Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[rnicklaus]

What is being said here is that if one is to put a signal into an LA2A and the gain control is not touched, as the threshold is lowered and attenuation is seen and heard, that gain is being added in between threshold, ratio and output?

[zmix]

rnicklaus wrote on Thu, 27 April 2006 12:47

What is being said here is that if one is to put a signal into an LA2A and the gain control is not touched, as the threshold is lowered and attenuation is seen and heard, that gain is being added in between threshold, ratio and output?

In order to fully understand your question, I think it is important for you to see how the LA-2 operates, and where the gain exists in the circuit.

The LA-2 operates thusly:

There is a pot wired as an attenuator on the input, just after the T4 (gain reduction element), labled "gain" (on the front panel).

The signal is tapped before this "gain" control and fed to another pot wired as an attenuator which feeds the amplifier driving the EL panel of the T4 attenuator. This pot is labled "Reduction".

The RATIO is selected by a switch which controls how isolated the sidechain is from the GR element (thereby altering the amount of feedback from the GR element to the sidechain).

As is true with most compressor/ limiters, makeup gain is adjusted by attenuating the input to a fixed gain amplifier.
The UREI 1176 has over 45dB of makeup gain.
The LA-2 has -I believe- 30dB of gain in the amplifier circuit.

If no attenuation was applied to the LA-2 you would have enough signal to roast a turkey.

[rnicklaus]

But if full attenuation is applied and no gain, what happens at that point, to the turkey?

Is gain added?

I get how the 1176 works with a fixed threshold and input gain needs to be added to get more gain reduction (and attack and release)

But when one gets into the LA2A, DBX 160 series, LA3, etc, is gain being added when GR occurs?

[zmix]

rnicklaus wrote on Thu, 27 April 2006 16:14

But if full attenuation is applied and no gain, what happens at that point, to the turkey? Is gain added? I get how the 1176 works with a fixed threshold and input gain needs to be added to get more gain reduction (and attack and release) But when one gets into the LA2A, DBX 160 series, LA3, etc, is gain being added when GR occurs?

What exactly are you asking?

The LA-2, LA-3 and 1176 all have input attenuators and gain reduction devices followed by fixed gain amplifiers, so if your question is simply "is gain added" the answer is yes for all the units you mentioned.

The output amplifier in the  LA-2 adds 30dB, the LA-3 adds 30dB or 50dB depending on the gain switch and the 1176 adds 45dB.

The dbx 160 also has a fixed gain amplifier and uses a control voltage to the VCA to adjust the output level.

[rnicklaus]

I am asking that when you take the LA2A, LA3, DBX 160, set the ratio let's say to 8:1 - output gain at unity - send a unity gain (0 DB) signal to the compressor - drop the threshold to - 9, is there any make up gain happening inside the compressor?  

[zmix]

rnicklaus wrote on Thu, 27 April 2006 18:34

I am asking that when you take the LA2A, LA3, DBX 160, set the ratio let's say to 8:1 - output gain at unity - send a unity gain (0 DB) signal to the compressor - drop the threshold to - 9, is there any make up gain happening inside the compressor?

A signal itself cannot be 'unity gain'. That term defines a transfer  function from input to output of a system. In the post above I detailed exactly how much gain is 'happening' in those units. I don't understand what you want to know exactly, based on your unclear wording, but I'll see if I can answer your question.

Only one of the compressors mentioned has any of the actual adjustments you specify, so I'll use that one as an example.

The dbX 160 has a V/I converter, followed by a VCA and an I/V converter and an output amplifier. The dbx 202 VCA is only capable of attenuation from unity gain so the output amplifier provides around 20dB of  fixed gain.

If the output level control on the dbX 160 is set to 0dB the VCA is actually attenuating the signal by 20dB and the output amplifier is restoring the level to 0. This could be defined as 'makeup gain' but the concept of makeup gain is systemic so by definition we disregard this gain.

If you set the threshold to -9dB and send the 0db signal to the unit at 8:1 the output will be reduced. If you switch the meter to read "output" you can see this. There is an adjustment, also called 'output',  and it is available specifically to compensate for this loss. In the dbX 160 the output control works to adjust the overall attenuation of the VCA by adding an offset to the CV generated by the detector. Adjusting the level so that the system is at unity gain to compensate for this loss is adding what is called 'makeup gain".  

