O
Orchestrator1
New member
Does a compressor alter the tonal quality of an instrument after compression.
Rex.
Rex.
rayc
retroreprobate
The short answer is YEP.
Some comps are preferred because of the colouration that is inherent in there process.
Realistically any processing will alter the signal, & therefore the tone, to some degree be it audible or not.
Some comps are preferred because of the colouration that is inherent in there process.
Realistically any processing will alter the signal, & therefore the tone, to some degree be it audible or not.
dualflip
Audio Engineer
The long one is YES too, aside from the fact that the compressor alters the envelope of a signal, it also adds coloration depending the electronics that it has in it, for example in a hardware compressor that uses transformers or tubes, the sound will inherit this, just by passing audio through a compressor (even if you dont compress ) youll be coloring the sound, of course some of them are designed to color the sound very very little, almost nothing, and others are designed to color the sound a lot, in the end ill answer something like : YES, but some more than others.
SouthSIDE Glen
independentrecording.net
Another longer answer is YES, EVERY compressor potentially changes to tonal quality of an instrument, not mainly through the act of coloration (distortion above and beyond the pure compression itself), but rather by the pure act of compression itself.
G.
G.
dualflip
Audio Engineer
I agree with you, but i have to say that coloration goes beyond distortion, coloration depends on the type of electronic components used, design, etc.. it may be as subtle as a low-end bump or a completely drastical change that a piece of gear impregnates in the sound. Just wanted to point this out because a lot of people think that coloration, or warmth as said commonly is only achieved by the use of "tubes driven hard" thats a lie, in fact there are some tube pre-amps out there (Avalon, Millenia among others) that are designed to be very very transparent.
NL5
Unpossible!
One word - HARMONICS 
SouthSIDE Glen
independentrecording.net
I'm using the term distortion in it's pure sense: any difference in the output from the input - any change in the signal - is a distortion of the original signal. The only question is to whether any given difference is intended or incidental. The coloration caused by circuit design is a form of incidental distortion; pushing that circuit out of equilibrium is a form of intended distortion.
But that's beside the point I wanted to make. My point is that even a theoretical "colorless" compression design is going to change the tonal balance of an instrument running through it, and do so in a way that's potentially more distortive of the original signal and it's tonal characteristics than any further circuit coloration would be.
Any time you set a threshold on a compressor, you are in effect separating the signal into two loudness zones. If the signal is coming from a musical instrument, complete with it's full range of forments, harmonics and so forth, these timbral elements are also going to be divided into two zones. Some of them are going to be at energy levels that fall mostly below the threshold.
So, when one performs the compression, the relative energy/volume levels of those tones riding under the threshold and those riding over the threshold will automatically be brought closer together. This is - by it's very definition - a change in the tonal balance of the signal, regardless of whether there's any extra incidental coloration or not.
G.
KineticSound
The VOICE
What about frequency response?
Glenn:
What happens to the frequencies? Let's say I start with a sound wave, Picture 1. Then I compress the sound wave, and it becomes Picture 2. This is how I've always thought of compression, but as it applies to the dynamic range of the sound wave...
When I'm applying compression, I understand what you're saying about my tonal balance changing, because I'm now blending 100% of my low-energy sound with a smaller percentage of my high-energy sound, but how does this apply to the frequencies of the sound wave? Would it mean that I am compressing more of the low-frequency end of the spectrum, since low frequencies carry more energy? That would tell me that I should end up with a brighter sound, but that's not what actually happens...
Are the dynamics of the sound wave so much more prominent that I just don't notice how the frequency balance is changing - i.e. it's so subtle that I notice the dynamic change over the tonal change?
Glenn:
What happens to the frequencies? Let's say I start with a sound wave, Picture 1. Then I compress the sound wave, and it becomes Picture 2. This is how I've always thought of compression, but as it applies to the dynamic range of the sound wave...
When I'm applying compression, I understand what you're saying about my tonal balance changing, because I'm now blending 100% of my low-energy sound with a smaller percentage of my high-energy sound, but how does this apply to the frequencies of the sound wave? Would it mean that I am compressing more of the low-frequency end of the spectrum, since low frequencies carry more energy? That would tell me that I should end up with a brighter sound, but that's not what actually happens...
Are the dynamics of the sound wave so much more prominent that I just don't notice how the frequency balance is changing - i.e. it's so subtle that I notice the dynamic change over the tonal change?
