COMPRESSOR INSERTS
The side-chain normally “listens” directly to the signal being compressed, though an insert point or “key” input is often fitted to enable other processors to be connected into the side-chain signal path, or to allow the side-chain to be fed from a different source altogether. For example, you could patch an equaliser into the side chain. Why should you want to equalise the side-chain signal of a compressor?
By boosting signals in the 5 to 8kHz range using an external equaliser, the compressor can be made to respond mainly to loud, toppy sounds in the 5 to 8kHz range – a technique often used for “de-essing” sibilant vocal tracks. Example: in this instance, the equaliser would be adjusted to boost the sibilant portion of the sound by first monitoring the output of the filter and tuning into the offending part of the spectrum by ear. If 10dB of boost is applied in this range, then the compressor will be 10dB more sensitive to sibilant t frequencies than to the rest of the audio spectrum. To make setting up easy, some compressors (and gates) have a side-chain listen switch which, when activated, replaces the normal output with the side-chain control signal so that it can be monitored over the control room speakers during setting up.
IN practice, a dedicated de-esser gives better results than a compressor hooked up to an equaliser, but if you need to de-ess something in a hurry and don’t have access to a dedicated unit, this arrangement will work providing you don’t process too heavily, otherwise the vocals will become “lispy” sounding. That’s because we only really want to bring down the level of sibilant sounds, but with this arrangement, whenever a sibilant sound is detected, the level of the whole signal is reduced. A commercial de-esser is more likely only to attenuate the troublesome frequencies, so it has fewer audible side effects.
PULSAR COMPRESSORS
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
DUCKING
Another reason for needing access to the side-chain is to enable the level of one signal to control another. By feeding music through the compressor and connecting a voice input to the side-chain, the music can be reduced in volume to make way for the voiceover. When the voice stops, the music returns to its previous level at a rate determined by the release control. This technique, often used by Djs, is called “ducking”. To achieve the effect, you will need some kind of mixer which has a mic channel with a direct output or insert send to feed the compressor-s side-chain. The reason for this is that the side-chain input on most compressors will only accept line level signals and so the mic signal needs to be amplified before it can be used.
Another reason for needing access to the side-chain is to enable the level of one signal to control another. By feeding music through the compressor and connecting a voice input to the side-chain, the music can be reduced in volume to make way for the voiceover. When the voice stops, the music returns to its previous level at a rate determined by the release control. This technique, often used by Djs, is called “ducking”. To achieve the effect, you will need some kind of mixer which has a mic channel with a direct output or insert send to feed the compressor-s side-chain. The reason for this is that the side-chain input on most compressors will only accept line level signals and so the mic signal needs to be amplified before it can be used.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
SOFT OR HARD KNEE
While most conventional compressors work on the threshold or hard knee principle, there is another popular type which takes the so-called “soft knee” approach. There, some compression may be applied to all signals, regardless of level, but the compression ratio is very low for small signals and automatically increases as the signal level increases. This makes for a very easy compressor to set up as there’s no threshold or ratio control, just a compression amount know. However, not all soft knee compressors are this simple. More commonly there will be a threshold in the same way as for a conventional hard knee compressor, but the compression ratio still increases progressively as the signal approaches the threshold. By the time the signal reaches the threshold, the full compression ratio set by the user is in force. In either case, the idea is to higher level signals that are compressed.
As a rule, soft knee compressors sound less obtrusive than hard knee compressors, but may not exercise such firm gain control. Ideally, you need a model that’s switchable between hard and soft knee so that you can match the type of compression to the sound you’re processing.
While most conventional compressors work on the threshold or hard knee principle, there is another popular type which takes the so-called “soft knee” approach. There, some compression may be applied to all signals, regardless of level, but the compression ratio is very low for small signals and automatically increases as the signal level increases. This makes for a very easy compressor to set up as there’s no threshold or ratio control, just a compression amount know. However, not all soft knee compressors are this simple. More commonly there will be a threshold in the same way as for a conventional hard knee compressor, but the compression ratio still increases progressively as the signal approaches the threshold. By the time the signal reaches the threshold, the full compression ratio set by the user is in force. In either case, the idea is to higher level signals that are compressed.
