors that derive from RefCounted<> to/in the implementation file. (Saves 4 megabytes from libglue.a alone. 1 meg off libbrowser.a. Hundred of kilobyte savings in a large number of .a files; only libmedia.a grew and it's only 100k.) BUG=none TEST=compiles Review URL: http://codereview.chromium.org/3452030 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@60863 0039d316-1c4b-4281-b951-d872f2087c98
* `#pragma once` for app, base, chrome, gfx, ipc, net, skia, viewsthakis@chromium.org2010-07-261-0/+1
| | | | | | | | | BUG=50273 TEST=everything still builds, build is 10% faster on windows, same speed on mac/linux TBR: erg git-svn-id: svn://svn.chromium.org/chrome/trunk/src@53716 0039d316-1c4b-4281-b951-d872f2087c98
* Move functions from skia/ext to app/gfx where possible: most of skia_utils.* ↵pkasting@chromium.org2009-09-241-8/+8
| | | | | | | | | | | | | and image_operations.* can be moved because they are not used by WebKit code. This also fixes the spelling of "Convolusion" to "Convolution" and updates some copyrights. This is a re-do of r26975, this time with WebKit update and some fixes to compile on Mac and Linux. BUG=none TEST=none git-svn-id: svn://svn.chromium.org/chrome/trunk/src@27031 0039d316-1c4b-4281-b951-d872f2087c98
* Reverting 26975.pkasting@chromium.org2009-09-231-8/+8
| | | | | | Review URL: http://codereview.chromium.org/222011 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@26979 0039d316-1c4b-4281-b951-d872f2087c98
* Move functions from skia/ext to app/gfx where possible: most of skia_utils.* ↵pkasting@chromium.org2009-09-231-8/+8
| | | | | | | | | | | | and image_operations.* can be moved because they are not used by WebKit code. This also fixes the spelling of "Convolusion" to "Convolution" and updates some copyrights. BUG=none TEST=none Review URL: http://codereview.chromium.org/207059 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@26975 0039d316-1c4b-4281-b951-d872f2087c98
* Remove a bunch of base dependencies from skia/ext. The only nontrivial change isbrettw@chromium.org2009-01-141-2/+0
| | | | | | | | in bitmap_platform_device_mac. The refcounting now matches the way the Windows file works. Review URL: http://codereview.chromium.org/17627 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@8015 0039d316-1c4b-4281-b951-d872f2087c98
* Move Image operations and convolver to the skia namespace and clean up a few ↵brettw@google.com2008-12-121-11/+13
| | | | | | | | (but not all) base types. Review URL: http://codereview.chromium.org/13726 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6921 0039d316-1c4b-4281-b951-d872f2087c98
* Reverting 6709,6708,6706.brettw@google.com2008-12-101-11/+13
| | | | | | Review URL: http://codereview.chromium.org/13345 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6710 0039d316-1c4b-4281-b951-d872f2087c98
* Use crappy apple define instead of OS_ defines since those don't exist for Skia.brettw@google.com2008-12-101-1/+1
| | | | | | Review URL: http://codereview.chromium.org/13709 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6709 0039d316-1c4b-4281-b951-d872f2087c98
* Remove base dependency from image_operations and associated unittests. I ↵brettw@google.com2008-12-101-12/+10
| | | | | | | | | also renamed the namespace from "gfx" to "skia". This also adds back the image_operations_unittest to a project (this had gotten lost in my move). Review URL: http://codereview.chromium.org/13143 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6706 0039d316-1c4b-4281-b951-d872f2087c98
* Move convolver and image_operations from base/gfx to skia/ext. This is justbrettw@google.com2008-12-031-0/+137
| | | | | | | | like my previous change except does no namespace renaming and doesn't touch skia_utils. Review URL: http://codereview.chromium.org/13080 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6290 0039d316-1c4b-4281-b951-d872f2087c98
* Revert my skia file moves because of layout test failures.brettw@google.com2008-12-031-137/+0
| | | | | | Review URL: http://codereview.chromium.org/12892 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6266 0039d316-1c4b-4281-b951-d872f2087c98
* Move convolver, image_operations, and skia_utils from base/gfx to skia/ext.brettw@google.com2008-12-021-0/+137
This changes the namespace in those files from "gfx" to "skia". I split skia_utils into two parts, the Windows specific part is now in a separate file called skia_utils_win. There were several obsolete includes of these headers which I removed. I also removed img_resize_perftest which isn't used and has bitrotted. Review URL: http://codereview.chromium.org/12842 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6248 0039d316-1c4b-4281-b951-d872f2087c98
generated by cgit v1.1 at 2025-02-01 01:03:39 +0000