主题：1/3oct smoothing 是什么意思，怎样实现

平滑，类似于归一化处理，方法有多种多样。
最简单的，就是点与点之间直线相连，这样一来，曲线上就有明显的尖角。通常就不会用此方法了。

归一化方法主要有如下几种，供大家参考：
1、线性函数转换，表达式如下：
y=(x-MinValue)/(MaxValue-MinValue)
说明：x、y分别为转换前、后的值，MaxValue、MinValue分别为样本的最大值和最小值。
2、对数函数转换，表达式如下：
y=log10(x)
说明：以10为底的对数函数转换。
3、反余切函数转换，表达式如下：
y=atan(x)*2/PI


还有其它的方法。
各种测试软件，算法有所差别。

http://www.loudspeakers.net/files/literature/tech%20notes/SMOOTH_TECH.pdf 

The audio industry lacks a standard for how much or how little data should be smoothed when
presenting a particular form of information. For this reason, each manufacturer has taken it
upon themselves to create their own company, or even product, standard. It is impossible to
look at any two manufacturers’ frequency response curves, for example, and compare them
since each uses a different test method and smoothing standard. These smoothing
“standards” often are not documented and vary from product to product within a
manufacturer’s product line. Sometimes smoothing is done so that curve relates better to what
is actually heard. Sometimes it is done to prove a point, and many times it is done for
competitive reasons; “Ours sounds better than theirs, so our curve needs to look at least as
good as theirs or no one will buy ours no matter how good it sounds”. This is the point where
science becomes fiction and real data disappears.

谢谢晴朗.

不理解，三分之一倍频比六分之一倍频 更平滑 ， 本人认为六分之一倍频取样更密集，应该比三分之一倍频平滑

如果均以20Hz为起点，1/3oct的点数就是20*1.256（2的1/3平），再以前一个数据当作基数乘以1.256，同理可得1/6oct（1.1225）和1/12oct（1.0595）。

可见在相同频宽范围内，1/3oct点数是最少的，1/12oct点数是最多的。点数取得越多，频响曲线显示越精确，峰峰谷谷也就越多，曲线也就显得不是很平滑。反之，点数越少的也就显示得越平滑。

 

不知是否正确，如有错误，请不啬抛砖。