Attachments
NL5
Unpossible!
I'm sure Glen is typing a much better answer, but here goes - any instrument makes a sound by producing a fundamental frequency and many "harmonics" of that frequency. Their relative balances make up the "sound". So, each of these frequencies have an associated amount of energy to them. A compressor will squash the higher energy ones, bringing out the lower energy harmonics. Thus, altering the "tonality" of the instrument.
edit - and it's the #1 reason I use a compressor. Automation is so much better at gain control........
edit - and it's the #1 reason I use a compressor. Automation is so much better at gain control........
SouthSIDE Glen
independentrecording.net
Kinetic,
NL5 provided a key clue when he referred to harmonics, which will be a major (but not the only) source of lower-energy frequencies.
Your diagrams are perhaps just a bit - unintentionally - misleading because they are showing what I assume is intended to be a sine wave of a specific frequency. Applicable, perhaps, to an analog synth voicing a single oscillator or some other instrument that is a source of a pure tone. Even then, a real-world sine wave is really just a collection of harmonic frequencies summed together into a "larger" sine wave.
But for a maybe easier to visualize - ane more real world - example: For most instruments, the waveform will be much more complex. What peeks/peaks above the threshold may be lower-frequency stuff, but it can also include high-frequency transients, not to mention higher frequency elements that are in effect modulating the lower frequency carrier. Also there can be low frequency "mud" and subharmonics and so forth that are lower in overall volume than the "main" forments. All in all, it can be a real mish mosh.
If you were to take a signal (even your sine wave) and slice it into 2 or 3 or 4 horizontal volume range slices, and take a spectrogram or FFT analysis of each slice, you'd find each has it's own spectral distribution or fingerprint. So it naturally follows that if you change the relative volumes of these "zones" you are going to be changing the composite spectral fingerprint of the overall signal.
I think we've all experienced a situation where something sounded "good" - say a 2mix buss, for example - but when we needed to reign in it's dynamics a bit with a compressor what came out the other end sounded honk-ish or harsh or some other unpleasant adjective, because we inadvertantly adjusted the tonal balance to unintentionally emphasize some of the more unpleasant formants that were at relative low levels before we compressed. This has nothing to do with the "coloration" of the compressor itself, but rather just that the act of compression itself exposed the previously low-level some blemishes in the sound.
EDIT: NL5, my answer (as usual) was only longer. Yours was better
- G.
G.
NL5 provided a key clue when he referred to harmonics, which will be a major (but not the only) source of lower-energy frequencies.
Your diagrams are perhaps just a bit - unintentionally - misleading because they are showing what I assume is intended to be a sine wave of a specific frequency. Applicable, perhaps, to an analog synth voicing a single oscillator or some other instrument that is a source of a pure tone. Even then, a real-world sine wave is really just a collection of harmonic frequencies summed together into a "larger" sine wave.
But for a maybe easier to visualize - ane more real world - example: For most instruments, the waveform will be much more complex. What peeks/peaks above the threshold may be lower-frequency stuff, but it can also include high-frequency transients, not to mention higher frequency elements that are in effect modulating the lower frequency carrier. Also there can be low frequency "mud" and subharmonics and so forth that are lower in overall volume than the "main" forments. All in all, it can be a real mish mosh.
If you were to take a signal (even your sine wave) and slice it into 2 or 3 or 4 horizontal volume range slices, and take a spectrogram or FFT analysis of each slice, you'd find each has it's own spectral distribution or fingerprint. So it naturally follows that if you change the relative volumes of these "zones" you are going to be changing the composite spectral fingerprint of the overall signal.
I think we've all experienced a situation where something sounded "good" - say a 2mix buss, for example - but when we needed to reign in it's dynamics a bit with a compressor what came out the other end sounded honk-ish or harsh or some other unpleasant adjective, because we inadvertantly adjusted the tonal balance to unintentionally emphasize some of the more unpleasant formants that were at relative low levels before we compressed. This has nothing to do with the "coloration" of the compressor itself, but rather just that the act of compression itself exposed the previously low-level some blemishes in the sound.
EDIT: NL5, my answer (as usual) was only longer. Yours was better
- G.G.
KineticSound
The VOICE
Bad example. I drew that quickly just to show the effect of the compressor clamping on the peak at the top of the wave that passed the theoretical threshold.
I think I grasp what you're saying, however. That the compressor may be clamping down on the most energetic portion of the dynamic curve and leaving the harmonic "artifacts" that came along for the ride unaffected. If I'm off base, let me know.
