Skip to content

utils

Functions:

RektPartial

RektPartial(rekt_partial: Callable[P, R])

Bases: Generic[P, R]

Class decorator that wraps the rekt_partial function and extends its functionality.

It is not meant to be used directly.

Methods:

  • __call__
  • abs

    Creates a rectangular mask to apply fixes only within the masked area,

  • rel
Source code
343
344
def __init__(self, rekt_partial: Callable[P, R]) -> None:
    self._func = rekt_partial

__call__

__call__(*args: args, **kwargs: kwargs) -> R
Source code
346
347
def __call__(self, *args: P.args, **kwargs: P.kwargs) -> R:
    return self._func(*args, **kwargs)

abs

abs(
    clip: VideoNode,
    func: Callable[Concatenate[VideoNode, P0], VideoNode],
    width: int,
    height: int,
    offset_x: int = 0,
    offset_y: int = 0,
    *args: args,
    **kwargs: kwargs
) -> ConstantFormatVideoNode

Creates a rectangular mask to apply fixes only within the masked area, significantly speeding up filters like anti-aliasing and scaling.

Parameters:

  • clip

    (VideoNode) –

    The source video clip to which the mask will be applied.

  • func

    (Callable[Concatenate[VideoNode, P0], VideoNode]) –

    The function to be applied within the masked area.

  • width

    (int) –

    The width of the rectangular mask.

  • height

    (int) –

    The height of the rectangular mask.

  • offset_x

    (int, default: 0 ) –

    The horizontal offset of the mask from the top-left corner, defaults to 0.

  • offset_y

    (int, default: 0 ) –

    The vertical offset of the mask from the top-left corner, defaults to 0.

Returns:

  • ConstantFormatVideoNode

    A new clip with the applied mask.

Source code
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
def abs(
    self,
    clip: vs.VideoNode,
    func: Callable[Concatenate[vs.VideoNode, P0], vs.VideoNode],
    width: int,
    height: int,
    offset_x: int = 0,
    offset_y: int = 0,
    *args: P0.args,
    **kwargs: P0.kwargs,
) -> ConstantFormatVideoNode:
    """
    Creates a rectangular mask to apply fixes only within the masked area,
    significantly speeding up filters like anti-aliasing and scaling.

    :param clip:            The source video clip to which the mask will be applied.
    :param func:            The function to be applied within the masked area.
    :param width:           The width of the rectangular mask.
    :param height:          The height of the rectangular mask.
    :param offset_x:        The horizontal offset of the mask from the top-left corner, defaults to 0.
    :param offset_y:        The vertical offset of the mask from the top-left corner, defaults to 0.
    :return:                A new clip with the applied mask.
    """
    nargs = (clip, func, offset_x, clip.width - width - offset_x, offset_y, clip.height - height - offset_y)
    return self._func(*nargs, *args, **kwargs)  # type: ignore

rel

rel(*args: args, **kwargs: kwargs) -> R
Source code
349
350
def rel(self, *args: P.args, **kwargs: P.kwargs) -> R:
    return self._func(*args, **kwargs)

freeze_replace_squaremask

freeze_replace_squaremask(
    mask: VideoNode,
    insert: VideoNode,
    mask_params: tuple[int, int, int, int],
    frame: int,
    frame_range: tuple[int, int],
) -> ConstantFormatVideoNode
Source code
271
272
273
274
275
276
277
278
279
280
281
282
def freeze_replace_squaremask(
    mask: vs.VideoNode,
    insert: vs.VideoNode,
    mask_params: tuple[int, int, int, int],
    frame: int,
    frame_range: tuple[int, int],
) -> ConstantFormatVideoNode:
    start, end = frame_range

    masked_insert = replace_squaremask(mask[frame], insert[frame], mask_params)

    return insert_clip(mask, masked_insert * (end - start + 1), start)

max_planes

max_planes(
    *_clips: VideoNode | Iterable[VideoNode], resizer: KernelLike = Bilinear
) -> ConstantFormatVideoNode
Source code
50
51
52
53
54
55
56
57
58
59
60
61
def max_planes(
    *_clips: vs.VideoNode | Iterable[vs.VideoNode], resizer: KernelLike = Bilinear
) -> ConstantFormatVideoNode:
    clips = flatten_vnodes(_clips)

