The AM Forum
June 13, 2024, 03:12:07 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
Author Topic: Processing 101 - Mike W2ZE  (Read 12631 times)
0 Members and 1 Guest are viewing this topic.
Radio Syracuse
Offline Offline

Posts: 1149

Syracuse Radio W2INR

« on: June 04, 2005, 01:43:06 PM »

Processing 101
A discussion and definition of commonly used term

By Mike Raide, W2ZE

  The purpose of this tutorial is not to recommend or condemn anyone particular unit(s), or to say it has to be done this way or it’s wrong.  I have found over the years, that opinions are like elbows, everyone has one, and certainly, everyone in the AM community has different opinions about processing. The purpose of this discussion is to define general terms and examine how they can help everyone achieve the “signature sound” that they want to achieve from their station.
Oh where, oh where to start!

Well, lets start with the unit for measuring power and voltage ratios in audio and RF. The Bel is named after Alexander Graham Bell, who not only invented the telephonium, but did a lot of work with sound and how humans perceive sound. The decibel is 1/10 of a Bel, and is the most commonly used unit. The decibel is calculated using the following formula:


So, we have an amplifier that takes a 1w signal and makes it a 20 watt signal out, so we calculate it by using the above formula:

10Log x (20/1) = 13.01dB

The amp has a power gain of 13 dB.

When power is doubled, the gain is 3 dB, but when voltage is doubled the gain is 6 dB. When power is reduced in half, the loss is 3dB, but when voltage is reduced in half, the loss is 6dB. Voltage gain is found using the following formula:


What starts to get a little confusing, it when you open a manual and look at the spec.’s of a particular piece of gear and it measures it’s input level or output level as say dBm or dBv or even dBu.

What that tells us is what reference the manufacturer used to find its output ratings. Many pieces of professional gear use some sort of metering to display levels, and are measured with VU (volume unit) meters, and are marked with decibels on the meter (1).

Volume units are referenced to 1mW of power produced by a 1KHz tone across a terminated 600 ohm load (2). Using Ohm’s law, this gives us 0.775V of audio, and references 0 dB on the meters, and the dBm is used to reference all power levels above and below this level.

The dBu is a measurement made unterminated or measured in an open circuit, and dBv is decibel-volts measured using an open circuit as well, but using 1V as its reference. Since these circuits are unterminated, or not working into a load of some sort, these are measurements of voltage only, not power. When setting up your gear, it is important that every piece of equipment be calibrated to some sort of standard level, usually using one of these reference levels.

  Whew! That was a mouthful. Now lets get into some terms that you may have heard in passing or in a QSO on the air, and find out how they help or hurt us.

Dynamic Range

Dynamic range is the basis of what we all are trying to accomplish when we start processing our audio .The definition of dynamic range is defined simply in two ways:

1 – To describe the actual range of signal fluctuations that are going through the equipment and,

2 – To define the maximum allowable range of signal fluctuations that can be put through the equipment (3).

Basically, the dynamic range is the quietest sounds all the way to the loudest sounds going through the equipment. Let's, for example, say the noise floor (noise the piece of equipment makes in operation) is –100 dB. The highest level our equipment sees coming in is measured at +3 dB. Our dynamic range is 103 dB.

This now gets us to peak-to-average ratio. Peak levels are the highest level of the audio at any given time. The average levels are where the audio stays around most of the time. A station that has high peak and low average audio has a lot of dynamic range, which has a more “open” or natural sound. A station that reduces the peak to average ratio of the audio increases loudness. Density is where the audio peaks are kept uniform throughout, but reduces the dynamic range significantly (4).

So, what can accomplish these things we just talked about? Well, here we go…


 Compression as defined by Bob Orban:

” Compression reduces the dynamic range of program material by reducing the gain of material whose average or root mean square (RMS) level exceeds the threshold of compression. The amount of by which the gain is reduced is called the gain reduction”(5).

Simply said, a compressor reduces the dynamic range to a reasonable level so the average audio is increased. The compressor threshold is an adjustment on a compressor that sets the level where you want the system to start compressing. Once audio starts to exceed the setting, the compressor starts to work. Any audio coming in below the threshold is not affected.

Once the compressor kicks in, the amount of gain reduction is adjusted in terms of compression ratio. The ratio is the amount of gain coming in versus the amount of gain coming out. Say a ratio of 5:1 is used. That means for every 5dB of audio gain coming in, only 1 dB of gain comes out. So if you have a threshold of say –30 dB set, and you now have an audio level that’s at –20dB, with a compression ratio set at 5:1, the output now at the compressor is –28dB.The rate at which the compression starts is called the knee. Some compressors have a hard knee, where they start acting right at the threshold. There are compressors with a soft knee, which start acting at or around 6 dB or so and gradually bring the signal into compression.

The attack time tells the compressor how fast to act once the threshold is crossed, and the release time tells the compressor to let go of the compression once the signal falls below the threshold. Faster attack and release times also reduce the dynamic range.

