Limiting vs Compression

[Ewizard]

Limiting is a bit misunderstood by most. By definition, it limits the peak output by compressing the media as much as is needed to control the dynamics to not go beyond the threshold.

A compressor works similarly except that the threshold is the level at which the compressor engages. It is a ratio-based system, so a 1:1 ratio means that for every db over the threshold, the compressor will reduce the level by 1db. A 4:1 ratio means that for every db over the threshold the signal goes, it will reduce output by 4db. The input and the output are not linked, so you can in theory compress more level down than what goes into the compressor.

Both are anti dynamics tools. In short, they both reduce the dynamic range of the media. This can have a downside. They only control the peak or RMS level of the media, so as you raise the level of the media, the noise floor comes up with it, while it reduces the peak level of that media. What you ultimately hear is a signal that sits in a particular volume range and the noise floor sets with it. I.E. quiet becomes louder and louder is just not much louder than the quiet.

When you hear a distorted guitar, it is very compressed; quite literally. Since distortion is the chopping off of the top of the guitar's signal, it is pretty much a limited signal. So to say a distorted guitar has dynamics is very much an oxymoronic statement. What you actually notice is that the range between fully broken up and slightly dirty is wide, but the level that you hear is fairly neutral or even. Not dynamic in actuality, but the distortion you hear goes through a state of compression and then into limiting.

Using a compressor before the amp can do all kinds of fun things. Depending on how it's set up you can get snappy sounds, to muted sustaining sounds. As an effect, it is very useful and can more or less be used as a boost, all the way to making your notes bloom.

 

[Chistopher]

When you hear a distorted guitar, it is very compressed; quite literally. Since distortion is the chopping off of the top of the guitar's signal, it is pretty much a limited signal. So to say a distorted guitar has dynamics is very much an oxymoronic statement. What you actually notice is that the range between fully broken up and slightly dirty is wide, but the level that you hear is fairly neutral or even. Not dynamic in actuality, but the distortion you hear goes through a state of compression and then into limiting.

Interestingly enough, distortion does not compress, it's just a really aggressive limiter. Distortion occurs when a signal has more amplitude than the system is able to reproduce, flattening the signal. It sounds like it is compression because guitar players use so much distortion that often times the only frequency that doesn't clip is the noise floor, which at that point is usually a lot closer to level of the rest of the signal than it usually is. So when you go play slightly dirty, some of the frequencies are close to the volume where they clip, but if you dig in, you increase the level to the point that more of those frequencies are at clipping level.

 

[Mincer]

Limiting is a bit misunderstood by most. By definition, it limits the peak output by compressing the media as much as is needed to control the dynamics to not go beyond the threshold.

A compressor works similarly except that the threshold is the level at which the compressor engages. It is a ratio-based system, so a 1:1 ratio means that for every db over the threshold, the compressor will reduce the level by 1db. A 4:1 ratio means that for every db over the threshold the signal goes, it will reduce output by 4db. The input and the output are not linked, so you can in theory compress more level down than what goes into the compressor.

Both are anti dynamics tools. In short, they both reduce the dynamic range of the media. This can have a downside. They only control the peak or RMS level of the media, so as you raise the level of the media, the noise floor comes up with it, while it reduces the peak level of that media. What you ultimately hear is a signal that sits in a particular volume range and the noise floor sets with it. I.E. quiet becomes louder and louder is just not much louder than the quiet.

When you hear a distorted guitar, it is very compressed; quite literally. Since distortion is the chopping off of the top of the guitar's signal, it is pretty much a limited signal. So to say a distorted guitar has dynamics is very much an oxymoronic statement. What you actually notice is that the range between fully broken up and slightly dirty is wide, but the level that you hear is fairly neutral or even. Not dynamic in actuality, but the distortion you hear goes through a state of compression and then into limiting.

Using a compressor before the amp can do all kinds of fun things. Depending on how it's set up you can get snappy sounds, to muted sustaining sounds. As an effect, it is very useful and can more or less be used as a boost, all the way to making your notes bloom.

This description is exactly how I always understood it.

 

[BlackhawkRise]

Limiting is a bit misunderstood by most. By definition, it limits the peak output by compressing the media as much as is needed to control the dynamics to not go beyond the threshold.

A compressor works similarly except that the threshold is the level at which the compressor engages. It is a ratio-based system, so a 1:1 ratio means that for every db over the threshold, the compressor will reduce the level by 1db. A 4:1 ratio means that for every db over the threshold the signal goes, it will reduce output by 4db. The input and the output are not linked, so you can in theory compress more level down than what goes into the compressor.

Both are anti dynamics tools. In short, they both reduce the dynamic range of the media. This can have a downside. They only control the peak or RMS level of the media, so as you raise the level of the media, the noise floor comes up with it, while it reduces the peak level of that media. What you ultimately hear is a signal that sits in a particular volume range and the noise floor sets with it. I.E. quiet becomes louder and louder is just not much louder than the quiet.