As a rule, soft knee compressors sound less obtrusive than hard knee compressors, but may not exercise such firm gain control. Ideally, you need a model that’s switchable between hard and soft knee so that you can match the type of compression to the sound you’re processing.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
BUILT-IN GATES (do not mistake with Bill Gates… he he…)
Because a compressor can’t differentiate between a small wanted signal and low level unwanted noise, it tends to emphasise such noise by setting itself to maximum gain during quiet passages such as gaps between words or phrases in a vocal track. As a result, there is now a tendency for some manufacturers to build simple gates or expanders into their compressors in order to mute the signal at such times. This gate/expander would typically have only a threshold control and an LED to show when it’s closed. The attack and release times are normally preset or electronically linked to those of the compressor.
Suffice it to say here, that the gate should be set with as low a threshold as is practical to deal with the noise, otherwise quieter sections of wanted sounds could be muted. This kind of gating is generally of value only when working on separate tracks within a mix, as complete mixes tend to have few, if any periods of complete silence in which the gage could operate without making its presence known.
Noise is a particular problem when compressing signals already recorded onto analogue tape, simply because tale noise will be increased in level by the compressor. For this reason, it is usual to apply some degree of compression to vocal tracks during recording to ensure a healthy signal level going onto tape at all times. Though digital recording systems tend to be quieter, there’s often some hum and mic amp hiss recorded along with the original signal, so gating may still be desirable.
At the mixing stage, further compression may be applied, as needed, paying due regard to the effect of compression on any background noise present. As a general rule, every dB of compression is a dB of deterioration in the signal to noise ratio of the signal passing through it. In other words, if you compress a signal to produce 10dB of gain reduction, the noise during quieter sections will be increased by 10dB.
Because a compressor can’t differentiate between a small wanted signal and low level unwanted noise, it tends to emphasise such noise by setting itself to maximum gain during quiet passages such as gaps between words or phrases in a vocal track. As a result, there is now a tendency for some manufacturers to build simple gates or expanders into their compressors in order to mute the signal at such times. This gate/expander would typically have only a threshold control and an LED to show when it’s closed. The attack and release times are normally preset or electronically linked to those of the compressor.
Suffice it to say here, that the gate should be set with as low a threshold as is practical to deal with the noise, otherwise quieter sections of wanted sounds could be muted. This kind of gating is generally of value only when working on separate tracks within a mix, as complete mixes tend to have few, if any periods of complete silence in which the gage could operate without making its presence known.
Noise is a particular problem when compressing signals already recorded onto analogue tape, simply because tale noise will be increased in level by the compressor. For this reason, it is usual to apply some degree of compression to vocal tracks during recording to ensure a healthy signal level going onto tape at all times. Though digital recording systems tend to be quieter, there’s often some hum and mic amp hiss recorded along with the original signal, so gating may still be desirable.
At the mixing stage, further compression may be applied, as needed, paying due regard to the effect of compression on any background noise present. As a general rule, every dB of compression is a dB of deterioration in the signal to noise ratio of the signal passing through it. In other words, if you compress a signal to produce 10dB of gain reduction, the noise during quieter sections will be increased by 10dB.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
STEREO LINKING
Two-channel compressors generally include a stereo link switch for applications that involve the compressing of a stereo signal. Using two independent compressors in this role can cause image shifts, especially if there are loud peaks on one side of the mix that aren’t on the other. What’s needed is for both channels to undergo exactly the same amount of gain reduction at all times, regardless of whether the biggest peaks are in the left or right channel, and that’s where the “stereo link” switch is needed. This combines the side-chain signals in such a way that the compressor responds to an average of the two channel levels, and often you’ll find that the compressor can be adjusted using only the controls of the first channel. Other designs may average the control settings of the two channels.
It’s interesting to note that for proper stereo operation, each channel still needs a separate envelope follower as part of the side chain. If you were simply to combine the left and right side-chain signals (as some cheaper models do), then extract the combined envelope; you could get quite different results caused by phase cancellations between the left and right channels. This would, in turn, lead to inadequate amounts of gain reduction being applied to some sounds.