    assert check_variable_format(clips, max_planes)

    resizer = Kernel.ensure_obj(resizer, max_planes)

    width, height, fmt = clips[0].width, clips[0].height, clips[0].format.replace(subsampling_w=0, subsampling_h=0)

    return ExprOp.MAX.combine(split(resizer.scale(clip, width, height, format=fmt)) for clip in clips)

normalize_mask

normalize_mask(mask: VideoNode, clip: VideoNode) -> ConstantFormatVideoNode
normalize_mask(
    mask: Callable[[VideoNode, VideoNode], VideoNode],
    clip: VideoNode,
    ref: VideoNode,
) -> ConstantFormatVideoNode
normalize_mask(
    mask: EdgeDetectT | RidgeDetectT,
    clip: VideoNode,
    *,
    ridge: bool = ...,
    **kwargs: Any
) -> ConstantFormatVideoNode
normalize_mask(
    mask: GeneralMask, clip: VideoNode, ref: VideoNode
) -> ConstantFormatVideoNode
normalize_mask(
    mask: GenericMaskT,
    clip: VideoNode,
    ref: VideoNode | None = ...,
    *,
    ridge: bool = ...,
    **kwargs: Any
) -> ConstantFormatVideoNode
normalize_mask(
    mask: GenericMaskT,
    clip: VideoNode,
    ref: VideoNode | None = None,
    *,
    ridge: bool = False,
    **kwargs: Any
) -> ConstantFormatVideoNode
Source code
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
def normalize_mask(
    mask: GenericMaskT, clip: vs.VideoNode, ref: vs.VideoNode | None = None, *, ridge: bool = False, **kwargs: Any
) -> ConstantFormatVideoNode:
    if isinstance(mask, (str, type)):
        return normalize_mask(EdgeDetect.ensure_obj(mask, normalize_mask), clip, ref, ridge=ridge, **kwargs)

    if isinstance(mask, EdgeDetect):
        if ridge and isinstance(mask, RidgeDetect):
            cmask = mask.ridgemask(clip, **kwargs)
        else:
            cmask = mask.edgemask(clip, **kwargs)
    elif isinstance(mask, GeneralMask):
        cmask = mask.get_mask(clip, ref)
    elif callable(mask):
        if ref is None:
            raise CustomValueError("This mask function requires a ref to be specified!")

        cmask = mask(clip, ref)
    else:
        cmask = mask

    return depth(cmask, clip, range_in=ColorRange.FULL, range_out=ColorRange.FULL)

region_abs_mask

region_abs_mask(
    clip: VideoNode, width: int, height: int, left: int = 0, top: int = 0
) -> ConstantFormatVideoNode
Source code
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
def region_abs_mask(
    clip: vs.VideoNode, width: int, height: int, left: int = 0, top: int = 0
) -> ConstantFormatVideoNode:
    assert check_variable_format(clip, region_rel_mask)

    def _crop(w: int, h: int) -> ConstantFormatVideoNode:
        cropped = vs.core.std.CropAbs(clip, width, height, left, top)
        return vs.core.std.AddBorders(cropped, left, w - width - left, top, h - height - top)

    if 0 in {clip.width, clip.height}:
        if complexpr_available:
            return norm_expr(
                clip, _get_region_expr(clip, left, left + width, top, top + height, 0, True), func=region_rel_mask
            )

        return vs.core.std.FrameEval(clip, lambda f, n: _crop(f.width, f.height), clip)

    return region_rel_mask(clip, left, clip.width - width - left, top, clip.height - height - top)

region_rel_mask

region_rel_mask(
    clip: VideoNode,
    left: int = 0,
    right: int = 0,
    top: int = 0,
    bottom: int = 0,
) -> ConstantFormatVideoNode
Source code
84
85
86
87
88
89
90
91
92
93
94
def region_rel_mask(
    clip: vs.VideoNode, left: int = 0, right: int = 0, top: int = 0, bottom: int = 0
) -> ConstantFormatVideoNode:
    assert check_variable_format(clip, region_rel_mask)

    if complexpr_available:
        return norm_expr(clip, _get_region_expr(clip, left, right, top, bottom, 0), func=region_rel_mask)

    cropped = vs.core.std.Crop(clip, left, right, top, bottom)

    return vs.core.std.AddBorders(cropped, left, right, top, bottom)

rekt_partial

rekt_partial(
    clip: VideoNode,
    func: Callable[Concatenate[VideoNode, P0], VideoNode],
    left: int = 0,
    right: int = 0,
    top: int = 0,
    bottom: int = 0,
    *args: args,
    **kwargs: kwargs
) -> ConstantFormatVideoNode

Creates a rectangular mask to apply fixes only within the masked area, significantly speeding up filters like anti-aliasing and scaling.

