VirtualBox

source: vbox/trunk/src/libs/libvorbis-1.3.7/doc/vorbis-fidelity.html@ 103131

Last change on this file since 103131 was 96468, checked in by vboxsync, 2 years ago

libs/libvorbis-1.3.7: Re-exporting, hopefully this time everything is there. bugref:10275

  • Property svn:eol-style set to native
  • Property svn:keywords set to Author Date Id Revision
File size: 6.5 KB
Line 
1<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
2<html>
3<head>
4
5<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15"/>
6<title>Ogg Vorbis Documentation</title>
7
8<style type="text/css">
9body {
10 margin: 0 18px 0 18px;
11 padding-bottom: 30px;
12 font-family: Verdana, Arial, Helvetica, sans-serif;
13 color: #333333;
14 font-size: .8em;
15}
16
17a {
18 color: #3366cc;
19}
20
21img {
22 border: 0;
23}
24
25#xiphlogo {
26 margin: 30px 0 16px 0;
27}
28
29#content p {
30 line-height: 1.4;
31}
32
33h1, h1 a, h2, h2 a, h3, h3 a {
34 font-weight: bold;
35 color: #ff9900;
36 margin: 1.3em 0 8px 0;
37}
38
39h1 {
40 font-size: 1.3em;
41}
42
43h2 {
44 font-size: 1.2em;
45}
46
47h3 {
48 font-size: 1.1em;
49}
50
51li {
52 line-height: 1.4;
53}
54
55#copyright {
56 margin-top: 30px;
57 line-height: 1.5em;
58 text-align: center;
59 font-size: .8em;
60 color: #888888;
61 clear: both;
62}
63</style>
64
65</head>
66
67<body>
68
69<div id="xiphlogo">
70 <a href="https://xiph.org/"><img src="fish_xiph_org.png" alt="Fish Logo and Xiph.Org"/></a>
71</div>
72
73<h1>Ogg Vorbis: Fidelity measurement and terminology discussion</h1>
74
75<p>Terminology discussed in this document is based on common terminology
76associated with contemporary codecs such as MPEG I audio layer 3
77(mp3). However, some differences in terminology are useful in the
78context of Vorbis as Vorbis functions somewhat differently than most
79current formats. For clarity, then, we describe a common terminology
80for discussion of Vorbis's and other formats' audio quality.</p>
81
82<h2>Subjective and Objective</h2>
83
84<p><em>Objective</em> fidelity is a measure, based on a computable,
85mechanical metric, of how carefully an output matches an input. For
86example, a stereo amplifier may claim to introduce less that .01%
87total harmonic distortion when amplifying an input signal; this claim
88is easy to verify given proper equipment, and any number of testers are
89likely to arrive at the same, exact results. One need not listen to
90the equipment to make this measurement.</p>
91
92<p>However, given two amplifiers with identical, verifiable objective
93specifications, listeners may strongly prefer the sound quality of one
94over the other. This is actually the case in the decades old debate
95[some would say jihad] among audiophiles involving vacuum tube versus
96solid state amplifiers. There are people who can tell the difference,
97and strongly prefer one over the other despite seemingly identical,
98measurable quality. This preference is <em>subjective</em> and
99difficult to measure but nonetheless real.</p>
100
101<p>Individual elements of subjective differences often can be qualified,
102but overall subjective quality generally is not measurable. Different
103observers are likely to disagree on the exact results of a subjective
104test as each observer's perspective differs. When measuring
105subjective qualities, the best one can hope for is average, empirical
106results that show statistical significance across a group.</p>
107
108<p>Perceptual codecs are most concerned with subjective, not objective,
109quality. This is why evaluating a perceptual codec via distortion
110measures and sonograms alone is useless; these objective measures may
111provide insight into the quality or functioning of a codec, but cannot
112answer the much squishier subjective question, "Does it sound
113good?". The tube amplifier example is perhaps not the best as very few
114people can hear, or care to hear, the minute differences between tubes
115and transistors, whereas the subjective differences in perceptual
116codecs tend to be quite large even when objective differences are
117not.</p>
118
119<h2>Fidelity, Artifacts and Differences</h2>
120
121<p>Audio <em>artifacts</em> and loss of fidelity or more simply
122put, audio <em>differences</em> are not the same thing.</p>
123
124<p>A loss of fidelity implies differences between the perceived input and
125output signal; it does not necessarily imply that the differences in
126output are displeasing or that the output sounds poor (although this
127is often the case). Tube amplifiers are <em>not</em> higher fidelity
128than modern solid state and digital systems. They simply produce a
129form of distortion and coloring that is either unnoticeable or actually
130pleasing to many ears.</p>
131
132<p>As compared to an original signal using hard metrics, all perceptual
133codecs [ASPEC, ATRAC, MP3, WMA, AAC, TwinVQ, AC3 and Vorbis included]
134lose objective fidelity in order to reduce bitrate. This is fact. The
135idea is to lose fidelity in ways that cannot be perceived. However,
136most current streaming applications demand bitrates lower than what
137can be achieved by sacrificing only objective fidelity; this is also
138fact, despite whatever various company press releases might claim.
139Subjective fidelity eventually must suffer in one way or another.</p>
140
141<p>The goal is to choose the best possible tradeoff such that the
142fidelity loss is graceful and not obviously noticeable. Most listeners
143of FM radio do not realize how much lower fidelity that medium is as
144compared to compact discs or DAT. However, when compared directly to
145source material, the difference is obvious. A cassette tape is lower
146fidelity still, and yet the degradation, relatively speaking, is
147graceful and generally easy not to notice. Compare this graceful loss
148of quality to an average 44.1kHz stereo mp3 encoded at 80 or 96kbps.
149The mp3 might actually be higher objective fidelity but subjectively
150sounds much worse.</p>
151
152<p>Thus, when a CODEC <em>must</em> sacrifice subjective quality in order
153to satisfy a user's requirements, the result should be a
154<em>difference</em> that is generally either difficult to notice
155without comparison, or easy to ignore. An <em>artifact</em>, on the
156other hand, is an element introduced into the output that is
157immediately noticeable, obviously foreign, and undesired. The famous
158'underwater' or 'twinkling' effect synonymous with low bitrate (or
159poorly encoded) mp3 is an example of an <em>artifact</em>. This
160working definition differs slightly from common usage, but the coined
161distinction between differences and artifacts is useful for our
162discussion.</p>
163
164<p>The goal, when it is absolutely necessary to sacrifice subjective
165fidelity, is obviously to strive for differences and not artifacts.
166The vast majority of codecs today fail at this task miserably,
167predictably, and regularly in one way or another. Avoiding such
168failures when it is necessary to sacrifice subjective quality is a
169fundamental design objective of Vorbis and that objective is reflected
170in Vorbis's design and tuning.</p>
171
172<div id="copyright">
173 The Xiph Fish Logo is a
174 trademark (&trade;) of Xiph.Org.<br/>
175
176 These pages &copy; 1994 - 2005 Xiph.Org. All rights reserved.
177</div>
178
179</body>
180</html>
Note: See TracBrowser for help on using the repository browser.

© 2024 Oracle Support Privacy / Do Not Sell My Info Terms of Use Trademark Policy Automated Access Etiquette