Two-channel compressors generally include a stereo link switch for applications that involve the compressing of a stereo signal. Using two independent compressors in this role can cause image shifts, especially if there are loud peaks on one side of the mix that aren’t on the other. What’s needed is for both channels to undergo exactly the same amount of gain reduction at all times, regardless of whether the biggest peaks are in the left or right channel, and that’s where the “stereo link” switch is needed. This combines the side-chain signals in such a way that the compressor responds to an average of the two channel levels, and often you’ll find that the compressor can be adjusted using only the controls of the first channel. Other designs may average the control settings of the two channels.
It’s interesting to note that for proper stereo operation, each channel still needs a separate envelope follower as part of the side chain. If you were simply to combine the left and right side-chain signals (as some cheaper models do), then extract the combined envelope; you could get quite different results caused by phase cancellations between the left and right channels. This would, in turn, lead to inadequate amounts of gain reduction being applied to some sounds.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
VARIATIONS ON A THEME
Not all compressors work in exactly the same way or have the same control layout. One alternative to the variable threshold system is to have a fixed threshold preceded by a variable gain input stage. This has the same effect as varying the threshold while having a fixed input level. – although an output gain control is still desirable to balance up the overall gain through the compressor.
Compressors with gain control elements based on LEDs and photocells have a smooth, vintage sound characterised by a relatively slow attack and fast release time. They are currently very popular, because even though the processed sound is quite obviously being compressed, the side effects are quite flattering to vocals and instruments such as acoustic guitar.
A number of compressors are also built with tubes in the gain control stage, and these again tend to produce a characteristic sound, which some describe as warm and musical. The precise sound depends on how the circuitry is designed, and whether the tube runs off a full high voltage supply or a reduced DC voltage. See the section on tube equipment for more details.
Compressor designs are also available that use a complex side-chain design to maintain a long term average signal level as well as compressing peaks in the signal in the normal way. They may also include a facility that detects periods of silence between sounds and prevent the compressor gain from rising. Done properly, this brings about a significant improvement in signal to noise ratio without recourse to gates. A multiband limiter might well be used in line with a radio transmitter to ensure that the average signal level is always constant, but there are also applications in mastering to a digital medium where any level overshoots, however brief, cannot be tolerated.
Not all compressors work in exactly the same way or have the same control layout. One alternative to the variable threshold system is to have a fixed threshold preceded by a variable gain input stage. This has the same effect as varying the threshold while having a fixed input level. – although an output gain control is still desirable to balance up the overall gain through the compressor.
Compressors with gain control elements based on LEDs and photocells have a smooth, vintage sound characterised by a relatively slow attack and fast release time. They are currently very popular, because even though the processed sound is quite obviously being compressed, the side effects are quite flattering to vocals and instruments such as acoustic guitar.
A number of compressors are also built with tubes in the gain control stage, and these again tend to produce a characteristic sound, which some describe as warm and musical. The precise sound depends on how the circuitry is designed, and whether the tube runs off a full high voltage supply or a reduced DC voltage. See the section on tube equipment for more details.
Compressor designs are also available that use a complex side-chain design to maintain a long term average signal level as well as compressing peaks in the signal in the normal way. They may also include a facility that detects periods of silence between sounds and prevent the compressor gain from rising. Done properly, this brings about a significant improvement in signal to noise ratio without recourse to gates. A multiband limiter might well be used in line with a radio transmitter to ensure that the average signal level is always constant, but there are also applications in mastering to a digital medium where any level overshoots, however brief, cannot be tolerated.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
AUTO MODE
Another refinement which is now becoming more common on lower priced processors is auto attack and release, though some models only use an auto release in combination with a fixed attack. Instead of the user fixing an attack and release time, the compressor incorporates a circuit which analyses the dynamic characteristics of the incoming signal and varies the attack and release times so as to produce the least obtrusive compression. This is valuable where the signal dynamics are constantly changing, making fixed settings inappropriate. An example of this is the bass guitar which may be slapped, pulled and played conventionally during the course of a single piece of music. All these styles have different dynamics, and a compressor which can continually optimise its settings is clearly an advantage under these conditions. Similarly, if a complete mix is being compressed to increase the average signal level, a fixed attack and release time is unlikely to be optimum for all parts of the song. Again, a compressor with a auto setting should give consistently better results.
