|
Chris@1
|
1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
|
|
Chris@1
|
2 <html>
|
|
Chris@1
|
3 <head>
|
|
Chris@1
|
4
|
|
Chris@1
|
5 <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15"/>
|
|
Chris@1
|
6 <title>Ogg Vorbis Documentation</title>
|
|
Chris@1
|
7
|
|
Chris@1
|
8 <style type="text/css">
|
|
Chris@1
|
9 body {
|
|
Chris@1
|
10 margin: 0 18px 0 18px;
|
|
Chris@1
|
11 padding-bottom: 30px;
|
|
Chris@1
|
12 font-family: Verdana, Arial, Helvetica, sans-serif;
|
|
Chris@1
|
13 color: #333333;
|
|
Chris@1
|
14 font-size: .8em;
|
|
Chris@1
|
15 }
|
|
Chris@1
|
16
|
|
Chris@1
|
17 a {
|
|
Chris@1
|
18 color: #3366cc;
|
|
Chris@1
|
19 }
|
|
Chris@1
|
20
|
|
Chris@1
|
21 img {
|
|
Chris@1
|
22 border: 0;
|
|
Chris@1
|
23 }
|
|
Chris@1
|
24
|
|
Chris@1
|
25 #xiphlogo {
|
|
Chris@1
|
26 margin: 30px 0 16px 0;
|
|
Chris@1
|
27 }
|
|
Chris@1
|
28
|
|
Chris@1
|
29 #content p {
|
|
Chris@1
|
30 line-height: 1.4;
|
|
Chris@1
|
31 }
|
|
Chris@1
|
32
|
|
Chris@1
|
33 h1, h1 a, h2, h2 a, h3, h3 a {
|
|
Chris@1
|
34 font-weight: bold;
|
|
Chris@1
|
35 color: #ff9900;
|
|
Chris@1
|
36 margin: 1.3em 0 8px 0;
|
|
Chris@1
|
37 }
|
|
Chris@1
|
38
|
|
Chris@1
|
39 h1 {
|
|
Chris@1
|
40 font-size: 1.3em;
|
|
Chris@1
|
41 }
|
|
Chris@1
|
42
|
|
Chris@1
|
43 h2 {
|
|
Chris@1
|
44 font-size: 1.2em;
|
|
Chris@1
|
45 }
|
|
Chris@1
|
46
|
|
Chris@1
|
47 h3 {
|
|
Chris@1
|
48 font-size: 1.1em;
|
|
Chris@1
|
49 }
|
|
Chris@1
|
50
|
|
Chris@1
|
51 li {
|
|
Chris@1
|
52 line-height: 1.4;
|
|
Chris@1
|
53 }
|
|
Chris@1
|
54
|
|
Chris@1
|
55 #copyright {
|
|
Chris@1
|
56 margin-top: 30px;
|
|
Chris@1
|
57 line-height: 1.5em;
|
|
Chris@1
|
58 text-align: center;
|
|
Chris@1
|
59 font-size: .8em;
|
|
Chris@1
|
60 color: #888888;
|
|
Chris@1
|
61 clear: both;
|
|
Chris@1
|
62 }
|
|
Chris@1
|
63 </style>
|
|
Chris@1
|
64
|
|
Chris@1
|
65 </head>
|
|
Chris@1
|
66
|
|
Chris@1
|
67 <body>
|
|
Chris@1
|
68
|
|
Chris@1
|
69 <div id="xiphlogo">
|
|
Chris@1
|
70 <a href="http://www.xiph.org/"><img src="fish_xiph_org.png" alt="Fish Logo and Xiph.Org"/></a>
|
|
Chris@1
|
71 </div>
|
|
Chris@1
|
72
|
|
Chris@1
|
73 <h1>Ogg Vorbis: Fidelity measurement and terminology discussion</h1>
|
|
Chris@1
|
74
|
|
Chris@1
|
75 <p>Terminology discussed in this document is based on common terminology
|
|
Chris@1
|
76 associated with contemporary codecs such as MPEG I audio layer 3
|
|
Chris@1
|
77 (mp3). However, some differences in terminology are useful in the
|
|
Chris@1
|
78 context of Vorbis as Vorbis functions somewhat differently than most
|
|
Chris@1
|
79 current formats. For clarity, then, we describe a common terminology
|
|
Chris@1
|
80 for discussion of Vorbis's and other formats' audio quality.</p>
|
|
Chris@1
|
81
|
|
Chris@1
|
82 <h2>Subjective and Objective</h2>
|
|
Chris@1
|
83
|
|
Chris@1
|
84 <p><em>Objective</em> fidelity is a measure, based on a computable,
|
|
Chris@1
|
85 mechanical metric, of how carefully an output matches an input. For
|
|
Chris@1
|
86 example, a stereo amplifier may claim to introduce less that .01%
|
|
Chris@1
|
87 total harmonic distortion when amplifying an input signal; this claim
|
|
Chris@1
|
88 is easy to verify given proper equipment, and any number of testers are
|
|
Chris@1
|
89 likely to arrive at the same, exact results. One need not listen to
|
|
Chris@1
|
90 the equipment to make this measurement.</p>
|
|
Chris@1
|
91
|
|
Chris@1
|
