Help Understanding Compressor

[Lunax]

Well I basically feel like I understand the basic functions of the Ableton Compressor. But one part that confuses me so much, is the knee chart. It might be because I'm overlooking something or don't understand a certain part, but when I change the ratio the output doesn't seem to be accurate with the ratio I put. I'll give an example.

Ex: My threshold is at -24.00 dB and my ratio is at 12.00. And on the high knee chart it goes upward until it reaches the threshold line indicated on the chart and it doesn't increase as drastically. Which is what I expect. But it clearly ends at -6.00 dB. Wouldn't end around -2.00 dB because there is only two 12s in 24? Not 6? Since for every 12.00 dB above the ratio it will increas the output by 1 dB.

I'm either just really stupid when it comes to chart and ratios or I overlooked another component of the chart. (or both) It's hard to describe the problem over text, it would be much easier to show you guys in person. But if anyone knows what I'm doing wrong help would be appreciated A LOT. I've looked at youtube tutorials and everything and I just can't seem to figure out what's going on.

[Linear Phase]

squash dat ship... squeeze it, juice it, crush it! lol.. how about laying off the threshold, and try and keep those ratios to 4:1 and under. as for knee, that's all about the curve of the compression.. a harder knee is probably good on drums fer example, while soft would be good on vocals... that said.. having too hard a knee on your drums is not going to be a good sound.

you have to be careful with the tools.. they sort of do things, that are not needed, in a mix.. like compression of 13:1 with a thresh of -45, and knee of infinite

[fishmonkey]

Well I basically feel like I understand the basic functions of the Ableton Compressor. But one part that confuses me so much, is the knee chart. It might be because I'm overlooking something or don't understand a certain part, but when I change the ratio the output doesn't seem to be accurate with the ratio I put. I'll give an example.

Ex: My threshold is at -24.00 dB and my ratio is at 12.00. And on the high knee chart it goes upward until it reaches the threshold line indicated on the chart and it doesn't increase as drastically. Which is what I expect. But it clearly ends at -6.00 dB. Wouldn't end around -2.00 dB because there is only two 12s in 24? Not 6? Since for every 12.00 dB above the ratio it will increas the output by 1 dB.

I'm either just really stupid when it comes to chart and ratios or I overlooked another component of the chart. (or both) It's hard to describe the problem over text, it would be much easier to show you guys in person. But if anyone knows what I'm doing wrong help would be appreciated A LOT. I've looked at youtube tutorials and everything and I just can't seem to figure out what's going on.

you are getting confused by the makeup gain. turn that off and the graph will look the way you expect it to.

[Lunax]

http://t.co/Gy2wcpSK
Heres a pic. Tell me if I labeled the output and input right. The arrow is pointing at where I think the orange line would be once it passes the threshold. But It's not where I think it should be. This is with makeup off.(I'm too stupid to post pictures directly on to the forum, so the link takes you to my twitter account where I posted the picture)

Does the orange line after the threshold on the graph indicate the amount of gain reduction in dB? Or just the output after the ratio? Or what, whats it supposed to show? Because if its the output it doesn't make much sense to me.

[fishmonkey]

the graph is correct, however you are reading it wrong.

input level is the x-axis, output level is the y-axis.

when the input level is below threshold (-24 dB), the output level is the same as the input.

then for each 6 dB increase in input level over threshold, the output level increases by 6/12 dB (half a dB).

therefore, at +6 dB input, the output level is -24 + 2.5 dB (-21.5 dB), which is exactly what the graph shows.

[Lunax]

why would it increase by half of a dB? and why would you add 2.5 dB to the output? wouldn't it be 2.00 dB?

[fishmonkey]

a ratio of 12 means that for every 1 dB that the input is over threshold, the output is squashed to 1/12 dB over threshold.

e.g. if the input is 12 dB over threshold, then the output will be 12/12 dB or only 1 dB over threshold. it follows that for every 6 dB that the input is over threshold (one square on the graph grid), the output will be 6/12 dB over threshold.

on the graph, the maximum input level is at +6 dB. the difference between that and the threshold is 24 + 6 = 30 dB, i.e. the maximum input level on the graph is 30 dB over the specified threshold. and 30/12 dB = 2.5 dB.

[Lunax]

And what did you mean by "when the input level is below threshold (-24 dB), the output level is the same as the input."? were you referring to the picture I posted and can you please explain more on what that means.

[Lunax]

Okay I think I finally figured it out. But I think I was looking at it differently then you were explaining.

I looked at it this way, and it seems to be correct.

If my threshold is -24 dB then I imagine my output as being -24 dB, and if my ratio is 2.00 then it adds 12 dB to the output making it at -12 dB.

And if my threshold is at -12 dB and my ratio is at 4 dB. Then I imagine the output starting out at -12 dB and increasing from there. Meaning that since there is three 4s in a 12 then I add three dB to the -12 making it -9 dB.

IDK if this is the correct way to look at the chart but it hasn't been proven inaccurate so far.

[fishmonkey]

And what did you mean by "when the input level is below threshold (-24 dB), the output level is the same as the input."? were you referring to the picture I posted and can you please explain more on what that means.

the threshold determines at what input level the compressor starts to act. at or below the threshold, the compressor does nothing, it passes the audio straight through. so if you set the threshold to -24 dB and the input is -24 dB or below, then the output will be exactly the same level (assuming that makeup gain is turned off).

look at the graph, that's why the line is on a 45 degree slope up to the threshold level.

[fishmonkey]

Okay I think I finally figured it out. But I think I was looking at it differently then you were explaining.

I looked at it this way, and it seems to be correct.

If my threshold is -24 dB then I imagine my output as being -24 dB, and if my ratio is 2.00 then it adds 12 dB to the output making it at -12 dB.

And if my threshold is at -12 dB and my ratio is at 4 dB. Then I imagine the output starting out at -12 dB and increasing from there. Meaning that since there is three 4s in a 12 then I add three dB to the -12 making it -9 dB.

IDK if this is the correct way to look at the chart but it hasn't been proven inaccurate so far.

nope, sorry.

assuming that makeup gain is turned off, below the threshold, the compressor does nothing. for all input values greater than the threshold, the output level is equal to the threshold plus the extra level that is above threshold divided by the ratio.

here are some examples with threshold = -24 dB, ratio = 4:

if the input level is -30 dB, then the output level is also -30 dB.

if the input level is -16 dB, then this is 8 dB over the threshold. you divide this 8 dB by the ratio (8/4 dB = 2 dB) and add it to the threshold to get the output level. therefore the output level in this case is -24 dB + 2 dB = -22 dB.

if the input level is -20 dB (4 dB over threshold), the output level = -24 + (4/4) = -23 dB.

i'm not sure how to make it clearer.

[fishmonkey]

more succintly, assuming that makeup gain is off:

if input level <= threshold, output level = input level

if input level > threshold, output level = threshold + ((input level - threshold) / ratio)