Intro To Signal Processing

ADVANCED RECORDING PRIMER

CHAPTER 3 - PART 1, BY ROBERT DENNIS

SIGNAL PROCESSING

HISTORY AND USES:

Signal Processing is the process of using devices (or programs) that alter the level or time of an original signal. The history of signal processing is that it came into being to make up for a lack in the recording process where the "real" sound of the instrument or performance could not be captured properly. The aim was to "restore" recorded sound to something closer to the real sound that was heard by the ear.

In modern "pop" music the use of signal processing usually extends to making the music with monosodium glutamate added to the sound. The instruments become "hyped, "more-there, "bigger, "louder, "more-striking, " than really possible.

Soon after engineers began over-hyping the sound, excessive amounts or unusual uses of signal processing were applied to "create a different sound" one that could never be considered related to natural sound put out by instruments.

TYPES OF SIGNAL PROCESSING:

Signal Processing can be broken down into three general types:

PROCESSING TYPE	DESCRIPTION	USED FOR
DYNAMICS	Automatic gain changes under certain circumstances under the categories of Compression, Limiting, Expansion & Gating	Raising apparent volume, reducing noise and leakage, preventing overload, modifying attack or sustain, reducing sibilance or pops, keeping a consistent level, reducing the need for other processing.
EQUALIZATION	Changing the gain of specific bands of frequencies under the categories of filtering. peak EQ, shelf EQ, and notch EQ.	Reducing masking, reducing noise and leakage, modifying attack, reducing sibilance or pops, keeping a consistent level, making image more present or distant.
DELAY	Changing the timing of signals and adding repeats under the categories of delay, reverberation, flanging, doubling, echoes, chorusing.	Making image more present or distant, modifying attack and sustain, making a fuller sound, "unreal" processing, stereo simulation, increasing body.

SIGNAL PROCESSING FOR ANALOG & DIGITAL RECORDING

Equalization Processing For Analog and Digital Recording:

Equalization processing is increasing or decreasing the level of the audio at certain sets of frequencies and not affecting the audio level of other frequencies for the purpose of changing the sound of the instrument being miced or picked up. The board equalizer allows the engineer to make these changes.

Analog recording techniques call for much of the needed equalization processing to be applied before the signal is recorded. Program equalization that involves boosting the mid-range or high-frequency energy is done before recording for two major reasons. The biggest reason is that equalization done after the multitrack recording tends to bring up the noise with the signal. The old pros will quickly tell you that if you want more highs, to boost them before recording. Another major reason for high-frequency boost during analog recording is the "blurring" of the high frequency energy and transients in analog recording. If you were recording high-hat with analog tape, it would come back somewhat duller sounding upon playback. Recording engineers often boost high frequency energy at 15 kHz so that when the high-hat is played back, it has as much of the sizzle sound as it originally did. The adding of the equalization during recording makes the high hat less real sounding because you are adding more phase distortion with the board equalizer, but at least it has the same amount of sizzle which is also a part of making the instrument sound as real as it can.

Digital recording, with its very low noise and lack of phase-distortion does not need to have program equalization applied to overcome the limitations of the recording system. There is no real benefit to adding program equalization as the tracks are being cut. The board equalizer will cause phase-distortion. Since there is little or no phase distortion in digital recording the phase distortion of program equalization becomes obvious quickly. The general rule for digital recording therefore becomes to add little or no equalization during recording.

Another point is that one of the biggest reasons to use program equalization is the fact that there is a hearing limitation of masking. When two sound sources put out energy close to the same frequency, the louder signal will tend to make the softer signal disappear. Because of this instruments may sound "dull" or "hard to hear" when they are played with the other instruments of the production. When you are recording the basic session, all of the overdubbed instruments are not there yet and the masking that they will cause is hard to judge. For this reason, program equalization is best applied during mixdown; program equalization applied during recording may not overcome the masking in mixdown and you would have done it differently if you had heard all of the instruments.

Use Of Compression in Analog and Digital Recording:

A Compressor acts to reduce gain of the highest part of the audio. This action makes the loud passages not go as high as they did originally. On a vocal, this is usually set so that the compressor reduces gain by about 6 dB on the loud passages. When the compressor is used before recording, the gain has to be readjusted so that the loudest passages still cause the meter to reach "0." By making this adjustment, the recording engineer is making the low-level passages go to the recorder 6 dB louder.

When compression is used in this manner while recording vocals, the noise of the track is half as loud as compared to the signal. For this reason, many recording engineers will not record a lead vocal onto analog tape without compression. One of the biggest reasons to use compression is to overcome the analog tape noise of the recording system.

Compression is used for other purposes also. The compressor makes the average level louder and gives more apparent volume to the recording compared to a recording made without compression. This attribute of the sound change is especially important to the very loud styles of music (Heavy Metal and Hard Rock). Compression for these styles of music may have a gain reduction of 10-12 dB. Compression done during the mixdown of an analog recording will bring up the low level tape noise as well as even out the volume changes of the vocal. This is another reason that compression is generally done as the recording takes place.