[rnicklaus]

It very well may be that I don't understand this thread, but I don't think my question is unclear at all.

This is said with full respect for your explanations and knowledge.

If I can take a signal in and out of a dbx 160 and apply no output gain, set it to 10:1 turn the threshold to minimum, no real signal will be at the output of the unit.

There does not, to my logical brain, seem to be any make up gain going on at this point.

[zmix]

If you plug a Les Paul into a 100 Watt Marshall and turn the amp all the way up and the guitar all the way down you won't get any "real" signal out of it either.

Obviously not a 100 watt amp.

[rnicklaus]

zmix wrote on Thu, 27 April 2006 17:06

If you plug a Les Paul into a 100 Watt Marshall and turn the amp all the way up and the guitar all the way down you won't get any "real" signal out of it either. Obviously not a 100 watt amp.

That has nothing to do with my question or your graphs, but I am done anyway, so no matter.

Of course I meant no added output gain.

[Ronny]

zmix wrote on Thu, 27 April 2006 11:53

Ronny, Take a close look at your graph. Now imagine what happens to the curves you have added when you set the makeup gain so that unity gain (Vin=Vout) occurs at 0dB, voila these curves are exactly the SAME.... Is this really that confusing? I asked you to plot how you are actually USING your compression so that these concepts would be crystal clear to you. Would you really use a compressor as indicated on your additional traces - with 19dB of loss upon insertion?? Hell no, so please normalize your graph and try your argument again. -CZ

Chuck, bear with me please, I'm not clear on how to plot a tranfer function on one of my comps. I have over 50 different compressors and every one of the manuals that shows compression graphs has the same type graph that I displayed. The data represents exactly what is happening at the compression stage. On Morgan's graph the signals are not at unity gain. The signal enters the compressor from the bottom in gain. Unity in my explaination means "input gain" All compressors have input gain at unity. If the threshold is not reached it passes the same input gain through output "at unity". I understand that some comps restructure the gain as in the LA-2A but before therehold is met, input and output levels are the same. The signal "has to enter the compressor where the X and Y lines read the same decibel levels" as it's before threshold and before the signal is processed, thus input and output levels must be the same. What I'm having a problem with is Morgan's two lines show input at -40dB on the 2:1 line and input near -30dB "output" on the 20"1 line, the input gain at compressor input in reality is the same as the output gain "until the compression threshold level is reached" from the bottom left. So no this is not crystal clear to me. I think I understand what you are saying how the rotation point is showing makeup gain, but looking at it this way is new to me because I just can't understand how an input signal can be input at -40dB or -33dB output when the graph shows unity gain at -50dB. My blue lines represent every compression graph that I've seen up until you showed Morgans. Are you telling me that you don't understand what I'm relating? I admit that I don't understand fully what you are relating as the graph's just don't represent real gain going into a compressor. Perhaps they must do this to show make-up gain, but that's still not clear to me.

There is no -19dB loss at insertion on my graph, signal enters at unity gain -50dB input -50dB output, when threshold is met at -20dB, the 20:1 ratio "attenuates" the signal by -19dB for every 20dB that exceeds the threshold level. Output gain structure, not showing on my graph can be set at any level to make-up for the attenuation that is happening after threshold of the compression stage. On your graph the 20:1 line shows input at -50dB and output at -33dB, "before the threshold level is met and before the compression stage is activated" and threhold applied at -3dB output, this is why it's hard for me to understand this. I'm not clear on how you are seeing a gain rise after threshold is met, because even on Morgans graphs, the lines go to the right of the input gain, signifying what I see as gain reduction "from input", same as my graph if he moved his two lines down to unity gain at the mid center 45 degree line that represents unity gain on his and all graphs that I'm familiar with and the line that divides the X-Y access, so that "output level of the compression stage" shows on right side of the mid center diagonal line as gain reduction. Compressors reduce gain when threshold is met if ratio is more than 1:1, therefore the reduction line will always be below input (unity) gain level to the right side of the diagonal mid center line when ratio is applied. The amount of ratio applied determines the level the signal will be when it reaches the ouput gain structure. In this case the 2:1 ratio is reducing the signal by -10dB when the signal leaves the compression section and goes to the output gain structure.    

[zmix]

Ronny wrote on Thu, 27 April 2006 06:06

Chuck, bear with me please, I'm not clear on how to plot a tranfer function on one of my comps.

