Audio Level¶
Abstract |
Audio Level |
Authors |
Walter Fan |
Status |
WIP |
Updated |
2021-12-29 |
概述¶
音频电平 Audio Level 就是指声音的高低，它主要是指声波的振幅，我们先了解几个概念
声压¶
有效声压
声压级 SPL (Sound Press Level)¶
声音的有效声压与基准声压之比，取以10 为底的对数，再乘以 20， 即为声压级
声压级
度量单位¶
分贝（dB）是十分之一贝尔（B）：1B = 10dB。1贝尔的两个功率量的比值是10:1，1贝尔的两个场量的比值是{displaystyle {sqrt {10}}:1}{sqrt {10}}:1[1]。
使用分贝有很多便利之处：
分贝实际上是对数值，因此可以用常用的数量来表示非常大的比值，可以清楚地表示非常大的数量变化。
多部件系统的整体增益（如级联的放大器）可以直接用各部件的增益分贝相加而求得。不必把这些增益值相乘（例如log(A × B × C) = log(A) + log(B) + log(C)）。
人对强度的感知，如声音或者光照，更接近与强度的对数成正比而不是强度值本身，依据韦伯定理，因此分贝值可用于描述感知级别或级差。
dB与后缀的组合，指出计算比值时的参考值。例如dBm指示功率值与1毫瓦的比值的分贝数。
如果计算分贝时的参考值明确、确切地给出，那么分贝数值可以作为绝对量，如同被测量的功率量或者场量。例如，20dBm即为100毫瓦。
SI国际单位制不允许使用分贝与后缀的组合形式如dBm, dBu, dBA，等等[13]。但这种不遵从SI单位制的表示却广泛应用于很多场合。
dBV¶
dB(VRMS) – voltage relative to 1 volt, regardless of impedance.[3] This is used to measure microphone sensitivity, and also to specify the consumer line-level of −10 dBV, in order to reduce manufacturing costs relative to equipment using a +4 dBu line-level signal.[45]
dBm¶
dB(mW) – power relative to 1 milliwatt. In audio and telephony, dBm is typically referenced relative to a 600 Ω impedance,[56] which corresponds to a voltage level of 0.775 volts or 775 millivolts.
dB SPL¶
dB（声压级，sound pressure level）–在空气或其它气体中的声压，参考值为20微帕斯卡（μPa） = 2×10−5 Pa，这是人能听到的最安静的声音。大致相当于3米外蚊子飞行的声音。经常被缩写为”dB”，这造成了很多误解以为”dB”是个有量纲的绝对单位。对于水声或其它液体，参考值是1 μPa[18]。
dB SPL (sound pressure level) – for sound in air and other gases, relative to 20 micropascals (μPa), or 2×10−5 Pa, approximately the quietest sound a human can hear. For sound in water and other liquids, a reference pressure of 1 μPa is used.[51] An RMS sound pressure of one pascal corresponds to a level of 94 dB SPL.
Audio Level¶
RFC6464 中规定声音电平 audio level 以一字节扩展头的最低 7 个比特来表示，此字节的最高一位是一个标志位 “V”。
最低7位比特：表示 audio level 单位是 -dBov，其值从 0（最大声） 到 127（静音） ，代表着 dBov 的 0（最大声） 到 -127（静音）
最高1位比特：表示这个包否包含语音(voice activity), 如果在 SDP 协商中没有启用 “vad”(Voice Activity Detection) 语音活动检测，就忽略这个比特
Metrics¶
audioLevel¶
type: double
Only exists for audio. Represents the audio level of the receiving track. For audio levels of tracks attached locally, see RTCAudioSourceStats instead.
The value is between 0..1 (linear), where 1.0 represents 0 dBov, 0 represents silence, and 0.5 represents approximately 6 dBSPL change in the sound pressure level from 0 dBov.
The audioLevel is averaged over some small interval, using the algorithm described under totalAudioEnergy. The interval used is implementation dependent.
totalAudioEnergy¶
type: double
Only exists for audio. Represents the audio energy of the receiving track. For audio energy of tracks attached locally, see RTCAudioSourceStats instead.
This value must be computed as follows: for each audio sample that is received (and thus counted by totalSamplesReceived), add the sample’s value divided by the highest-intensity encodable value, squared and then multiplied by the duration of the sample in seconds. In other words, duration * Math.pow(energy/maxEnergy, 2).
This can be used to obtain a root mean square (RMS) value that uses the same units as audioLevel, as defined in [RFC6464]. It can be converted to these units using the formula Math.sqrt(totalAudioEnergy/totalSamplesDuration). This calculation can also be performed using the differences between the values of two different getStats() calls, in order to compute the average audio level over any desired time interval. In other words, do Math.sqrt((energy2 - energy1)/(duration2 - duration1)).
For example, if a 10ms packet of audio is produced with an RMS of 0.5 (out of 1.0), this should add 0.5 * 0.5 * 0.01 = 0.0025 to totalAudioEnergy. If another 10ms packet with an RMS of 0.1 is received, this should similarly add 0.0001 to totalAudioEnergy. Then, Math.sqrt(totalAudioEnergy/totalSamplesDuration) becomes Math.sqrt(0.0026/0.02) = 0.36, which is the same value that would be obtained by doing an RMS calculation over the contiguous 20ms segment of audio.
If multiple audio channels are used, the audio energy of a sample refers to the highest energy of any channel.