The low-level noise in digital recording means that compression need not be done to overcome the limitations of the recording system. If compression is done during the mix, the recorded noise will not be noticeably increased. Compression done to change the average level of the vocal or to make it "more intense" can be better judged during the mixdown. Because of these factors, compression is used to much less of a degree during digital recording.

CHOOSING THE TYPE OF PROCESSING:

By looking at the above chart, you can easily see that accomplishing a certain task can often be done with more than one type of processing. Which processing you choose (or which combination you choose) is what can make you perform like a professional engineer or amateur recordist. For example, you may ask yourself which processing is best to modify the attack and sustain of an instrument. The example below discusses the pros and cons of using each of the processing types to get more "presence" on a vocal:

NON PROCESSING SOLUTIONS

Not everything you do needs different signal processing or effects. In the overall picture it is often best to adopta rule that "less processing is best."

Bring Up Fader

Why It Works	It brings up the vocal.
Advantages	Doesn't change vocal quality.
Disadvantages	Sometimes bringing up the vocal can cause other instruments to be harder to hear (because a frequency element in the voce now covers up another instrument part). Bringing the vocal up can sometimes reduce the "excitement" of the mix because the instruments are now lower than they were (relative to the voice). The vocal can sound too loud in spots.

Ride The Fader

Why It Works	Brings up vocal lines that were hard to hear.
Advantages	Doesn't change vocal quality.
Disadvantages	If it is done incorrectly, it can make matters worse.

DYNAMICS PROCESSING SOLUTIONS

The Compression & Limiting functions on your dynamics processing will turn down the louder passages, automatically. Now the vocal will not peak as high and it is usually brought up to be the same peak level as it was before using compression or limiting. This effectively increases the "average level" making the low-level passages louder.

Compress The Vocal

How It Works	It effectively brings up the passages that are being covered up, directly helping each line being heard.
Advantages	It does not "color" or change the vocal quality.
Disadvantages	The voice can, again, start to cover up instruments that are low in the mix. Using too much compression tends to "take the life out" of the vocal and in extreme cases causes a "pumping" (where you hear noise and leakage come up on low-level passages) or a "restricted" sound on the loud passages.

Compress The Mix

Compressing the overall mix tends to make everything heard better. Sometimes it does one good to process the overall mix with compression rather than the voice. The disadvantages for overall compression are much the same as the disadvantages of compressing the vocal only.

Limit The Mix

Limiting is an extreme form of compression, allowing no signal to get above a preset level (called the Threshold). Running your limiter with the quickest attack time possible and 5 ms. or so release time can reduce those quick spikes that cause the overall level to be low. Set the "threshold control" so these peaks are reduced about 6 dB. Now the whole mix can come up in level. Anytime you increase the overall level, the mix will sound better. Maybe the vocal really doesn't need adjustment.

EQUALIZATION SOLUTIONS

Equalization changes the level of the signal for a particular band of frequencies. Applying equalization to a vocal will tend to accent certain harmonic frequencies over others.

Boost Presence or Projection Frequencies

How It Works	The presence frequencies (around 5 kHz) and projection frequencies (around 3 kHz) will make the vocal cut though the music track.
Advantages	This often allows the vocal to be mixed near the level of the track, creating "excitement"
Disadvantages	The hyping of these frequencies with EQ makes the voice less natural. If overdone, the vocal becomes harsh and irritating and can "thin-out.".

Boost Presence on Entire Mix

Even after the mix is complete, boosting 3-5 kHz can bring out the vocal in the mix, along with the possible disadvantages listed above.

DELAY SOLUTIONS

The use of delay and reverb effects to "bring out" things in the mix is the most neglected technique in today's mixing.

Reduce Reverb While Adding Delay

How It Works	Reverberation makes things sound more distant. Reducing reverberation on the vocal will put the vocal more "up front." Unfortunately this usually causes the vocal to "sound bare." Adding a Slap or Space Echo to the vocal helps give the character of more reverberation without sending the vocal further back in the mix.
Advantages	This solution does not change vocal quality.
Disadvantages	The slap or space echo tends to make the recording sound somewhat like it was made in an outdoor arena, especially if overdone. The vocal can also become harder to understand.

Customize Effects Parameters For Vocal

How It Works	Pre-delay on the reverberation makes the vocal sound more present. Certain programs, such as Hall programs tend to have a high "pre" or "initial" delay. Other programs, such as "Plate" and "Room" tend to have shorter pre-delay parameters. Using the correct kind of reverb program, (Plate or Room on instruments and Hall on vocals) or adjustment of parameters on the reverb will help bring the vocal forward in the mix.
Advantages	This works without changing the balance of the mix.
Disadvantages	If the programs are too dissimilar, it begins to sound like the vocal and the band are in different rooms. To help with this, adjust the reverb time parameter of both units to match or at least be close. Another disadvantage is that it takes an additional effects unit for the vocal sound.

One of the ways to get a really good mix is to use all of your signal processing tools on the different instruments that make up the production. There are many other settings that can be used with EQ/Delay/Dynamics to achieve the best mix, one of which is the "Exciting Compressor" and others which you will find in the RID Pro Audio Specialist and EQ & Mixing study modules in the Alexander Magazine online training service.