David
NL5
Unpossible!
....and this is why I mix thru a 2-buss comp and limiter. However, one really needs to understand what they are doing, so as not to overload them. It's definitely an art, that requires some learning......
X
XLR
______
Sometimes more, sometimes less. Depends. What instrument? What compressor? What settings? How much gain reduction? Does it sound to you like the tone quality changes?
suprstar
It aint ez being green
NOTHING happens to the frequencies, it's all about amplitude.. In this very simple example, the 1st pic is a standard sin wav, the 2nd pic is a highly compressed sin wav, the slope at zero is a LOT higher, and when it gets to the peaks, it stays there longer than it should. It's not a pure tone anymore, and thus does not sound exactly like one anymore. This example would be a compression with a -inf threshold. If your threshold was high, then just the peaks would look like the compressed wav, and the parts near zero would look loke the standard wav.
Attachments
SouthSIDE Glen
independentrecording.net
Well, yes and no on both counts. You're saying the right thing, if I read you right, but the above statement is a bit contradictory.
Sure, the fundamental frequency of the sine remains the same, but the harmonic content does change. As you say later, the two different versions of the sine do sound different in the same way as a sine wave and a triangle wave of the same frequency will sound different, even if they are of the same peak amplitude (though the triangle will sound more different, of course.)
They sound different because the amplitudes of the constituent harmonic frequencies change. This is what changes the "tone" of the signal between the two different wave shapes. If you take a spectral analysis of the two different sides of your graph, they will indeed look different as well as sound different, even if you were to even out the gain at output, because even though the fundamental frequency and peak amplitude of the wave is not changing, the harmonic content is.
Now, this is all whiteboard stuff. XLR does imply a good point in that just how *audible* these changes may or may not be depends greatly upon all the factors listed. if you have a high threshold with a soft knee just to smooth some wild transients, for example, the actual audibility of tonal change will probably be questionable, and probably masked by the distortion introduced by the "color" of the cpmpressor itself.
However, the strict OP question of whether compression changes the tonal balance of an instrument, has a strict and definitive answer of "yes, it does", regardless of any further coloration.
G.
KineticSound
The VOICE
Frequencies
suprstar,
If I'm understanding Glenn and NL5's discourse, then that's technically correct, but in a practical sense the harmonic frequencies will end up amplified.
OK, more graphs! (below). In the input signal, I have big dynamic range that is full of my dominant frequency, right? And somewhere below that I have all my harmonic frequencies. Then I clamp on a compressor, and the signal changes to "compressor output". My dynamic range is reduced, and I'm going to end up amplifying the signal. As I think I understand it now, the harmonic frequencies that were once buried are now right up there with my somewhat flattened dominant frequencies. The frequencies haven't changed, but the balance between dominant and harmonic has changed. Am I on the right track?
suprstar,
If I'm understanding Glenn and NL5's discourse, then that's technically correct, but in a practical sense the harmonic frequencies will end up amplified.
OK, more graphs! (below). In the input signal, I have big dynamic range that is full of my dominant frequency, right? And somewhere below that I have all my harmonic frequencies. Then I clamp on a compressor, and the signal changes to "compressor output". My dynamic range is reduced, and I'm going to end up amplifying the signal. As I think I understand it now, the harmonic frequencies that were once buried are now right up there with my somewhat flattened dominant frequencies. The frequencies haven't changed, but the balance between dominant and harmonic has changed. Am I on the right track?
Attachments
suprstar
It aint ez being green
That's true, I understand, I was being overly simplistic.
definately, while a 440 A is always that, it's the harmonic differences that make different 440's sound like flutes, guitars, pianos, etc..
suprstar
It aint ez being green
I think so, sounds correct to me,
iqi616
Yet another Mike
These are the best pics so far that show the action of the compressor. Some people think that the compressor acts on each wave distorting it but that only happens on bass instruments if the attack and release are too fast. A compressor/limiter is a voltage controlled amplifier and the control voltage tracks the peak voltages. Some compressors let you select the RMS (average) signal level instead of the peaks.
Your illustration shows the first cycle getting through so this compressor is either a little bit slow reacting or the attack control has been set to let the leading edge through - that's good for emphasizing transients. In fact that is one way that compressors affect the perceived tone of the sound - attenuated transients make the sound less bright and accentuated transients make the sound brighter.