tough compressor ques.

  • Thread starter Thread starter Levey
  • Start date Start date
L

Levey

New member
How do you make an expander out of a compressor?

I know you CAN do it but I just dont know HOW to do it. No tampering with the circuitry allowed either. Ive been killing myself over this one.
 
It depends, not all compressors have that ability. If the one you have dosn'y have that abilty, yeah, youll be killing yourself for a long...long time.

SoMm
 
Some compressors have ratios that go less than 1:1, ie 0.5:1. Thats acting like an expander in those ratio settings.
 
yeah i forgot to mention it has to be one mono compressor with no other dynamic devices built in, expander etc. that would solve my problems wouldnt it, but this question DOES have an answer. just one mono compressor by itself with the usual controls: threshold, ratio, attack and release times, gain.

the compressor with ratios such as .5:1 would still be compression right. an odd compression, however it would look like an expander slope of input vs. output.
 
i dont know anyone who would want to do this including myself. its one of the many ballbusters assigned by my engineering professor. i think i give up
 
An expander is more functionally a noise gate rather than a compressor, but in terms of the process - it works like a compressor in reverse.

Compressors reduce the gain of a signal exceeding a given threshold, while expanders apply gain reduction to signals falling below a certain threshold to acheive a more subtle muting effect than you would get with a gate.

What an expander ISN'T though, is a way restore dynamics lost thru overcompression...

SoM is right though -- not all compressors will allow the feature.
 
Use the board..

A simple way to do it could be to split the input signal in two, compress one of them, and send to two channels in your mixer.

Then, using phase reverse on one of the channels, you subtract the compressed signal from the direct one.

At sub-threshold levels you could have any gain reduction down to completely cancelled, set by the relative gain of the two signals. When over-threshold - ie. actually compressing - you subtract relatively less, and thus get higher overall gain..

Or have I missed something?

Jakob Erland
 
Re: Use the board..

gyraf said:
A simple way to do it could be to split the input signal in two, compress one of them, and send to two channels in your mixer.

Then, using phase reverse on one of the channels, you subtract the compressed signal from the direct one.

At sub-threshold levels you could have any gain reduction down to completely cancelled, set by the relative gain of the two signals. When over-threshold - ie. actually compressing - you subtract relatively less, and thus get higher overall gain..

Or have I missed something?

Jakob Erland
Click to expand...

No you didn't miss anything. That is the only way.

I love these kind of quizzes...any more?
 
Technically I don't think this techique is (upward) expansion. An expander actually increases the loudness of passages that are acsending in volume. What you are talking about is parallel compression. Mastering engineers do this pretty often. It important to make sure you use delay to prevent phase shift and a compressor that has look ahead abilities.
If you look up Richard Hulse, he did some experimenting on parallel compression and develope some basic guidelines.

SoMm
 
Naa..

SoMm,

Nope, this IS expansion. Loud passages will be louder. Weak ones will be weaker.

What you are thinking about - the parallel compression - works by exactly NOT phase reversing one part of the signal. A very different thing...

Jakob E.
 
quote:
--------------------------------------------------------------------------------
Originally posted by gyraf
A simple way to do it could be to split the input signal in two, compress one of them, and send to two channels in your mixer.

Then, using phase reverse on one of the channels, you subtract the compressed signal from the direct one.

At sub-threshold levels you could have any gain reduction down to completely cancelled, set by the relative gain of the two signals. When over-threshold - ie. actually compressing - you subtract relatively less, and thus get higher overall gain..
--------------------------------------------------------------------------------

gyraf,

im going to be greedy and ask if you could just clarify a few things for me since it sounds as though youve had experience with this before.

- you split the signal... one goes direct to the console and one goes to the mono compressor and then to the console.

- one channel is phase reversed and both channels are then summed to one mono channel.

i am having trouble visualizing what exactly is happening to the audio after the phase reversal and summing of both channels. what are the characteristics of the audio resulting from the summation and how are these characteristics controlled?.. by just the compressor (threshold, makeup gain, etc) or the console levels also.

when i think of compression i think of a block or rectangular shape representing the dynamic range of the audio. the block has a line through it representing the threshold. the part of the block above the threshold is getting compressed according to the ratio. then the whole block is moved upward by a specified amound of gain.

when this compressed signal is then combined with its original signal (the direct signal into the other console channel) what is the result? and how does the expansion come out of this? i appreciate any help. i dont know why im not getting this.
 
