Talk:Blur
From Avisynth wiki
(Difference between revisions)
Raffriff42 (Talk | contribs) (MMX precision?) |
Raffriff42 (Talk | contribs) m (ack) |
||
| (4 intermediate revisions by 2 users not shown) | |||
| Line 1: | Line 1: | ||
| − | + | [[Category:Talk]] | |
| + | ==== MMX ==== | ||
:''A Known issue, with the MMX routines is the lack of full 8 bit precision in the calculations. '' | :''A Known issue, with the MMX routines is the lack of full 8 bit precision in the calculations. '' | ||
: O RLY? Can't find citation. 18:26, 21 September 2014 (CEST) | : O RLY? Can't find citation. 18:26, 21 September 2014 (CEST) | ||
| + | :: Ian fixed that apparently a long time ago (v2.58). See http://avisynth2.cvs.sourceforge.net/viewvc/avisynth2/avisynth/src/filters/focus.cpp?r1=1.11&r2=1.12 . You can remove that comment. [[User:Admin|Admin]] 18:42, 21 September 2014 (CEST) | ||
| + | |||
| + | ==== Kernel tech stuff deleted ==== | ||
| + | :I removed the following as I think it of interest to developers only. | ||
| + | <div {{NotaBeneWidthIndent|48|2|none}} > | ||
| + | This is a simple 3x3-kernel blurring filter using the kernel [(1−1/2^amount)/2, 1/2^amount, (1−1/2^amount)/2]. The largest allowable argument for Blur is log2(3) (which is about 1.58), which corresponds to a ({{Frac|1|3}}, {{Frac|1|3}}, {{Frac|1|3}}) kernel. A value of 1.0 gets you a ({{Frac|1|4}}, {{Frac|1|2}}, {{Frac|1|4}}) kernel. | ||
| + | </div> | ||
| + | :please feel free to add it back if it is important to the article. --[[User:Raffriff42|Raffriff42]] 06:49, 28 January 2016 (CET) | ||
| + | ::I see that you are not a developer ;) They too read the documentation. You might argue whether it is the right place for it, but such stuff should be some where. More generally, what i truly hate is many developers don't mention or reference the algorithms they use to implement a filter. Makes it very hard to understand what a filter does, let alone fix bugs in it. [[User:Admin|Admin]] 18:48, 16 March 2016 (CET) | ||
| + | :::OK that makes a lot of sense. I like that you put the information under "Developer notes." --[[User:Raffriff42|Raffriff42]] 20:45, 16 March 2016 (CET) | ||
Latest revision as of 20:45, 16 March 2016
[edit] MMX
- A Known issue, with the MMX routines is the lack of full 8 bit precision in the calculations.
- O RLY? Can't find citation. 18:26, 21 September 2014 (CEST)
- Ian fixed that apparently a long time ago (v2.58). See http://avisynth2.cvs.sourceforge.net/viewvc/avisynth2/avisynth/src/filters/focus.cpp?r1=1.11&r2=1.12 . You can remove that comment. Admin 18:42, 21 September 2014 (CEST)
[edit] Kernel tech stuff deleted
- I removed the following as I think it of interest to developers only.
This is a simple 3x3-kernel blurring filter using the kernel [(1−1/2^amount)/2, 1/2^amount, (1−1/2^amount)/2]. The largest allowable argument for Blur is log2(3) (which is about 1.58), which corresponds to a (1/3, 1/3, 1/3) kernel. A value of 1.0 gets you a (1/4, 1/2, 1/4) kernel.
- please feel free to add it back if it is important to the article. --Raffriff42 06:49, 28 January 2016 (CET)
- I see that you are not a developer ;) They too read the documentation. You might argue whether it is the right place for it, but such stuff should be some where. More generally, what i truly hate is many developers don't mention or reference the algorithms they use to implement a filter. Makes it very hard to understand what a filter does, let alone fix bugs in it. Admin 18:48, 16 March 2016 (CET)
- OK that makes a lot of sense. I like that you put the information under "Developer notes." --Raffriff42 20:45, 16 March 2016 (CET)
- I see that you are not a developer ;) They too read the documentation. You might argue whether it is the right place for it, but such stuff should be some where. More generally, what i truly hate is many developers don't mention or reference the algorithms they use to implement a filter. Makes it very hard to understand what a filter does, let alone fix bugs in it. Admin 18:48, 16 March 2016 (CET)
