Skip to content

rff

Functions:

apply_rff_array 

apply_rff_array(
    old_array: Sequence[T],
    rff: Sequence[int],
    tff: Sequence[int],
    prog_seq: Sequence[int],
) -> list[T]
Source code in vssource/rff.py 
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
def apply_rff_array(old_array: Sequence[T], rff: Sequence[int], tff: Sequence[int], prog_seq: Sequence[int]) -> list[T]:
    array_double_rate = list[T]()

    for prog, arr, rffv, tffv in zip(prog_seq, old_array, rff, tff):
        repeat_amount = (3 if rffv else 2) if prog == 0 else ((6 if tffv else 4) if rffv else 2)

        array_double_rate.extend([arr] * repeat_amount)

    # assert (len(array_double_rate) % 2) == 0
    if (len(array_double_rate) % 2) != 0:
        warnings.warn("uneven amount of fields removing last\n")
        array_double_rate = array_double_rate[:-1]

    # It seems really weird thats its allowed to have rff stuff across
    # vob boundries even for multi angle stuff i have seen this so often though it is ok to remove the warnings
    for i, f1, f2 in zip(count(), array_double_rate[::2], array_double_rate[1::2]):
        if f1 != f2:
            warnings.warn(f"Ambiguous pattern due to rff {f1}!={f2} on index {i}\n")

    # 1::2 because there might be the case where there is a one frame cell at the end and if we cut off the last field
    # we will miss that vobid with 0::2
    return array_double_rate[1::2]

apply_rff_video 

apply_rff_video(
    node: VideoNode,
    rff: list[int],
    tff: list[int],
    prog: list[int],
    prog_seq: list[int],
) -> ConstantFormatVideoNode
Source code in vssource/rff.py 
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
def apply_rff_video(
    node: vs.VideoNode, rff: list[int], tff: list[int], prog: list[int], prog_seq: list[int]
) -> ConstantFormatVideoNode:
    assert len(node) == len(rff) == len(tff) == len(prog) == len(prog_seq)

    fields = list[dict[str, int]]()
    tfffs = node.std.RemoveFrameProps(["_FieldBased", "_Field"]).std.SeparateFields(True)

    for i, current_prg_seq, current_prg, current_rff, current_tff in zip(count(), prog_seq, prog, rff, tff):
        if not current_prg_seq:
            if current_tff:
                first_field = 2 * i
                second_field = 2 * i + 1
            else:
                first_field = 2 * i + 1
                second_field = 2 * i

            fields += [
                {"n": first_field, "tf": current_tff, "prg": False, "repeat": False},
                {"n": second_field, "tf": not current_tff, "prg": False, "repeat": False},
            ]

            if current_rff:
                assert current_prg
                repeat_field = deepcopy(fields[-2])
                repeat_field["repeat"] = True
                fields.append(repeat_field)
        else:
            assert current_prg

            field_count = 1
            if current_rff:
                field_count += 1 + int(current_tff)

            fields += [
                {"n": 2 * i, "tf": 1, "prg": True, "repeat": False},
                {"n": 2 * i + 1, "tf": 0, "prg": True, "repeat": False},
            ] * field_count

    # TODO: mark known progressive frames as progressive

    # assert (len(fields) % 2) == 0
    if (len(fields) % 2) != 0:
        warnings.warn("uneven amount of fields removing last\n")
        fields = fields[:-1]

    for a, tf, bf in zip(count(), fields[::2], fields[1::2]):
        if tf["tf"] == bf["tf"]:
            bf["tf"] = not bf["tf"]

            warnings.warn(f"Invalid field transition at {a}")

    for fcurr, fnext in zip(fields[::2], fields[1::2]):
        if fcurr["tf"] == fnext["tf"]:
            raise CustomRuntimeError(
                f"Found invalid stream with two consecutive {'top' if fcurr['tf'] else 'bottom'} fields!"
            )

    final = remap_frames(tfffs, [x["n"] for x in fields])

    def _set_field(n: int, f: vs.VideoFrame) -> vs.VideoFrame:
        f = f.copy()

        f.props.pop("_FieldBased", None)
        f.props._Field = fields[n]["tf"]

        return f

    final = vs.core.std.ModifyFrame(final, final, _set_field)

    woven = final.std.DoubleWeave()[::2]

    def _set_repeat(n: int, f: vs.VideoFrame) -> vs.VideoFrame:
        f = f.copy()
        if fields[n * 2]["repeat"]:
            f.props["RepeatedField"] = 1
        elif fields[n * 2 + 1]["repeat"]:
            f.props["RepeatedField"] = 0
        else:
            f.props["RepeatedField"] = -1
        return f

    woven = vs.core.std.ModifyFrame(woven, woven, _set_repeat)

    # TODO: this seems to not work or atleast useless since its disable for non progressive sequence which is rare
    def _update_progressive(n: int, f: vs.VideoFrame) -> vs.VideoFrame:
        fout = f.copy()

        tf = fields[n * 2]
        bf = fields[n * 2 + 1]

        if tf["prg"] and bf["prg"]:
            fout.props["_FieldBased"] = 0

        return fout

    return vs.core.std.ModifyFrame(woven, woven, _update_progressive)

cut_array_on_ranges 

cut_array_on_ranges(array: list[T], ranges: list[tuple[int, int]]) -> list[T]
Source code in vssource/rff.py 
136
137
def cut_array_on_ranges(array: list[T], ranges: list[tuple[int, int]]) -> list[T]:
    return [array[i] for i in flatten([range(rrange[0], rrange[1] + 1) for rrange in ranges])]