Parameters:

  • clip

    (VideoNode) –

    The source video clip to which the mask will be applied.

  • func

    (Callable[Concatenate[VideoNode, P0], VideoNode]) –

    The function to be applied within the masked area.

  • left

    (int, default: 0 ) –

    The left boundary of the mask, defaults to 0.

  • right

    (int, default: 0 ) –

    The right boundary of the mask, defaults to 0.

  • top

    (int, default: 0 ) –

    The top boundary of the mask, defaults to 0.

  • bottom

    (int, default: 0 ) –

    The bottom boundary of the mask, defaults to 0.

Returns:

  • ConstantFormatVideoNode

    A new clip with the applied mask.

Source code
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
@RektPartial
def rekt_partial(
    clip: vs.VideoNode,
    func: Callable[Concatenate[vs.VideoNode, P0], vs.VideoNode],
    left: int = 0,
    right: int = 0,
    top: int = 0,
    bottom: int = 0,
    *args: P0.args,
    **kwargs: P0.kwargs,
) -> ConstantFormatVideoNode:
    """
    Creates a rectangular mask to apply fixes only within the masked area,
    significantly speeding up filters like anti-aliasing and scaling.

    :param clip:            The source video clip to which the mask will be applied.
    :param func:            The function to be applied within the masked area.
    :param left:            The left boundary of the mask, defaults to 0.
    :param right:           The right boundary of the mask, defaults to 0.
    :param top:             The top boundary of the mask, defaults to 0.
    :param bottom:          The bottom boundary of the mask, defaults to 0.
    :return:                A new clip with the applied mask.
    """

    assert check_variable(clip, rekt_partial._func)

    def _filtered_func(clip: vs.VideoNode, *args: P0.args, **kwargs: P0.kwargs) -> ConstantFormatVideoNode:
        return func(clip, *args, **kwargs)  # type: ignore

    if left == top == right == bottom == 0:
        return _filtered_func(clip, *args, **kwargs)

    cropped = clip.std.Crop(left, right, top, bottom)

    filtered = _filtered_func(cropped, *args, **kwargs)

    check_ref_clip(cropped, filtered, rekt_partial._func)

    if complexpr_available:
        filtered = core.std.AddBorders(filtered, left, right, top, bottom)

        ratio_w, ratio_h = 1 << clip.format.subsampling_w, 1 << clip.format.subsampling_h

        vals = list(
            filter(
                None,
                [
                    ("X {left} >= " if left else None),
                    ("X {right} < " if right else None),
                    ("Y {top} >= " if top else None),
                    ("Y {bottom} < " if bottom else None),
                ],
            )
        )

        return norm_expr(
            [clip, filtered],
            [*vals, ["and"] * (len(vals) - 1), "y x ?"],
            left=[left, left / ratio_w],
            right=[clip.width - right, (clip.width - right) / ratio_w],
            top=[top, top / ratio_h],
            bottom=[clip.height - bottom, (clip.height - bottom) / ratio_h],
            func=rekt_partial._func,
        )

    if not (top or bottom) and (right or left):
        return core.std.StackHorizontal(
            list(
                filter(
                    None,
                    [
                        clip.std.CropAbs(left, clip.height) if left else None,
                        filtered,
                        clip.std.CropAbs(right, clip.height, x=clip.width - right) if right else None,
                    ],
                )
            )
        )

    if (top or bottom) and (right or left):
        filtered = core.std.StackHorizontal(
            list(
                filter(
                    None,
                    [
                        clip.std.CropAbs(left, filtered.height, y=top) if left else None,
                        filtered,
                        clip.std.CropAbs(right, filtered.height, x=clip.width - right, y=top) if right else None,
                    ],
                )
            )
        )