这个度量指标只存在于音频中。它表示接收轨道的音频能量。有关本地附加轨道的音频能量，请参阅 RTCAudioSourceStats。
该值必须按如下方式计算：
对于接收到的每个音频样本（因此按 totalSamplesReceived 计数），将样本值除以最高强度的可编码值，平方，然后乘以以秒为单位的样本持续时间。
duration * Math.pow(energy/maxEnergy, 2).
这可用于获得与 audioLevel 使用相同单位的均方根 (RMS) 值，如 [RFC6464] 中所定义。可以使用公式 Math.sqrt(totalAudioEnergy/totalSamplesDuration) 将其转换为这些单位。也可以使用两个不同 getStats() 调用的值之间的差异来执行此计算，以便计算任何所需时间间隔内的平均音频电平。换句话说，做 Math.sqrt((energy2 - energy1)/(duration2 - duration1))。
例如，如果生成一个 10 毫秒的音频数据包，其 RMS 为 0.5（超出 1.0），则应将 0.5 * 0.5 * 0.01 = 0.0025 添加到 totalAudioEnergy。如果接收到另一个 RMS 为 0.1 的 10ms 数据包，这应该类似地将 0.0001 添加到 totalAudioEnergy。然后，Math.sqrt(totalAudioEnergy/totalSamplesDuration) 变为 Math.sqrt(0.0026/0.02) = 0.36，这与通过对连续 20 毫秒音频段进行 RMS 计算获得的值相同。
如果使用多个音频通道，则样本的音频能量是指任一通道的最高能量。
示例代码¶
RFC6464 中有计算 Audio Level 的 java 代码
显示/隐藏代码
1 /*
2 Copyright (c) 2011 IETF Trust and the persons identified
3 as authors of the code. All rights reserved.
4
5 Redistribution and use in source and binary forms, with
6 or without modification, is permitted pursuant to, and subject
7 to the license terms contained in, the Simplified BSD License
8 set forth in Section 4.c of the IETF Trust's Legal Provisions
9 Relating to IETF Documents (http://trustee.ietf.org/license-info).
10 */
11
12 /**
13 * Calculates the audio level of specific samples of a signal
14 * relative to overload.
15 */
16 public class AudioLevelCalculator
17 {
18
19 /**
20 * Calculates the audio level of a signal with specific
21 * <tt>samples</tt>.
22 *
23 * @param samples the samples whose audio level we need to
24 * calculate. The samples are specified as an <tt>int</tt>
25 * array starting at <tt>offset</tt>, extending <tt>length</tt>
26 * number of elements, and each <tt>int</tt> element in the
27 * specified range representing a sample whose audio level we
28
29 * need to calculate. Though a sample is provided in the
30 * form of an <tt>int</tt> value, the sample size in bits
31 * is determined by the caller via <tt>overload</tt>.
32 *
33 * @param offset the offset in <tt>samples</tt> at which the
34 * samples start.
35 *
36 * @param length the length of the signal specified in
37 * <tt>samples<tt>, starting at <tt>offset</tt>.
38 *
39 * @param overload the overload (point) of <tt>signal</tt>.
40 * For example, <tt>overload</tt> can be {@link Byte#MAX_VALUE}
41 * for 8-bit signed samples or {@link Short#MAX_VALUE} for
42 * 16-bit signed samples.
43 *
44 * @return the audio level of the specified signal.
45 */
46 public static int calculateAudioLevel(
47 int[] samples, int offset, int length,
48 int overload)
49 {
50 /*
51 * Calculate the root mean square (RMS) of the signal.
52 */
53 double rms = 0;
54
55 for (; offset < length; offset++)
56 {
57 double sample = samples[offset];
58
59 sample /= overload;
60 rms += sample * sample;
61 }
62 rms = (length == 0) ? 0 : Math.sqrt(rms / length);
63
64 /*
65 * The audio level is a logarithmic measure of the
66 * rms level of an audio sample relative to a reference
67 * value and is measured in decibels.
68 */
69 double db;
70
71 /*
72 * The minimum audio level permitted.
73 */
74 final double MIN_AUDIO_LEVEL = -127;
75
76 /*
77 * The maximum audio level permitted.
78 */
79 final double MAX_AUDIO_LEVEL = 0;
80
81 if (rms > 0)
82 {
83 /*
84 * The "zero" reference level is the overload level,
85 * which corresponds to 1.0 in this calculation, because
86 * the samples are normalized in calculating the RMS.
87 */
88 db = 20 * Math.log10(rms);
89
90 /*
91 * Ensure that the calculated level is within the minimum
92 * and maximum range permitted.
93 */
94 if (db < MIN_AUDIO_LEVEL)
95 db = MIN_AUDIO_LEVEL;
96 else if (db > MAX_AUDIO_LEVEL)
97 db = MAX_AUDIO_LEVEL;
98 }
99 else
100 {
101 db = MIN_AUDIO_LEVEL;
102 }
103
104 return (int)Math.round(db);
105 }
106 }
FAQ¶
Gain 和 Volume 的区别¶
Gain是指针对输入的声音信号的增益调节放大倍数，Volume是指针对输出的音量大小。
Gain指增益：增益就是用来增强输入进音箱电路的信号的。增益的作用就是放大信号。将输入的信号放大一定的倍数，而单位通常是我们熟悉的分贝
Volume指音量：就是声音的响度，指音量的大小，也就是最终输出的声音大小。