Another refinement which is now becoming more common on lower priced processors is auto attack and release, though some models only use an auto release in combination with a fixed attack. Instead of the user fixing an attack and release time, the compressor incorporates a circuit which analyses the dynamic characteristics of the incoming signal and varies the attack and release times so as to produce the least obtrusive compression. This is valuable where the signal dynamics are constantly changing, making fixed settings inappropriate. An example of this is the bass guitar which may be slapped, pulled and played conventionally during the course of a single piece of music. All these styles have different dynamics, and a compressor which can continually optimise its settings is clearly an advantage under these conditions. Similarly, if a complete mix is being compressed to increase the average signal level, a fixed attack and release time is unlikely to be optimum for all parts of the song. Again, a compressor with a auto setting should give consistently better results.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
SPLIT BAND COMPRESSION
It was explained earlier why compressors usually cause some dulling of the processed sound, due to low frequency peaks such as loud bass drum beats causing attenuation of the whole signal including low level, high frequency sounds. One way round this is to use a compressor or limiter that splits the input into two or more frequency bands, then compresses each independently before recombining them. This ensures that a loud peak occurring in one part of the audio spectrum won’t affect the level of frequencies in other bands, but as you might imagine, this is complicated and expensive due to the fact that each band needs its own compressor circuit as well as the filters necessary to split the bands. Also, some method has to be arranged to make the attack and release times of each band complement each other.
Low cost, two-band compressors are quite common, but in more discerning applications, special circuitry has to be used to combine the bands intelligently, otherwise unwanted phase changes and other anomalies can be introduced. Typically, a sophisticated multi-band compressor or limiter might cost four or five times as much as a conventional one.
Note that a two-band compressor can make quite an effective de-esser, especially if you can adjust the frequency split point between the bands. By setting the split at 4 or 5kHz, loud sibilant sounds will only cause compression of frequencies above 5kHz, leaving the lower band unaffected.
It was explained earlier why compressors usually cause some dulling of the processed sound, due to low frequency peaks such as loud bass drum beats causing attenuation of the whole signal including low level, high frequency sounds. One way round this is to use a compressor or limiter that splits the input into two or more frequency bands, then compresses each independently before recombining them. This ensures that a loud peak occurring in one part of the audio spectrum won’t affect the level of frequencies in other bands, but as you might imagine, this is complicated and expensive due to the fact that each band needs its own compressor circuit as well as the filters necessary to split the bands. Also, some method has to be arranged to make the attack and release times of each band complement each other.
Low cost, two-band compressors are quite common, but in more discerning applications, special circuitry has to be used to combine the bands intelligently, otherwise unwanted phase changes and other anomalies can be introduced. Typically, a sophisticated multi-band compressor or limiter might cost four or five times as much as a conventional one.
Note that a two-band compressor can make quite an effective de-esser, especially if you can adjust the frequency split point between the bands. By setting the split at 4 or 5kHz, loud sibilant sounds will only cause compression of frequencies above 5kHz, leaving the lower band unaffected.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
DIGITAL COMPRESSORS
At the time of writing, there are relatively few stand-alone digital hardware compressors, though they are found built into digital mixers or as software plug-ins for digital audio workstations. In theory, a digital compressor can do everything an analogue compressor can do, though it has the additional advantage of being able to delay the signal slightly so as to allow the side-chain signal to “see” what’s coming a fraction of a second before it arrives at the gain control section. This is known as a “look-ahead” system, and it can help a compressor or limiter deal more effectively with signal peaks. By having a little advance warning of peaks, compressors can ensure that no overshoot will take place, and this is again very important in digital mastering applications. In most other respects, software-based digital compressors emulate the controls and parameters of an analogue compressor, though the larger area of a computer screen makes possible the use of more intuitive graphical displays.