    return core.std.StackVertical(
        list(
            filter(
                None,
                [
                    clip.std.CropAbs(clip.width, top) if top else None,
                    filtered,
                    clip.std.CropAbs(clip.width, bottom, y=clip.height - bottom) if bottom else None,
                ],
            )
        )
    )

replace_squaremask

replace_squaremask(
    clipa: VideoNode,
    clipb: VideoNode,
    mask_params: tuple[int, int, int, int],
    ranges: FrameRangeN | FrameRangesN | None = None,
    blur_sigma: int | float | None = None,
    invert: bool = False,
    func: FuncExceptT | None = None,
    show_mask: bool = False,
) -> ConstantFormatVideoNode

Replace an area of the frame with another clip using a simple square mask.

This is a convenience wrapper merging square masking and framerange replacing functionalities into one function, along with additional utilities such as blurring.

Parameters:

  • clipa

    (VideoNode) –

    Base clip to process.

  • clipb

    (VideoNode) –

    Clip to mask on top of clipa.

  • mask_params

    (tuple[int, int, int, int]) –

    Parameters passed to squaremask. Expects a tuple of (width, height, offset_x, offset_y).

  • ranges

    (FrameRangeN | FrameRangesN | None, default: None ) –

    Frameranges to replace with the masked clip. If None, replaces the entire clip. Default: None.

  • blur_sigma

    (int | float | None, default: None ) –

    Post-blurring of the mask to help hide hard edges. If you pass an int, a :py:func:box_blur will be used. Passing a float will use a :py:func:gauss_blur instead. Default: None.

  • invert

    (bool, default: False ) –

    Invert the mask. This means everything but the defined square will be masked. Default: False.

  • func

    (FuncExceptT | None, default: None ) –

    Function returned for custom error handling. This should only be set by VS package developers. Default: :py:func:squaremask.

  • show_mask

    (bool, default: False ) –

    Return the mask instead of the masked clip.

Returns:

  • ConstantFormatVideoNode

    Clip with a squaremask applied, and optionally set to specific frameranges.

Source code
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
def replace_squaremask(
    clipa: vs.VideoNode,
    clipb: vs.VideoNode,
    mask_params: tuple[int, int, int, int],
    ranges: FrameRangeN | FrameRangesN | None = None,
    blur_sigma: int | float | None = None,
    invert: bool = False,
    func: FuncExceptT | None = None,
    show_mask: bool = False,
) -> ConstantFormatVideoNode:
    """
    Replace an area of the frame with another clip using a simple square mask.

    This is a convenience wrapper merging square masking and framerange replacing functionalities
    into one function, along with additional utilities such as blurring.

    :param clipa:           Base clip to process.
    :param clipb:           Clip to mask on top of `clipa`.
    :param mask_params:     Parameters passed to `squaremask`. Expects a tuple of (width, height, offset_x, offset_y).
    :param ranges:          Frameranges to replace with the masked clip. If `None`, replaces the entire clip.
                            Default: None.
    :param blur_sigma:      Post-blurring of the mask to help hide hard edges.
                            If you pass an int, a :py:func:`box_blur` will be used.
                            Passing a float will use a :py:func:`gauss_blur` instead.
                            Default: None.
    :param invert:          Invert the mask. This means everything *but* the defined square will be masked.
                            Default: False.
    :param func:            Function returned for custom error handling.
                            This should only be set by VS package developers.
                            Default: :py:func:`squaremask`.
    :param show_mask:       Return the mask instead of the masked clip.

    :return:                Clip with a squaremask applied, and optionally set to specific frameranges.
    """
    func = func or replace_squaremask

    assert check_variable(clipa, func) and check_variable(clipb, func)

    mask = squaremask(clipb[0], *mask_params, invert, func=func)

    if isinstance(blur_sigma, int):
        mask = box_blur(mask, blur_sigma)
    elif isinstance(blur_sigma, float):
        mask = gauss_blur(mask, blur_sigma)

    mask = core.std.Loop(mask, clipa.num_frames)

    if show_mask:
        return mask

    merge = clipa.std.MaskedMerge(clipb, mask)

    ranges = normalize_ranges(clipa, ranges)

    if len(ranges) == 1 and ranges[0] == (0, clipa.num_frames - 1):
        return merge

    return replace_ranges(clipa, merge, ranges)

squaremask

squaremask(
    clip: VideoNode,
    width: int,
    height: int,
    offset_x: int,
    offset_y: int,
    invert: bool = False,
    force_gray: bool = True,
    func: FuncExceptT | None = None,
) -> ConstantFormatVideoNode

Create a square used for simple masking.

