1SoX(7) Sound eXchange SoX(7)
2
6 soxexam - SoX Examples
7
9 Introduction
10 The core problem is that you need some experience in using effects in
11 order to say `that any old sound file sounds with effects absolutely
12 hip'. There isn't any rule-based system which tell you the correct set‐
13 ting of all the parameters for every effect. But after some time you
14 will become an expert in using effects.
15 Here are some examples which can be used with any music sample. (For a
17 sample where only a single instrument is playing, extreme parameter
18 setting may make well-known `typically' or `classical' sounds. Like‐
19 wise, for drums, vocals or guitars.)
20 Single effects will be explained and some given parameter settings that
22 can be used to understand the theory by listening to the sound file
23 with the added effect.
24 Using multiple effects in parallel or in series can result either in a
26 very nice sound or (mostly) in a dramatic overloading in variations of
27 sounds such that your ear may follow the sound but you will feel unsat‐
28 isfied. Hence, for the first time using effects try to compose them as
29 minimally as possible. We don't regard the composition of effects in
30 the examples because too many combinations are possible and you really
31 need a very fast machine and a lot of memory to play them in real-time.
32 However, real-time playing of sounds will greatly speed up learning
34 and/or tuning the parameter settings for your sounds in order to get
35 that `perfect' effect.
36 Basically, we will use the `play' front-end of SoX since it is easier
38 to listen sounds coming out of the speaker or earphone instead of look‐
39 ing at cryptic data in sound files.
40 For easy listening of file.xxx (`xxx' is any sound format):
42 play file.xxx effect-name effect-parameters
44 Or more SoX-like (for `dsp' output on a UNIX/Linux computer):
46 sox file.xxx -t ossdsp -2 -s /dev/dsp effect-name effect-parame‐
48 ters
49 or (for `au' output):
51 sox file.xxx -t sunau -2 -s /dev/audio effect-name effect-parame‐
53 ters
54 And for date freaks:
56 sox file.xxx file.yyy effect-name effect-parameters
58 Additional options can be used. However, in this case, for real-time
60 playing you'll need a very fast machine.
61 Notes:
63 I played all examples in real-time on a Pentium 100 with 32 MB and
65 Linux 2.0.30 using a self-recorded sample ( 3:15 min long in `wav' for‐
66 mat with 44.1 kHz sample rate and stereo 16 bit ). The sample should
67 not contain any of the effects. However, if you take any recording of a
68 sound track from radio or tape or CD, and it sounds like a live concert
69 or ten people are playing the same rhythm with their drums or funky-
70 grooves, then take any other sample. (Typically, less then four dif‐
71 ferent instruments and no synthesizer in the sample is suitable. Like‐
72 wise, the combination vocal, drums, bass and guitar.)
73 Echo
75 An echo effect can be naturally found in the mountains, standing some‐
76 where on a mountain and shouting a single word will result in one or
77 more repetitions of the word (if not, turn a bit around and try again,
78 or climb to the next mountain).
79 However, the time difference between shouting and repeating is the
81 delay (time), its loudness is the decay. Multiple echos can have dif‐
82 ferent delays and decays.
83 It is very popular to use echos to play an instrument with itself
85 together, like some guitar players (Queen's Brain May) or vocalists do.
86 For music samples of more than one instrument, echo can be used to add
87 a second sample shortly after the original one.
88 This will sound as if you are doubling the number of instruments play‐
90 ing in the same sample:
91 play file.xxx echo 0.8 0.88 60 0.4
93 If the delay is very short, then it sound like a (metallic) robot play‐
95 ing music:
96 play file.xxx echo 0.8 0.88 6 0.4
98 Longer delay will sound like an open air concert in the mountains:
100 play file.xxx echo 0.8 0.9 1000 0.3
102 One mountain more, and:
104 play file.xxx echo 0.8 0.9 1000 0.3 1800 0.25
106 Echos
108 Like the echo effect, echos stand for `ECHO in Sequel', that is the
109 first echos takes the input, the second the input and the first echos,
110 the third the input and the first and the second echos, ... and so on.
111 Care should be taken using many echos (see introduction); a single
112 echos has the same effect as a single echo.
113 The sample will be bounced twice in symmetric echos:
115 play file.xxx echos 0.8 0.7 700 0.25 700 0.3
117 The sample will be bounced twice in asymmetric echos:
119 play file.xxx echos 0.8 0.7 700 0.25 900 0.3
121 The sample will sound as if played in a garage:
123 play file.xxx echos 0.8 0.7 40 0.25 63 0.3
125 Chorus
127 The chorus effect has its name because it will often be used to make a
128 single vocal sound like a chorus. But it can be applied to other
129 instrument samples too.
130 It works like the echo effect with a short delay, but the delay isn't
132 constant. The delay is varied using a sinusoidal or triangular modula‐
133 tion. The modulation depth defines the range the modulated delay is
134 played before or after the delay. Hence the delayed sound will sound
135 slower or faster, that is the delayed sound tuned around the original
136 one, like in a chorus where some vocals are a bit out of tune.
137 The typical delay is around 40ms to 60ms, the speed of the modulation
139 is best near 0.25Hz and the modulation depth around 2ms.
