Forum rules
By using this "Production" sub-forum, you acknowledge that you have read, understood and agreed with our terms of use for this site. Click HERE to read them. If you do not agree to our terms of use, you must exit this site immediately. We do not accept any responsibility for the content, submissions, information or links contained herein. Users posting content here, do so completely at their own risk.
The sidechain in a compressor is the part of the compressor that "listens" to the audio input, and decide how much to lower the volume on the main chain. the side chain isn't the audio-out but a copy of the input. What we mean with sidechaining is replacing the original sidechain of the audio input with an external input, usually the kick.
So a compressor, whether it is sidechained or not, lowers the volume completely? Because i thought it lowered the loudest frequencies and kept the more quiet ones the same (depending on the threshold). So what you are saying is if i had perfect volume automations, i wouldn't need a compressor?
chillpenguin wrote:So a compressor, whether it is sidechained or not, lowers the volume completely? Because i thought it lowered the loudest frequencies and kept the more quiet ones the same (depending on the threshold). So what you are saying is if i had perfect volume automations, i wouldn't need a compressor?
It's not exactly frequency depended, you could make it by putting an eq in the chain though.
If you automate your volumes perfectly you don't need a compressor or a limiter. But it's really quite hard and
a compressor could also add a little texture to the audio.
When you route the kick through the sidechain input, it will compress other sounds whenever you hear the kick.
Would be a pain in the neck to record volume automation on every kick
Agent 47 wrote:Next time I can think of something, I will.
I know that it would be annoying to do the automations by hand but i was just trying to understand compression so i was asking if "in theory" compression is the same as automating the volume. Sounds like it is.
chillpenguin wrote:I know that it would be annoying to do the automations by hand but i was just trying to understand compression so i was asking if "in theory" compression is the same as automating the volume. Sounds like it is.
It isnt. The effect of compression does affect the volume of the output, but it is not the same as simply lowering the volume.
aka blinkesko Soundcloud Jesus Loves Electro - Burning Love (Augment remix)
chillpenguin wrote:I know that it would be annoying to do the automations by hand but i was just trying to understand compression so i was asking if "in theory" compression is the same as automating the volume. Sounds like it is.
It isnt. The effect of compression does affect the volume of the output, but it is not the same as simply lowering the volume.
Actually it is. The threshold and ratio just determine how much the volume is reduced when the signal goes over the threshold (or to continue with my metaphor, how much the hand on the fader lowers the volume), and the attack and release just determine how quickly the hand moves (a slow release, for example, would just mean that after the audio goes below the threshold, the hand on the fader is slow to return the fader to its normal position).
There is no mystical squashing of certain parts of the signal in compression. The volume of everything over the threshold is reduced by an amount determined by the ratio.
The confusing thing is sounds are usually made of multiple frequencies with varying amplitudes. So when you say "The volume of everything over the threshold is reduced by an amount determined by the ratio" i kind of picture the compressor reducing the volume of ONLY the frequencies going over the theshold. Because in a complex jumble of frequencies with varying amplitudes, it is very likely that some frequencies in a given sound will go over the threshold and some will not. Does the compress ONLY reduce the volume of what passes the threshold? Or does it reduce the volume of the entire sound when any one part of it goes over the threshold. If it's the latter than yeah it is exactly like a volume automation. But there seems to be a little confusion of what a compressor actually does. I did some research after posting my original question and i now believe a compressor is lowering the volume of the ENTIRE source if any one frequency goes over the threshold. I am still a little confused because people are saying different things and it is hard to know who to believe. But on an article i read on http://www.dnbscene.com they said that it is literally a volume automation. Hand on the fader style.
The compressor doesn't see all of the individual frequencies and compress each one differently. That would (probably) drastically change the timbre of the compressed sound. It just measures the level of the signal and treats it as a whole.
If you put a compressor on a drum buss it will treat it as one sound, one signal, that either goes over the threshold or doesn't. How is the compressor supposed to tell that the cymbal was what caused the over, and that it should only lower that?
the higher frequency oscillations ride the lower ones, its all combined into one wave form
if you look at a tune up close you can see the sine wave sub bending up and down with all the higher frequency waves running along it and each other, when it goes over the threshold the whole thing gets dipped, the amount of dip set by the ratio and the envelope of the dip set by the attack and release.
A 'multiband compressor' or a 'dynamic eq' will let you set frequency ranges to compress individually with different settings, but a regular compressor is purely volume automation.
ultimately its all being pumped into a speaker cone which vibrates, so its got to sum up to 1 line moving backwards and forwards from the centre, if you think about wave forms like that pic being a pattern of the cones vibrations back and forth it starts to make sense of how it creates multiple frequencies at the same time.
You are right though in a sense with what you are thinking, you can see the low frequency waves are giving all the volume to that sound, without them it would be very quiet, thats why some people go for the multiband compressors, so they can compress the low end without damaging the tops which arnt really causing much trouble on their own in terms of peaks.
Brian Oblivion wrote:the higher frequency oscillations ride the lower ones, its all combined into one wave form
if you look at a tune up close you can see the sine wave sub bending up and down with all the higher frequency waves running along it and each other, when it goes over the threshold the whole thing gets dipped, the amount of dip set by the ratio and the envelope of the dip set by the attack and release.
A 'multiband compressor' or a 'dynamic eq' will let you set frequency ranges to compress individually with different settings, but a regular compressor is purely volume automation.
good post.
This is why, if you're going to highpass and compress something, you put the filter before the compressor. (Unless you want that sound as a creative effect.)
I do volume automation in the form of one-shot lfo's on the gain. Faster than automation haha
I've actually been thinking about this lately and I'm trying to decide which method would work best. I want to get it set up to where the pumping can trick you into hearing the thump of a kick on laptop speakers/stuff without any bass. Also to have the pumping from the snare. Like in this tune:
AxeD wrote:This is as far as we can take this topic basically
Compression reduces overall dynamic range, thus lowering the volume. Sidechain compression is the same, but an
external input triggers the compressor.
Well not necessarily. Compression as has been discussed in this thread is very oversimplified. Yes a compressor can reduce output based on input etc, but how comps do this is far more nuanced, resulting in a more nuanced sound than simply lowering (or raising, depending) the volume.
Attack is how quickly the compressor reduces the volume after the signal goes over the threshold.
Release is how quickly the compressor returns to normal volume after the signal returns back below the threshold.
Lookahead is where the input is split, one signal goes off to the sidechain to trigger gain reduction, and the other signal (that is going to have its gain reduced) is delayed slightly.
This means that smoother sounding, slower attack settings can still be used to catch transients.
Faster than automation haha