At the time of writing, there are relatively few stand-alone digital hardware compressors, though they are found built into digital mixers or as software plug-ins for digital audio workstations. In theory, a digital compressor can do everything an analogue compressor can do, though it has the additional advantage of being able to delay the signal slightly so as to allow the side-chain signal to “see” what’s coming a fraction of a second before it arrives at the gain control section. This is known as a “look-ahead” system, and it can help a compressor or limiter deal more effectively with signal peaks. By having a little advance warning of peaks, compressors can ensure that no overshoot will take place, and this is again very important in digital mastering applications. In most other respects, software-based digital compressors emulate the controls and parameters of an analogue compressor, though the larger area of a computer screen makes possible the use of more intuitive graphical displays.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
COMPRESSOR APPLICATIONS
Compressors are useful for reducing the dynamic range of both signals being recorded onto a multitrack machine and signals being mixed from a multitrack to a master recorder.
The only real difference between a compressor and a limiter is in the compression ration and there is no hard and fast figure that divides the two, though to ensure no overshoot under any circumstances, a dedicated peak limiter is required, as a general-purpose compressor may not respond quickly enough. Strictly speaking, a ratio of infinity: 1 would be needed to ensure that the signal never exceeds the threshold but in practice, a compressor set with a ratio for over 10:1 has much the same audible effect.
Always be prepared to experiment with the compressors at your disposal: not only do ratio types sound somewhat different to soft knee types, but one make or model will usually sound quite unlike another – even when set up in a similar way. The subtle harmonic distortion created by valve compressors, for example, has led many engineers to favour them for their warm sound when used on vocals. Most importantly, don’t make the mistake of overcompressing: music depends on dynamics to convey its emotional message, and if you flatten the dynamics entirely, you may end up with something that’s easy to record, but musically unsatisfying.
The following compression setting should only be used as a guide or starting point. It is not intended that they be regarded as a set of rigidly defined rules; each situation is different and the final results must ultimately be determined by ear.
Compressors are useful for reducing the dynamic range of both signals being recorded onto a multitrack machine and signals being mixed from a multitrack to a master recorder.
The only real difference between a compressor and a limiter is in the compression ration and there is no hard and fast figure that divides the two, though to ensure no overshoot under any circumstances, a dedicated peak limiter is required, as a general-purpose compressor may not respond quickly enough. Strictly speaking, a ratio of infinity: 1 would be needed to ensure that the signal never exceeds the threshold but in practice, a compressor set with a ratio for over 10:1 has much the same audible effect.
Always be prepared to experiment with the compressors at your disposal: not only do ratio types sound somewhat different to soft knee types, but one make or model will usually sound quite unlike another – even when set up in a similar way. The subtle harmonic distortion created by valve compressors, for example, has led many engineers to favour them for their warm sound when used on vocals. Most importantly, don’t make the mistake of overcompressing: music depends on dynamics to convey its emotional message, and if you flatten the dynamics entirely, you may end up with something that’s easy to record, but musically unsatisfying.
The following compression setting should only be used as a guide or starting point. It is not intended that they be regarded as a set of rigidly defined rules; each situation is different and the final results must ultimately be determined by ear.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
VOCALS
Either a standard ratio compressor or a soft knee model can be used on vocals though my personal preference is for a soft knee type as it is possible to apply quite a lot of compression without creating an artificial sound. However, there are occasions when obvious compression makes a vocal sound more exciting, so try your hard-knee setting too. If a ratio type is used, select a ratio of around 4:1 to start with and then fine tune it by ear, though you may want to go as high as 8:1 for a hard compressed rock vocal sound.
Because we don’t normally want a sharp attack to the words in a vocal line, a fairly fast attack time should be used so that gain control starts immediately. A release time of around half a second is usually a good starting point and little or no hold time is necessary. Alternatively, use a long attack time of say 50mS and a fast release time of around 100mS to emulate a vintage photo-electric compressor. Try this in both hard and soft knee modes. The degree of compression require depends on the performer, and as with any type of signal processing, it is advisable to use as little as is necessary to achieve the desired result. For pop vocals, you might find that between 10 and 12kB of gain reduction during loud sections is fairly typical, though a well-trained vocalist may well need less.