140 A single delay will make the sample more overloaded:
142 play file.xxx chorus 0.7 0.9 55 0.4 0.25 2 -t
144 Two delays of the original samples sound like this:
146 play file.xxx chorus 0.6 0.9 50 0.4 0.25 2 -t 60 0.32 0.4 1.3 -s
148 A big chorus of the sample is (three additional samples):
150 play file.xxx chorus 0.5 0.9 50 0.4 0.25 2 -t 60 0.32 0.4 2.3 -t
152 40 0.3 0.3 1.3 -s
153 Flanger
155 The flanger effect is like the chorus effect, but the delay varies
156 between 0ms and maximal 5ms. It sound like wind blowing, sometimes
157 faster or slower including changes of the speed.
158 The flanger effect is widely used in funk and soul music, where the
160 guitar sound varies frequently slow or a bit faster.
161 Now, let's groove the sample:
163 play file.xxx flanger
165 listen carefully between the difference of sinusoidal and triangular
167 modulation:
168 play file.xxx flanger triangle
170 Reverb
172 The reverb effect is often used in audience hall which are to small or
173 contain too many many visitors which disturb (dampen) the reflection of
174 sound at the walls. Reverb will make the sound be perceived as if it
175 were in a large hall. You can try the reverb effect in your bathroom
176 or garage or sport hall by shouting loud some words. You'll hear the
177 words reflected from the walls.
178 The biggest problem in using the reverb effect is the correct setting
180 of the (wall) delays such that the sound is realistic and doesn't sound
181 like music playing in a tin can or has overloaded feedback which
182 destroys any illusion of playing in a big hall. To help you obtain
183 realistic reverb effects, you should decide first how long the reverb
184 should take place until it is not loud enough to be registered by your
185 ears. This is be done by varying the reverb time `t'. To simulate
186 small halls, use 200ms. To simulate large halls, use 1000ms. Clearly,
187 the walls of such a hall aren't far away, so you should define its set‐
188 ting be given every wall its delay time. However, if the wall is to
189 far away for the reverb time, you won't hear the reverb, so the nearest
190 wall will be best at `t/4' delay and the farthest at `t/2'. You can try
191 other distances as well, but it won't sound very realistic. The walls
192 shouldn't stand to close to each other and not in a multiple integer
193 distance to each other ( so avoid wall like: 200 and 202, or something
194 like 100 and 200 ).
195 Since audience halls do have a lot of walls, we will start designing
197 one beginning with one wall:
198 play file.xxx reverb 1 600 180
200 One wall more:
202 play file.xxx reverb 1 600 180 200
204 Next two walls:
206 play file.xxx reverb 1 600 180 200 220 240
208 Now, why not a futuristic hall with six walls:
210 play file.xxx reverb 1 600 180 200 220 240 280 300
212 If you run out of machine power or memory, then stop as many applica‐
214 tions as possible (every interrupt will consume a lot of CPU time which
215 for bigger halls is absolutely necessary).
216 Phaser
218 The phaser effect is like the flanger effect, but it uses a reverb
219 instead of an echo and does phase shifting. You'll hear the difference
220 in the examples comparing both effects. The delay modulation can be
221 sinusoidal or triangular, preferable is the later for multiple instru‐
222 ments. For single instrument sounds, the sinusoidal phaser effect will
223 give a sharper phasing effect. The decay shouldn't be to close to 1
224 which will cause dramatic feedback. A good range is about 0.5 to 0.1
225 for the decay.
226 We will take a parameter setting as before (gain-out is lower since
228 feedback can raise the output dramatically):
229 play file.xxx phaser 0.8 0.74 3 0.4 0.5 -t
231 The drunken loudspeaker system (now less alcohol):
233 play file.xxx phaser 0.9 0.85 4 0.23 1.3 -s
235 A popular sound of the sample is as follows:
237 play file.xxx phaser 0.89 0.85 1 0.24 2 -t
239 The sample sounds if ten springs are in your ears:
241 play file.xxx phaser 0.6 0.66 3 0.6 2 -t
243 Compander
245 The compander effect allows the dynamic range of a signal to be com‐
246 pressed or expanded. For most situations, the attack time (response to
247 the music getting louder) should be shorter than the decay time because
248 our ears are more sensitive to suddenly loud music than to suddenly
249 soft music.
250 For example, suppose you are listening to Strauss's `Also Sprach
252 Zarathustra' in a noisy environment such as a car. If you turn up the
253 volume enough to hear the soft passages over the road noise, the loud
254 sections will be too loud. You could try this:
255 play file.xxx compand 0.3,1 -90,-90,-70,-70,-60,-20,0,0 -5 0 0.2
257 The transfer function (`-90,...') says that very soft sounds between
259 -90 and -70 decibels (-90 is about the limit of 16-bit encoding) will
260 remain unchanged. That keeps the compander from boosting the volume on
261 `silent' passages such as between movements. However, sounds in the
262 range -60 decibels to 0 decibels (maximum volume) will be boosted so
263 that the 60dB dynamic range of the original music will be compressed
264 3-to-1 into a 20dB range, which is wide enough to enjoy the music but
265 narrow enough to get around the road noise. The -5 dB output gain is
266 needed to avoid clipping (the number is inexact, and was derived by
267 experimentation). The 0 for the initial volume will work fine for a
268 clip that starts with a bit of silence, and the delay of 0.2 has the
269 effect of causing the compander to react a bit more quickly to sudden
270 volume changes.