92 <p>However, given two amplifiers with identical, verifiable objective
|
|
Chris@1
|
93 specifications, listeners may strongly prefer the sound quality of one
|
|
Chris@1
|
94 over the other. This is actually the case in the decades old debate
|
|
Chris@1
|
95 [some would say jihad] among audiophiles involving vacuum tube versus
|
|
Chris@1
|
96 solid state amplifiers. There are people who can tell the difference,
|
|
Chris@1
|
97 and strongly prefer one over the other despite seemingly identical,
|
|
Chris@1
|
98 measurable quality. This preference is <em>subjective</em> and
|
|
Chris@1
|
99 difficult to measure but nonetheless real.</p>
|
|
Chris@1
|
100
|
|
Chris@1
|
101 <p>Individual elements of subjective differences often can be qualified,
|
|
Chris@1
|
102 but overall subjective quality generally is not measurable. Different
|
|
Chris@1
|
103 observers are likely to disagree on the exact results of a subjective
|
|
Chris@1
|
104 test as each observer's perspective differs. When measuring
|
|
Chris@1
|
105 subjective qualities, the best one can hope for is average, empirical
|
|
Chris@1
|
106 results that show statistical significance across a group.</p>
|
|
Chris@1
|
107
|
|
Chris@1
|
108 <p>Perceptual codecs are most concerned with subjective, not objective,
|
|
Chris@1
|
109 quality. This is why evaluating a perceptual codec via distortion
|
|
Chris@1
|
110 measures and sonograms alone is useless; these objective measures may
|
|
Chris@1
|
111 provide insight into the quality or functioning of a codec, but cannot
|
|
Chris@1
|
112 answer the much squishier subjective question, "Does it sound
|
|
Chris@1
|
113 good?". The tube amplifier example is perhaps not the best as very few
|
|
Chris@1
|
114 people can hear, or care to hear, the minute differences between tubes
|
|
Chris@1
|
115 and transistors, whereas the subjective differences in perceptual
|
|
Chris@1
|
116 codecs tend to be quite large even when objective differences are
|
|
Chris@1
|
117 not.</p>
|
|
Chris@1
|
118
|
|
Chris@1
|
119 <h2>Fidelity, Artifacts and Differences</h2>
|
|
Chris@1
|
120
|
|
Chris@1
|
121 <p>Audio <em>artifacts</em> and loss of fidelity or more simply
|
|
Chris@1
|
122 put, audio <em>differences</em> are not the same thing.</p>
|
|
Chris@1
|
123
|
|
Chris@1
|
124 <p>A loss of fidelity implies differences between the perceived input and
|
|
Chris@1
|
125 output signal; it does not necessarily imply that the differences in
|
|
Chris@1
|
126 output are displeasing or that the output sounds poor (although this
|
|
Chris@1
|
127 is often the case). Tube amplifiers are <em>not</em> higher fidelity
|
|
Chris@1
|
128 than modern solid state and digital systems. They simply produce a
|
|
Chris@1
|
129 form of distortion and coloring that is either unnoticeable or actually
|
|
Chris@1
|
130 pleasing to many ears.</p>
|
|
Chris@1
|
131
|
|
Chris@1
|
132 <p>As compared to an original signal using hard metrics, all perceptual
|
|
Chris@1
|
133 codecs [ASPEC, ATRAC, MP3, WMA, AAC, TwinVQ, AC3 and Vorbis included]
|
|
Chris@1
|
134 lose objective fidelity in order to reduce bitrate. This is fact. The
|
|
Chris@1
|
135 idea is to lose fidelity in ways that cannot be perceived. However,
|
|
Chris@1
|
136 most current streaming applications demand bitrates lower than what
|
|
Chris@1
|
137 can be achieved by sacrificing only objective fidelity; this is also
|
|
Chris@1
|
138 fact, despite whatever various company press releases might claim.