Providing there is no excessive crosstalk on the vocal track from other instruments or voices, it should be possible to use the compressor’s own expander/gate to clean up pauses between words and phrases (you can always use an external gate if your compressor isn’t fitted with one). This cleaning up process id desirable because the gain increase during quiet sections can often raise the level of breath sounds, lip “smacks”, teeth clicks and electronic hiss to an unacceptable degree – especially when high levels of compression are necessary.
It is common practice to compress vocals as they are recorded, but to err on the side of undercompression. This is so that an otherwise perfect take isn’t ruined by excessive compression; when you come to mix, a little more compression can always be applied if necessary – but it is much harder to repair the effects of overcompression.
If the vocalist has a particularly sibilant style and even changing the microphone type and position doesn’t cure the problem entirely, then the additional compression employed during the mix can be made frequency conscious to act as a “de-esser”. By inserting an equaliser in the side-chain peaked at somewhere between 4 and 10kHz, it should be possible to reduce the effect of overpronounced “S” sounds – but care must be taken not to make the processing obvious by overuse. The exact equaliser frequency may be found by switching the compressor to “side-chain listen” and then tuning the equaliser until the sibilance is most pronounced. For this reason, a sweep or parametric equaliser is preferable to a graphic as the frequency needing attention may fall between the slider frequencies making precise tuning impossible. As started earlier, a dedicated de-esser or split band compressor will generally produce better results if you have access to one.
Either a standard ratio compressor or a soft knee model can be used on vocals though my personal preference is for a soft knee type as it is possible to apply quite a lot of compression without creating an artificial sound. However, there are occasions when obvious compression makes a vocal sound more exciting, so try your hard-knee setting too. If a ratio type is used, select a ratio of around 4:1 to start with and then fine tune it by ear, though you may want to go as high as 8:1 for a hard compressed rock vocal sound.
Because we don’t normally want a sharp attack to the words in a vocal line, a fairly fast attack time should be used so that gain control starts immediately. A release time of around half a second is usually a good starting point and little or no hold time is necessary. Alternatively, use a long attack time of say 50mS and a fast release time of around 100mS to emulate a vintage photo-electric compressor. Try this in both hard and soft knee modes. The degree of compression require depends on the performer, and as with any type of signal processing, it is advisable to use as little as is necessary to achieve the desired result. For pop vocals, you might find that between 10 and 12kB of gain reduction during loud sections is fairly typical, though a well-trained vocalist may well need less.
Providing there is no excessive crosstalk on the vocal track from other instruments or voices, it should be possible to use the compressor’s own expander/gate to clean up pauses between words and phrases (you can always use an external gate if your compressor isn’t fitted with one). This cleaning up process id desirable because the gain increase during quiet sections can often raise the level of breath sounds, lip “smacks”, teeth clicks and electronic hiss to an unacceptable degree – especially when high levels of compression are necessary.
It is common practice to compress vocals as they are recorded, but to err on the side of undercompression. This is so that an otherwise perfect take isn’t ruined by excessive compression; when you come to mix, a little more compression can always be applied if necessary – but it is much harder to repair the effects of overcompression.
If the vocalist has a particularly sibilant style and even changing the microphone type and position doesn’t cure the problem entirely, then the additional compression employed during the mix can be made frequency conscious to act as a “de-esser”. By inserting an equaliser in the side-chain peaked at somewhere between 4 and 10kHz, it should be possible to reduce the effect of overpronounced “S” sounds – but care must be taken not to make the processing obvious by overuse. The exact equaliser frequency may be found by switching the compressor to “side-chain listen” and then tuning the equaliser until the sibilance is most pronounced. For this reason, a sweep or parametric equaliser is preferable to a graphic as the frequency needing attention may fall between the slider frequencies making precise tuning impossible. As started earlier, a dedicated de-esser or split band compressor will generally produce better results if you have access to one.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
DE-POPPING
By setting the equaliser frequency to around 50Hz, the compressor can also be used to reduce the level of @popping@ on a vocal recording. However, it has to be said that this is an area where prevention is far better than cure. A simple pop shield made from a pair of tights stretched over a wire frame or even a chip pan splashguard placed between the singer and the microphone normally eliminates all traces of popping witnout otherwise affecting the sound. If de-popping is necessary, a split-band compressor with a split point at around 200Hz will usually produce fewer objectionalbe side effects.
