日语五元音动画曲线文件输出(.adxlip)

指定参数 -enable_adxlip_compat 时会输出此文件。

以与 CriLipsMake.exe 输出文件兼容的格式输出。
仅记录日语五元音的有效值。

文件内容

分析结果以逗号分隔的文本文件输出。文件格式如下：

// input: created by CriLipsMake2
// framerate: 100 [fps]
// frame count,msec,width(0-1 def=0.583),height(0-1 def=0.000),toungue(0-1 def=0.000),A(0-1),I(0-1),U(0-1),E(0-1),O(0-1),Vol(dB)
0,0.0000,0.5830,0.0000,0.0000,0.0000,0.0000,0.0000,0.0000,0.0000,-96.0000
1,10.0000,0.5830,0.0000,0.0000,0.0000,0.0000,0.0000,0.0000,0.0000,-96.0000
2,20.0000,0.5830,0.0000,0.0000,0.0000,0.0615,0.0000,0.0000,0.0000,-96.0000
3,30.0000,0.5830,0.0000,0.0000,0.0000,0.1988,0.0000,0.0000,0.0000,-96.0000
4,40.0000,0.5830,0.0000,0.0000,0.0000,0.3530,0.0000,0.0000,0.0000,-96.0000
5,50.0000,0.5830,0.0000,0.0000,0.0000,0.5046,0.0000,0.0000,0.0000,-96.0000
6,60.0000,0.5830,0.0000,0.0000,0.0000,0.6423,0.0000,0.0000,0.0000,-96.0000
7,70.0000,0.5830,0.0000,0.0000,0.0000,0.7618,0.0000,0.0000,0.0000,-96.0000
8,80.0000,0.5830,0.0000,0.0000,0.0000,0.8605,0.0000,0.0000,0.0000,-96.0000
9,90.0000,0.5830,0.0000,0.0000,0.0000,0.9326,0.0000,0.0000,0.0000,-96.0000
10,100.0000,0.5830,0.0000,0.0000,0.0000,0.9700,0.0000,0.0000,0.0000,-96.0000

以'//'开头的行为注释部分，为头部。
当前版本头部输出3行。
头部以下的每一行都记录了口型数据的分析结果。

frame count

帧计数是根据帧率（固定为100FPS）依次递增的分析帧编号。

msec

该行口型数据对应的音频时间（毫秒）。
由 CriLipsMake2 输出时，固定为10毫秒。

width, height

CriLipsMake2 不支持此项。始终为闭口（0.0, 0.583）。

tongue position

根据分析结果写入相应的值。

A, I, U, E, O

分别表示以下值：

变形目标 “あ” 的混合量（0.0f ~ 1.0f）
变形目标 “い” 的混合量（0.0f ~ 1.0f）
变形目标 “う” 的混合量（0.0f ~ 1.0f）
变形目标 “え” 的混合量（0.0f ~ 1.0f）
变形目标 “お” 的混合量（0.0f ~ 1.0f）

如果模型数据具有可混合的变形目标，请使用此输出值作为变形量。

此外，输出的五个混合量中，值大于0的最多同时只有2个。
这是为了防止组合3个以上混合形状导致口型异常。

所有混合量的总和不会超过1.0。
如果 “あいうえお” 全为0.0，则表示“说话者闭口状态”。

Vol

分贝为0.0时音量最大，越负音量越小，本结果中-96表示无声。
请注意-96～0.0并不是线性增加。
此值作为应用程序的参考值输出。
例如需要根据音量调整嘴巴张开大小时可参考此值。

文件内容​

frame count​

msec​

width, height​

tongue position​

A, I, U, E, O​

Vol​