DctFilter
Raffriff42 (Talk | contribs) m (use Template:Frac (meh)) |
m (category) |
||
| Line 1: | Line 1: | ||
| − | {{ | + | {{FilterCat4|External_filters|Plugins|Denoisers|Spatial_denoisers}} |
<h1>DctFilter</h1> | <h1>DctFilter</h1> | ||
<h2>Abstract</h2> | <h2>Abstract</h2> | ||
Revision as of 16:49, 21 September 2014
Contents |
DctFilter
Abstract
author: Tom Barry
version: 0.0.1.4
download: V 0.0.1.4
category: Misc Plugins
requirements:
- YV12 & YUY2 Colorspace and screen multiples of 16x16
- SSEMMX support
Description
DctFilter is an experimental filter that, for each 8x8 block, will do a Discrete Cosine Transform (DCT), scale down the selected frequency values,
and then reverse the process with an Inverse Discrete Cosine Transform (IDCT).Usage
In your Avisynth file use commands similar to
Avisource("D:\wherever\myfile.avi")
DctFilter(1,1,1,1,1,1,.5,0)
Parameters
There are 8 positional floating point parameters, all of which must be specified as in the range (0.0 <= x <= 1.0). These correspond to scaling factors for the 8 rows and colums of the 8x8 DCT blocks. The leftmost parm corresponds to the top row, left column. This would be the DC component of the transform and should always be left as 1.0.
In the above example the highest frequency components in each row and column would be zeroed while the 2nd highest would be cut in half.
The row & column parms are multiplied together to get the scale factor for each of the 64 values in a block. So if the top left value was V[0,0] then in the example above the we would scale row 6, col 6 (V[6,6]) by .5 * .5 = .25.
Note that while they look like floating point parms above they really now only have 3 bit accuracy so the only actual values used are 0, 1/8, 1/4, 3/8 ... 1.0. But you can specify any value and it will be rounded to the nearest one.
DctFilterD - New with v 0.0.1.4
You can instead use DctFilterD(DiagCt). This works similar to above but will zero out DiagCt number of the lower right diagonals of the DCT, leaving other values unchanged. In an 8x8 DCT result matrix there are 15 possible diagonals. (visualize a chess board) The DiagCt parm must be an integer from 1-14 saying how many of these diagonals must be zeroed, starting from the lower right hand corner.
Comments
This new function DctFilterD(DiagCt) works on diagonals. It is simpler and either zeroes diagonals or leaves them alone.
In a 8x8 DCT result there are 15 possible diagonals (visualize a chess board). So if you specify:
DctFilterD(4)
then the 4 diagonals in the lower right corner of the DCT result will be set to 0. The DiagCt parm may be any integer value from 1-14.
I haven't tested this much but, like my results with custom quant tables, it appears if you set DiagCt very large you will start to get edge noise, something like ringing.