By setting the equaliser frequency to around 50Hz, the compressor can also be used to reduce the level of @popping@ on a vocal recording. However, it has to be said that this is an area where prevention is far better than cure. A simple pop shield made from a pair of tights stretched over a wire frame or even a chip pan splashguard placed between the singer and the microphone normally eliminates all traces of popping witnout otherwise affecting the sound. If de-popping is necessary, a split-band compressor with a split point at around 200Hz will usually produce fewer objectionalbe side effects.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
ACOUSTIC GUITAR
Steel strung acoustic guitars may be given a denser, more even tone through the use of compression – the settings, with the possible exception of attack time, being somewhat similar to those used for vocals. To give the guitar a nice zingy attack, the compressor’s attack control should be set to anything from 10 to 40ms so that the attack of each note or chord passes through unsuppressed. Release time can be from 0.1 to 0.25 seconds depending on the effect you want to achieve, and for a really glassy tone, try an opto compressor. Incidentally, a brighter-than-life sound can also be created by using an aural enhancer or exciter after the compressor.
Steel strung acoustic guitars may be given a denser, more even tone through the use of compression – the settings, with the possible exception of attack time, being somewhat similar to those used for vocals. To give the guitar a nice zingy attack, the compressor’s attack control should be set to anything from 10 to 40ms so that the attack of each note or chord passes through unsuppressed. Release time can be from 0.1 to 0.25 seconds depending on the effect you want to achieve, and for a really glassy tone, try an opto compressor. Incidentally, a brighter-than-life sound can also be created by using an aural enhancer or exciter after the compressor.
*MUSIC* The most Powerful Language in the world! *INDEED*
-
Nestor
- Posts: 6676
- Joined: Tue Mar 27, 2001 4:00 pm
- Location: Fourth Dimension Paradise, Cloud Nine!
BASS GUITAR
Again you can use either a ratio or soft knee-type of compressor setup much as you would for the acoustic guitar (to emphasise the attack of each note). Here however, the release time must be adjusted to match the individual’s playing style. Slap and pull bass playing may need a lot of compression to keep the level even and some of the notes could be quite short, so it will be necessary to set the release time as fast as you can without causing the sound to “pump”. With very fast release times, you’ll also need to set the hold time to 50ms or so in order to prevent distortion of the low frequencies. But this of course assumes that this type of distortion isn’t regarded as artistically viable – again, let your ears make the final decision – you might find a little low frequency distortion suits the sound.
It may also be necessary to increase the compression ratio to 5:1 or more to keep tight control of the louder notes, and as previously stated, a compressor with auto attack and release setting may be better if the playing style varies greatly during the course of the song. With slower playing or melodic fretless styles, the release time may be longer, say, up to half a second or more, and if the attack of the note doesn’t need to be emphasised, set a faster attack time so that the level is brought under control more quickly. If the playing style changes so much that there is no optimum attack and release time setting, try the auto mode.
Again you can use either a ratio or soft knee-type of compressor setup much as you would for the acoustic guitar (to emphasise the attack of each note). Here however, the release time must be adjusted to match the individual’s playing style. Slap and pull bass playing may need a lot of compression to keep the level even and some of the notes could be quite short, so it will be necessary to set the release time as fast as you can without causing the sound to “pump”. With very fast release times, you’ll also need to set the hold time to 50ms or so in order to prevent distortion of the low frequencies. But this of course assumes that this type of distortion isn’t regarded as artistically viable – again, let your ears make the final decision – you might find a little low frequency distortion suits the sound.
It may also be necessary to increase the compression ratio to 5:1 or more to keep tight control of the louder notes, and as previously stated, a compressor with auto attack and release setting may be better if the playing style varies greatly during the course of the song. With slower playing or melodic fretless styles, the release time may be longer, say, up to half a second or more, and if the attack of the note doesn’t need to be emphasised, set a faster attack time so that the level is brought under control more quickly. If the playing style changes so much that there is no optimum attack and release time setting, try the auto mode.
*MUSIC* The most Powerful Language in the world! *INDEED*