Limiter

From Avisynth wiki
Jump to navigationJump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

The standard known as Rec.601 or CCIR-601 (Studio Encoding Parameters of Digital Television...) defines standard quantization levels: for luminance, black is 16 and peak white is 235. This is the normal luminance range. For chrominance, the range is 16 to 240, or 128±112. "The signal may occasionally excurse beyond" these ranges according to Rec.601.BT.601-7 (Annex 1, Table 3)

Luminance or chrominance values outside their ranges are known to cause problems with some TV sets, so Limiter was created to clip them (or clamp them - both terms are used) to the ranges [min_luma, max_luma] and [min_chroma, max_chroma], which you may specify, but by default are the Rec.601 values listed above.

Limiter can also colorize out-of-range pixels (with show) to warn of a potential problem. See the example images below. Naturally, this option should not be enabled for final output.

AVS+ By default all parameters are non-autoscaling – they must be scaled to the target bit depth.
By using parameter paramscale the range parameters are treated to be of 8 bits range and will be scaled to the current bit depth automatically

Syntax and Parameters

Limiter(clip clip [, int min_luma, int max_luma, int min_chroma, int max_chroma, string show ] )
AVS+ Limiter(clip clip [, float min_luma, float max_luma, float min_chroma, float max_chroma, string show, bool paramscale ] )

clip  clip = (required)
Source clip. YUV only.
AVS+ also supports YUVA.
int  min_luma = 16d
Y (luma) lower bound.
int  max_luma = 235d
Y (luma) upper bound.
int  min_chroma = 16
U & V (chroma) lower bound.
int  max_chroma = 240d
U & V (chroma) upper bound.
string  show = (empty)
Colorizes the pixels outside the specified [min_X, max_X] range.
If set, show must take one of the following values:
show Effect
"luma" colorizes out of bound luma
"luma_grey" colorizes out of bound luma, and makes the remaining pixels grey
"chroma" colorizes out of bound chroma
"chroma_grey" colorizes out of bound chroma, and makes the remaining pixels grey
The colorizing is done as follows:
  • YUY2: j,k=1,2 (chroma shared between two horizontal pixels p1 and p2: Y1UY2V)
  • YV24: j,k=1 (no chroma shared)
  • YV12: j,k=11,12,21,22 (chroma shared between 2x2 pixels Y11UY12V: Y21UY22V)
Pixel Condition Colorization
 show="luma"  show="luma_grey"
Yj < min_luma   red (pj)   red (pj)
Yj > max_luma   green (pj)   green (pj)
Yj < min_luma and Yk > max_luma   yellow (pj and pk)   puke (pj)  olive (pk)
 show="chroma"  show="chroma_grey"
U < min_chroma   yellow   yellow
U > max_chroma   blue
V < min_chroma   cyan
V > max_chroma   red
U < min_chroma and V < min_chroma   green
U > max_chroma and V < min_chroma   teal
U < min_chroma and V > max_chroma   orange
U > max_chroma and V > max_chroma   magenta
AVS+ bool  paramscale = false
If set, minimum/maximum luma/chroma values are treated as they were in 8 bit range (but non-integer values are allowed), limiter will autoscale them to the current bit depth of the clip.
Conversion uses the bit-shift method for integer bit depth conversion (from 8 to 8-16 bits). For 32 bit float clips the conversion is value/255.0 for Y (luma) channel and (value-128)/255.0 for U/V chroma channels. Do not use paramscale when you want to limit 32 bit float chroma exactly to +/-0.5, give the parameters directly instead.
Default: does not scale at all, parameters are used as-is. Parameters now are of float type to handle 32 bit float values..


Examples

# 1. Original image
FFmpegSource2("sintel.mp4")
Trim(4685, 4685)
# 2. Exaggerated luma and chroma range
Levels(16, 1.0, 255-16, 0, 255, coring=false)
Tweak(sat=2.0, coring=false)
# (insert Limiter here)
ColorYUV(analyze=true)
# 3. Limiter("luma")
[...]
Limiter(16, 235, 16, 240, "luma")
ColorYUV(analyze=true)
# 4. Limiter("luma_grey")
[...]
Limiter(16, 235, 16, 240, "luma_grey")
ColorYUV(analyze=true)
# 5. Limiter("chroma")
[...]
Limiter(16, 235, 16, 240, "chroma")
ColorYUV(analyze=true)
# 6. Limiter("chroma_grey")
[...]
Limiter(16, 235, 16, 240, "chroma_grey")
ColorYUV(analyze=true)
# 7. Putting it all together
[...]
StackHorizontal(
\    Last
\  , StackVertical(
\       Histogram("levels", 0.5).Crop(Width, 0, 256, 256).AddBorders(0, 0, 0, Height/2-256)
\     , Histogram("color2"     ).Crop(Width, 0, 256, 256).AddBorders(0, 0, 0, Height/2-256)
\    )
\  , StackVertical(
\       BicubicResize(Width/2, Height/2).Limiter(16, 235, 16, 240, "luma_grey")
\     , BicubicResize(Width/2, Height/2).Limiter(16, 235, 16, 240, "chroma_grey")
\    )
\ )


Changes

AVS+ r2724 Added support for float.

Added paramscale

AVS+ Added support for 8-16bit color spaces.
v2.60 Added support for YV24.
v2.56 Added show to show out of bound luma/chroma.