1) SSHiQ(int rY, int rC, int tY, int tC, int str, boolean HQY, boolean HQC,
boolean field)
rY and rC are respectively the radii for luminance and chrominance (how far from the current pixel neighbours are used).
The greater the more smooth.
tY and tC are thresholds (details sensibility). The greater the more smooth (not exactly the same as fpr other
SSHiQ filters).
str is the strength (only used in HiQ mode), ie how much of the smoothed version should be used at best for calculating the new
pixel
HQY and HQC tell whether to use the HiQ mode, where the edge information automatically reduce the strength value.
(HQ == 0) ? smoothed_pixel : ( (strength-edge)*smoothed_pixel + (256-(strength-edge))*pixel_in )/256
field allows you to process independently fields (risks of worse blurring/color bleeding)
2) SmoothUV (used to reduce rainbows, as done by smartsmoothIQ)
SmoothUV(int radius, int threshold, boolean field)
radius is the range of operation
threshold is the one used in smartsmoothIQ
field is the same as above
This filter doesn't process the luma at all, and can be seen as a shorter version of
SSHiQ(rY, radius, -1, threshold, 255, false, false, field)