When you hear a distorted guitar, it is very compressed; quite literally. Since distortion is the chopping off of the top of the guitar's signal, it is pretty much a limited signal. So to say a distorted guitar has dynamics is very much an oxymoronic statement. What you actually notice is that the range between fully broken up and slightly dirty is wide, but the level that you hear is fairly neutral or even. Not dynamic in actuality, but the distortion you hear goes through a state of compression and then into limiting.

Using a compressor before the amp can do all kinds of fun things. Depending on how it's set up you can get snappy sounds, to muted sustaining sounds. As an effect, it is very useful and can more or less be used as a boost, all the way to making your notes bloom.

I see a few inaccuracies here. I have interned with a group as part of signal processing for 2 years. Audio signals are not my primary focus though, so keep that in mind. Also, a few of the words you are using are wrong, but I understand what you're saying and there is no need to nitpick.

The compression ratio doesn't work quite how you mention. a 4:1 ratio means that if the signal is 4 db over the threshold value, it will output a 1 db signal. The system you appear to be describing would invert the volumes of the system, making the quiet parts louder than the loud parts. They only control the peak value. A real-time compression circuit could not practically control the RMS level of the signal in any meaningful way as that would require calculating an average value across the entirety of the signal, ie the compressor would need to see your volume for the entire recorded song.

Correct me if I misunderstood your explanations, but I believe you also misunderstand compression. It sounds to me like you think a compressor is just a more subtle limiter. A limiter a limiter works as you described, if you set the limit the 50 dB at a 4:1 ratio and hit it with a signal, the dynamic range of the parts above 100 dB will roughly a 1/4 of what it was, while the parts under 50 dB will be left untouched. Using a compressor you have two limits, say one at 50 and one at 25 dB; for this example lets say they use the same ratio. The 25 dB threshold works the same but instead of cutting volume to decrease dynamic range, it will increase the volume. Because compressors have a tendency to boost noise, a well made compressor will generally employ filters to help combat this.

Here's what the knobs do on most compressors:

Attack - Adjusts the ratio for the volume ceiling, to high a ratio and it squashes picking dynamics, too low a value and there is no limit on max volume
Sustain - Adjust the ratio for the volume floor, to high adds noise, reduces picking subtlety, and makes mistakes obvious
Compression - Does different things on different pedals, but if you have an attack and sustain knob, this brings the threshold values closer or further together. Lower on the knob would make the effect unnoticeable, higher on the knob makes every single element of the signal the same volume
Level - Adjusts total volume, has no effect on tone or compression.

 

[Ewizard]

I worded the activity of compression wrong, but the point was that it is a ratio-based effect. I did grossly oversimplify the outcome. The actual result would be that if you have a 4:1 ratio set, then with a signal +4db above the threshold, you will only have +1db of output. I.E. +4db goes in and only +1db goes out. The way I described it, +1 goes in and only -3 goes out ( a net negative gain ). You can get a net negative gain either way, but I did say it wrong.

The attack is quite literally how fast the compressor will react to the incoming signal. Using the volume ceiling isn't a good way to explain it. A fast setting means that every transient that goes above the threshold triggers the compressor. With a long attack, the effect of the compressor is delayed X time, so long as the transient is still above the threshold. I.E. A long attack will wait X amount of time to engage, provided the incoming signal is still above the threshold during that hold time. This can allow kick, snare, or guitar picking to keep the original transient before the compressor clamps down on the signal. It is also a great tool to keep vocal consonants intelligible, by allowing them to slip through before the body of the vocal note is clamped down on.

Sustain functions in many guitar pedals is a mixed bag. Some have makeup gain, some allow you to add that in on your own. The all-in-one " sustain " knob usually adds makeup gain as part of the effect. The higher you set it, the more compression there is and the less dynamic range the incoming signal will have. Quiet is loud and loud is the same volume as what was quiet.

I would argue that with tube guitar amps, that distortion is a compression to limiting type of result. As the signal begins to distort, there is an amount of sag that does make the note bloom or squish if you will. And since the tube is also beginning to operate non-linearly around the threshold of distortion, there is an amount of deviation in the signal that can be seen as compression on a scope. The signal will be clipped, but not in the same manner as a solid-state amp literally chops the top of the signal off. It will compress a bit before it goes into full-on clipping and what we classify as limiting. limiting in a guitar amp is quite literally the tube shutting off grid conduction because of the bias of that tube is falling apart ( power is not there to support the incoming signal ). Full-on limiting in a tube amp is called blocking distortion. The incoming signal is so strong the tube simply shuts off during the time the signal is outside of the bias range it is set in. Grid stoppers are used to actually control this problem. A grid stopper slows the incoming current down ( reduces the transient ) so that the tube doesn't go into sudden cutoff, in effect acting as a compressor.

 

[Chistopher]

I've been on this earth for multiple years and I still don't quite know what the knobs on most compressors do. Sometimes attack is ratio, sometimes clamp is ratio, usually attack is a time delay, sometimes it's a steeper ratio that only affects the first X seconds above the threshold, sometimes it's the speed at which the compressor kicks in. I have no clue what blend or recovery does on fenders new compressor. Sometimes sensitivity means the same things as comp, sometimes it doesn't