Levey,

Think of it this way:

If you split your signal in two, route them to two channels in your board, and phase reverse one of them, you can cancel out the signal by setting the two faders to the same level.

Now, if you move one of them either up OR DOWN, you get LOUDER output.

What we do is have a compressor reducing gain over threshold, in effect the same as moving one fader, and hence actually increasing gain vhen compressing.

.. Hope this is a bit clear'er..

Jakob Erland
Gyraf Audio
 
jakob,

ok i think i have it now. you are changing the output and compression etc. with the faders. im going to try this when i have a session in less than an hour.

i didnt realize at first that you were from denmark. my past few springs have been spent in scandinavia, namely denmark. kind of odd for someone that lives in the u.s. to go to scandinavia for spring break. my brother lived in copenhagen for quite a while. i never had a bad time there. mange tak again for the help

rob
 
I realize now that my previous post was mistaken, however, I think what Jakob is describing is exactly the same thing and isn't expansion either.

Think about it this way. Say you keep the level of the uncompressed signal and the compressed (say 5:1), inverted signal the same. Below the threshold, you get total cancellation - above the threshold, the signal increases at a ratio equal to the inverse of the compression (.2:1). This is exactly the same as a compressor with ratios less than 1:1. An expander, on the other hand, actually attenuates the signal below the threshold by a given ratio and does nothing to the audio above the threshold. With the method described here, all signals below the threshold are attenuated by the same amount (set by the relative fader positions). With expansion the amount of gain reduction below the threshold is a function of the signal level - the lower the level, the bigger the decrease. There is a difference here.

It has been a long day though... ;)

I'm still stumped on this one at the moment, but I'm going to give it some thought tonight.
 
If you set compressor threshold to minus infinite, it's the same as an expander..
 
Then you're expanding everything and don't have a variable threshold above which the gain is unchanged. So you'd need to use the compressor again on the summed signal (unaltered plus inverted compressed signal) with a ratio twice that of the original compression. Then you'd have real expansion. I think that's it, what do you think?

I have a feeling your solution is probably what Levey's prof was after though....
 
gyraf said:
If you set compressor threshold to minus infinite, it's the same as an expander..
Click to expand...

Yes, If you have a compressor like that. For some reason I thik of an expander as increasing dynamic range, and upward expansion as a volume bump that leaves everything below the threshold alone.

If we go back to the first question, you can only answer it includes having a digital or VCA based compressor with values less than 1:1 on the compressor. Bob Katz said your basically reversing the phase on the sidechain signal. Not all compressors do this, and using the board violates the rule of circuitry modification. Bob also said that you almost always have to follow with a limiter and using output attenuation instead of make-up gain. Interesting technique Ill have to try out.

SoMm
 
OK, I tried to attach a file with some Output vs. Input curves to help explain this in case someone can't visualize it. (I'm really not trying to be patronizing, just trying to help...)

So, curve A represents the unaltered signal - a line with a slope of 1 - what goes in comes out. Curve B represents the compressed signal where beyond the kink in the curve (threshold), the output is reduced according to some ratio and the slope of the line decreases. Curve C is the output vs input for signal A minus signal B (the result of Jakob's method). Compare curve C to curve D (the curve for an expander) and you see there is a difference. With curve C, the output increases above the threshold according the inverse of the ratio in curve B. With the expander curve shown in D, below the threshold the output is decreased according to some ratio and above the threshold, the slope is still 1 and the signal is unchanged.

If you compress signal A with a threshold of minus infinity, you get the graph shown in curve E - the slope of the entire line decreases. Subtract curve E from the original signal, curve A, and you get curve F, the slope of the entire line increases like the lower half of the expander curve D. Compress the signal again and you get curve D, the same as the expander. I think this is the real answer to the question.

Hope the visual aids help a little. Let me know if that didn't make sense.
 

Attachments

  • curves.webp
    curves.webp
    6.7 KB · Views: 111
thanks a ton for all the help everyone. i get it now. FINALLY. thanks a lot for the drawings ebeam. they really helped. im going to keep them and put them in my notes. ha. awesome. im sure ill be back soon with another question because my professor likes to mess with our heads such as this. thanks again.
 
You must log in or register to reply here.
Back
Top