|
|
Chris@1
|
139 Subjective fidelity eventually must suffer in one way or another.</p>
|
|
Chris@1
|
140
|
|
Chris@1
|
141 <p>The goal is to choose the best possible tradeoff such that the
|
|
Chris@1
|
142 fidelity loss is graceful and not obviously noticeable. Most listeners
|
|
Chris@1
|
143 of FM radio do not realize how much lower fidelity that medium is as
|
|
Chris@1
|
144 compared to compact discs or DAT. However, when compared directly to
|
|
Chris@1
|
145 source material, the difference is obvious. A cassette tape is lower
|
|
Chris@1
|
146 fidelity still, and yet the degradation, relatively speaking, is
|
|
Chris@1
|
147 graceful and generally easy not to notice. Compare this graceful loss
|
|
Chris@1
|
148 of quality to an average 44.1kHz stereo mp3 encoded at 80 or 96kbps.
|
|
Chris@1
|
149 The mp3 might actually be higher objective fidelity but subjectively
|
|
Chris@1
|
150 sounds much worse.</p>
|
|
Chris@1
|
151
|
|
Chris@1
|
152 <p>Thus, when a CODEC <em>must</em> sacrifice subjective quality in order
|
|
Chris@1
|
153 to satisfy a user's requirements, the result should be a
|
|
Chris@1
|
154 <em>difference</em> that is generally either difficult to notice
|
|
Chris@1
|
155 without comparison, or easy to ignore. An <em>artifact</em>, on the
|
|
Chris@1
|
156 other hand, is an element introduced into the output that is
|
|
Chris@1
|
157 immediately noticeable, obviously foreign, and undesired. The famous
|
|
Chris@1
|
158 'underwater' or 'twinkling' effect synonymous with low bitrate (or
|
|
Chris@1
|
159 poorly encoded) mp3 is an example of an <em>artifact</em>. This
|
|
Chris@1
|
160 working definition differs slightly from common usage, but the coined
|
|
Chris@1
|
161 distinction between differences and artifacts is useful for our
|
|
Chris@1
|
162 discussion.</p>
|
|
Chris@1
|
163
|
|
Chris@1
|
164 <p>The goal, when it is absolutely necessary to sacrifice subjective
|
|
Chris@1
|
165 fidelity, is obviously to strive for differences and not artifacts.
|
|
Chris@1
|
166 The vast majority of codecs today fail at this task miserably,
|
|
Chris@1
|
167 predictably, and regularly in one way or another. Avoiding such
|
|
Chris@1
|
168 failures when it is necessary to sacrifice subjective quality is a
|
|
Chris@1
|
169 fundamental design objective of Vorbis and that objective is reflected
|
|
Chris@1
|
170 in Vorbis's design and tuning.</p>
|
|
Chris@1
|
171
|
|
Chris@1
|
172 <div id="copyright">
|
|
Chris@1
|
173 The Xiph Fish Logo is a
|
|
Chris@1
|
174 trademark (™) of Xiph.Org.<br/>
|
|
Chris@1
|
175
|
|
Chris@1
|
176 These pages © 1994 - 2005 Xiph.Org. All rights reserved.
|
|
Chris@1
|
177 </div>
|
|
Chris@1
|
178
|
|
Chris@1
|
179 </body>
|
|
Chris@1
|
180 </html>
|