This is a fast and simple mask that's useful for very rough and simple masking.

Parameters:

  • clip

    (VideoNode) –

    The clip to process.

  • width

    (int) –

    The width of the square. This must be less than clip.width - offset_x.

  • height

    (int) –

    The height of the square. This must be less than clip.height - offset_y.

  • offset_x

    (int) –

    The location of the square, offset from the left side of the frame.

  • offset_y

    (int) –

    The location of the square, offset from the top of the frame.

  • invert

    (bool, default: False ) –

    Invert the mask. This means everything but the defined square will be masked. Default: False.

  • force_gray

    (bool, default: True ) –

    Whether to force using GRAY format or clip format.

  • func

    (FuncExceptT | None, default: None ) –

    Function returned for custom error handling. This should only be set by VS package developers. Default: :py:func:squaremask.

Returns:

  • ConstantFormatVideoNode

    A mask in the shape of a square.

Source code
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
def squaremask(
    clip: vs.VideoNode,
    width: int,
    height: int,
    offset_x: int,
    offset_y: int,
    invert: bool = False,
    force_gray: bool = True,
    func: FuncExceptT | None = None,
) -> ConstantFormatVideoNode:
    """
    Create a square used for simple masking.

    This is a fast and simple mask that's useful for very rough and simple masking.

    :param clip:        The clip to process.
    :param width:       The width of the square. This must be less than clip.width - offset_x.
    :param height:      The height of the square. This must be less than clip.height - offset_y.
    :param offset_x:    The location of the square, offset from the left side of the frame.
    :param offset_y:    The location of the square, offset from the top of the frame.
    :param invert:      Invert the mask. This means everything *but* the defined square will be masked.
                        Default: False.
    :param force_gray:  Whether to force using GRAY format or clip format.
    :param func:        Function returned for custom error handling.
                        This should only be set by VS package developers.
                        Default: :py:func:`squaremask`.

    :return:            A mask in the shape of a square.
    """
    func = func or squaremask

    assert check_variable(clip, func)

    mask_format = (
        clip.format.replace(color_family=vs.GRAY, subsampling_w=0, subsampling_h=0) if force_gray else clip.format
    )

    if offset_x + width > clip.width or offset_y + height > clip.height:
        raise CustomValueError("mask exceeds clip size!", func)

    if complexpr_available:
        base_clip = vs.core.std.BlankClip(
            clip, None, None, mask_format.id, 1, color=get_lowest_values(mask_format, ColorRange.FULL), keep=True
        )
        exprs = [
            _get_region_expr(
                base_clip,
                offset_x,
                clip.width - width - offset_x,
                offset_y,
                clip.height - height - offset_y,
                "range_max x" if invert else "x range_max",
            )
        ]

        if mask_format.num_planes > 1:
            for i in range(1, mask_format.num_planes):
                p = plane(base_clip, i)
                ratio_x = p.width / base_clip.width
                ratio_y = p.height / base_clip.height
                exprs.append(
                    _get_region_expr(
                        p,
                        int(offset_x * ratio_x),
                        int((clip.width - width - offset_x) * ratio_x),
                        int(offset_y * ratio_y),
                        int((clip.height - height - offset_y) * ratio_y),
                        "range_max x" if invert else "x range_max",
                    )
                )

        mask = norm_expr(base_clip, tuple(exprs), func=func)
    else:
        base_clip = core.std.BlankClip(
            clip, width, height, mask_format.id, 1, color=get_peak_values(mask_format, ColorRange.FULL), keep=True
        )

        mask = core.std.AddBorders(
            base_clip,
            offset_x,
            clip.width - width - offset_x,
            offset_y,
            clip.height - height - offset_y,
            [0] * mask_format.num_planes,
        )
        if invert:
            mask = core.std.Invert(mask)

    if clip.num_frames == 1:
        return mask

    return core.std.Loop(mask, clip.num_frames)