Set up one of your compressors as you normally would use it. You MUST follow this step if any of this is going to make sense. Just get your favorite source spanked up with a compressor, popping it in and out of bypass till you like the setting, unpatch it and then then simply send a signal to the unit and  measure  and record the output signal as you increase the input level. You could use the posted graphs as a model, and run the input level from -50dB to +20 dB. Just put a dot on the graph at the intersection of each point, so you have  a line comprised of these measurement points, input along the Horizontal axis, and measured output along the vertical axis.

Quote:

I have over 50 different compressors and every one of the manuals that shows compression graphs has the same type graph that I displayed. The data represents exactly what is happening at the compression stage. On Morgan's graph the signals are not at unity gain.

Yes they are at unity gain, look at the reference level, which in this case is: "0dB" The curves all intersect at this point. Input Level (horizontal axis) =0dB, output Level (vertical axis) =0dB. This is the definition of unity gain.

Quote:

The signal enters the compressor from the bottom in gain. Unity in my explaination means "input gain" All compressors have input gain at unity.

Unity simply means that the signal level in =  signal level out. The word unity means "one", a gain of "1" is  called unity gain. Input signal multiplied by '1' produces the same level at the output.  

Quote:

If the threshold is not reached it passes the same input gain through output "at unity".

Not in most cases...(real world). Look at the graph again. at the 2:1 ratio the input signal of -50dB produces an output signal of -40dB. These low level signal are increased in level by 10dB. Further examination of that curve shows you that the 2:1 ratio at a -20dB threshold will require 10dB of gain to be added so that there is no insertion loss at the nominal operating level of "0dB"  

Quote:

I understand that some comps restructure the gain as in the LA-2A but before therehold is met, input and output levels are the same. The signal "has to enter the compressor where the X and Y lines read the same decibel levels" as it's before threshold and before the signal is processed, thus input and output levels must be the same. What I'm having a problem with is Morgan's two lines show input at -40dB on the 2:1 line and input near -30dB "output" on the 20"1 line, the input gain at compressor input in reality is the same as the output gain "until the compression threshold level is reached" from the bottom left. So no this is not crystal clear to me. I think I understand what you are saying how the rotation point is showing makeup gain, but looking at it this way is new to me because I just can't understand how an input signal can be input at -40dB or -33dB output when the graph shows unity gain at -50dB. My blue lines represent every compression graph that I've seen up until you showed Morgans. Are you telling me that you don't understand what I'm relating? I admit that I don't understand fully what you are relating as the graph's just don't represent real gain going into a compressor. Perhaps they must do this to show make-up gain, but that's still not clear to me. There is no -19dB loss at insertion on my graph, signal enters at unity gain -50dB input -50dB output, when threshold is met at -20dB, the 20:1 ratio "attenuates" the signal by -19dB for every 20dB that exceeds the threshold level.

Look at the graph you've posted: When the input signal reached 0dB, the output is only at -19dB.

Quote:

Output gain structure, not showing on my graph can be set at any level to make-up for the attenuation that is happening after threshold of the compression stage.

The graph is used to define the transfer function of the system. Input versus output. Makeup gain is a part of every compressor design, manual or automatic makes NO difference.  

Quote:

On your graph the 20:1 line shows input at -50dB and output at -33dB, "before the threshold level is met and before the compression stage is activated" and threhold applied at -3dB output, this is why it's hard for me to understand this. I'm not clear on how you are seeing a gain rise after threshold is met, because even on Morgans graphs, the lines go to the right of the input gain, signifying what I see as gain reduction "from input", same as my graph if he moved his two lines down to unity gain at the mid center 45 degree line that represents unity gain on his and all graphs that I'm familiar with and the line that divides the X-Y access, so that "output level of the compression stage" shows on right side of the mid center diagonal line as gain reduction. Compressors reduce gain when threshold is met if ratio is more than 1:1, therefore the reduction line will always be below input (unity) gain level to the right side of the diagonal mid center line when ratio is applied. The amount of ratio applied determines the level the signal will be when it reaches the ouput gain structure. In this case the 2:1 ratio is reducing the signal by -10dB when the signal leaves the compression section and goes to the output gain structure.

Ronny, just plot some of your favorite compressor settings and this will all be much more clear...