 Limiting is simply a compressor that has a very high compression ratio (in the order of 20:1), very fast attack time (in uSec.) and a very fast release time (less than 200uSec)(6).

Limiting is used to keep the audio waveform envelope down to a certain level as to prevent over modulation. Limiting also increases density, which can make the audio seem loud, but can be more fatiguing to the listener. In order for a limiter to be truly effective, they are usually set at a very high threshold, otherwise the gain reduction could be too high. Usually, they are designed to “stay out of the way” until needed. Often they are set to reduce no more than 6dB, but that is a really a matter of personal choice.

Clipping clips or chops off audio peaks at a set level. Clipping is mainly used to chop off very fast transient peaks. Clipping should be used with moderation, as too much clipping will distort the audio waveform, and cause perceived audio distortion. There are two types of limiting/clipping: symmetrical and asymmetrical. Symmetrical limiting limits both the positive and negative peaks of the audio waveform, while asymmetrical limits usually the negative side of the waveform, allowing the positive peaks to shoot upward.


  Expanders are just what their name implies. They expand or increase the dynamic range of a processor. They reduce the amount of gain set by the expansion threshold. Basically it increases dynamic range by not amplifying the noise and low level audio that would probably be increased or exaggerated by the compressor.

Gate or Gating

  Gate or gating is what expanders are to compressors. There are two types of gating: noise gate and compressor or limiter gate (7). The limiter gate stops any change in low noise level during long stops or pauses. It holds the compressor gain when the level drops below the gate Threshold. A noise gate reduces noise by reducing the gain when it falls below its threshold.

AGC Levelers

  AGC levelers are devices that do nothing to decrease or increase dynamic range, All its function in life is to make sure that compressor/limiters receive constant input levels. They usually have very slow attack and release times (sec.). If a level goes above its threshold, it gently pulls it down to a set level. At low gain times it gently pulls the level up, and keeps everything on an even keel so to speak.

  Last but not least is the subject of equalization. This probably where most people get into heated debates on and off the air. It is purely subjective to each individual that it would be impossible for me to even suggest how to set one up. All I can do is just describe the differences between the two different types, and let the end user take care of the rest.

   Like I said earlier, there are two types of equalization: Parametric and graphic equalizers. Eq’ing basically changes the shape of the audio signal applied to it. It can cut or boost certain audio frequencies applied to it. There are two types of curves associated with equalizers: Shelving and peaking curves. It is best described by Bob Orban as:

“Shelving starts at a certain gain, as the frequency changes, the gain increases (boost) or decreases (cut)"(Cool.

"A peaking curve is bell-shaped on the frequency axis. Unlike the shelving curve, it has a well defined peak frequency. The shape of the curve can be uniquely defined by three parameters: the amount of equalization (in dB), the frequency of maximum equalization (Hz), and the Q which is a dimensionless number that describes whether the curve is broad or sharp." (9)

 A parametric EQ provides several peaking curves, and are usually three or four band EQ’s. They can be used to notch out a certain frequency by dropping the range of equalization low and making the “Q” or bandwidth lower. They can also be used to expand a certain band of frequencies. Generally a parametric EQ gives the operator the most flexibility, but it can be very hard to use.

The graphic equalizer is the more familiar one to most people. It divides the frequency range into usually 8 up to 31 bands in octave’s or fractions of octaves. It provides a graphical interface on the front to show what frequencies are being cut and boosted. It is easier to use, but not very flexible as far as expanding the range of boost and the bandwidth of a set of frequencies.

In Closing

 Well, I’ve done my best to describe everything I can about the basics of processing. Hopefully, it’s not as confusing as it seems. Usually equipment manuals provide very good detail into processing and how to best utilize the equipment. Most of the information in this paper comes from the NAB engineering handbook. The handbook is a very good and detailed source, unfortunately, it’s a very expensive book, and not many have access to it.

This article should give everyone a good starting point. Most processing is very subjective to the individual, and real world "trial and error" is necessary. With the basic information provided, the user won’t be forced to start “taking stabs in the dark”, or feel intimidated not to try it out. Also, one has to take into account their personal budget. Processing can be range from around 100 dollars for the basic Mic compressor/eq to thousands of dollars for multi band peak limiters and AGC levelers. With the endless list of equipment in varying price levels, the combinations are limitless. Most important, go out and have fun!


Mike, W2ZE


1.         Nab Handbook 9th edition; page 342
2.         Nab Handbook 9th edition; page 342
3.         Symetrix Inc. 528e manual revision 1.4, page 2-1
4.         Nab Handbook 9th edition; page 401
5.         Nab Handbook 9th edition; page 397
6.         Nab Handbook 9th edition; page 398
7.         Symetrix Inc. 528e manual revision 1.4; page 2-3
8.         Nab Handbook 9th edition; page 399
9.         Nab Handbook 9th edition; page 399

G - The INR

Amateur Weather Station KNYSYRAC64
Creator - owner - - 2001 - 2010
Founding Member - NEAR-Fest
Pages: [1]   Go Up
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.05 seconds with 19 queries.