271 The following (one long) command shows how multi-band companding is
273 typically used in FM radio:
274 play file.xxx
276 vol -3 db filter 8000- 32 100 mcompand "0.005,0.1
277 -47,-40,-34,-34,-17,-33 0 0 0" 100 "0.003,0.05
278 -47,-40,-34,-34,-17,-33 0 0 0" 400 "0.000625,0.0125
279 -47,-40,-34,-34,-15,-33 0 0 0" 1600 "0.0001,0.025
280 -47,-40,-34,-34,-31,-31,-0,-30 0 0 0" 6400
281 "0,0.025 -38,-31,-28,-28,-0,-25 0 0 0" vol 27 db vol -12
282 db highpass 22 highpass 22 filter -17500 256 vol +12 db
283 vol -3 db lowp 17801
284 The audio file is played with a simulated FM radio sound (or broadcast
286 signal condition if the lowp at the end is skipped). Note that the
287 pipeline is set up with US-style 75us preemphasis.
288 Changing the Rate of Playback
290 You can use stretch to change the rate of playback of an audio sample
291 while preserving the pitch. For example to play at half the speed:
292 play file.wav stretch 2
294 To play a file at twice the speed:
296 play file.wav stretch 0.5
298 Other related options are `speed' to change the speed of play (and
300 changing the pitch accordingly), and pitch, to alter the pitch of a
301 sample. For example to speed a sample so it plays in half the time
302 (for those Mickey Mouse voices):
303 play file.wav speed 2
305 To raise the pitch of a sample 1 while note (100 cents):
307 play file.wav pitch 100
309 Reducing noise in a recording
312 First find a period of silence in your recording, such as the beginning
313 or end of a piece. If the first 1.5 seconds of the recording are
314 silent, do
315 sox file.wav -n trim 0 1.5 noiseprof /tmp/profile
317 Next, use the noisered effect to actually reduce the noise:
319 play file.wav noisered /tmp/profile
321 Making a recording
323 Thanks to Douglas Held for the following suggestion:
324 rec parameters filename other-effects silence 1 5 2%
326 This use of the silence effect allows you to start a recording session
328 but only start writing to disk once non-silence is detect. For example,
329 use this to start your favorite command line for recording and walk
330 over to your record player and start the song. No periods of silence
331 will be recorded.
332 Scripting with SoX
334 One of the benefits of a command-line tool is that it is easy to use it
335 in scripts to perform more complex tasks. In marine radio, a Mayday
336 emergency call is transmitted preceded by a 30-second alert sound. The
337 alert sound comprises two audio tones at 1300Hz and 2100Hz alternating
338 at a rate of 4Hz. The following shows how SoX can be used in a script
339 to construct an audio file containing the alert sound. The scripting
340 language shown is `Bourne shell' (sh) but it should be simple to trans‐
341 late this to another scripting language if you do not have access to
342 sh.
343 # Make sure we append to a file that's initially empty:
345 rm -f 2tones.raw
346 for freq in 1300 2200; do
348 sox -c1 -r8000 -n -t raw - synth 0.25 sine $freq vol 0.7 >> 2tones.raw
349 done
350 # We need 60 copies of 2tones.raw (0.5 sec) to get 30 secs of audio:
352 iterations=60
353 # Make sure we append to a file that's initially empty:
355 rm -f alert.raw
356 while [ $iterations -ge 1 ]; do
358 cat 2tones.raw >> alert.raw
359 iterations=`expr $iterations - 1`
360 done
361 # Add a file header and save some disc space:
363 sox -sw -c1 -r8000 alert.raw alert.ogg
364 play alert.ogg
366 If you try out the above script, you may want to hit Ctrl-C fairly soon
368 after the alert tone starts playing - it's not a pleasant sound! The
369 synth effect is used to generate each of the tones; -c1 -r8000 selects
370 mono, 8kHz sampling-rate audio (i.e. relatively low fidelity, suitable
371 for the marine radio transmission channel); each tone is generated at a
372 length of 0.25 seconds to give the required 4Hz alternation. Note the
373 use of `raw' as the intermediary file format; a self-describing
374 (header) format would just get in the way here. The self-describing
375 header is added only at the final stage; in this case, .ogg is chosen,
376 since lossy compression is appropriate for this application.
377 There are further practical examples of scripting with SoX available to
379 download from the SoX web-site [1].
380
382 sox(1), libst(3)
383 References
385 [1] SoX - Sound eXchange | Scripts, http://sox.source‐
386 forge.net/Docs/Scripts
387
389 This man page was written largely by Juergen Mueller
390 (jmueller@uia.ua.ac.be). Other SoX authors and contributors are listed
391 in the AUTHORS file that is distributed with the source code.
392soxexam January 31, 2007 SoX(7)