EQ is a tool like a sword in that it has two sharp sides to be aware of. EQ does something specific, the issue is when do you use it and when don’t you.
A background is needed on how things work to see how eq “fixes things”.
In a time invariant system, like a simple circuit made of analogue parts like R’s, L’s and C’s etc, one has a situation where if one alters the amplitude in any way with a filter, one finds there is always a corresponding phase shift that accompanies it. For example if you have a 2nd order (12 dB/oct) filter, one observes a 180 degree phase rotation going from way above to way below he filter corner F. A fourth order filter causes a 360 degree rotation, IE, 90 degrees per order or 5dB slope.
By Phase shift, I mean that the output is delayed or in special cases advanced relative to the input signal. Zero degrees shift means the output and input track in time while say a 180 degree shift means the output is delayed ½ cycle relative to the input.
Thus, if one wants to reproduce visually recognizable but a wide band signal like a square wave, one must have flat amplitude and no phase shift (time) for about 1/10 to 10X the square frequency.
Flat amplitude but having phase shift alters the shape just as not flat amplitude.
Since phase is also a view of time, when we adjust phase, we play with time as well.
So, we have a simple non-flat circuit made of R’s C’s and L’s which we will pretend we want to fix with eq. We measure it’s magnitude an phase response and say, boy, I can fix that with eq.
So you insert an eq in the line and by measuring the output and fiddling, you make it flat AND by doing so, one also fixes the phase shifts that went along with the amplitude flaws.
Naturally this part sounds a lot like the panacea coolaid pitch used to introduce audio gear.
I saved a long time to get my first EQ in the late 70’s and then built my fist third octave unit. Eventually the price fell to that of dirt like it does for everything electronic in the modern era.
So, the down side;
One is potentially adding a noise and distortion source, often “bypass” still goes through circuitry. This may or may not be an issue, if you can have it in circuit and in / bypass on your preamp and not hear a difference when set flat, then it’s not an issue.
The reason it has /got a “bad name” (aside from units suffering from the above issue).
An obvious use might be to “fix” ones loudspeakers and while if one measured them outdoors and limited the corrections to below say 300-500Hz and to no more than about + - 6 or 10 dB, than this too works well.
The problem is, in a room what one measures is both a mix of the normal signals BUT ALSO reflected signals which are delayed in time and add back into the sum at the mic position. This means a signal that is delayed and returns a half wavelength or an odd number of half wavelengths later arrives out of phase and cancels the original signal. This appears as a notch in the frequency response, if you have a high resolution measurement, it can be quite deep.
This kind of response alteration is NOT a time invariant effect and so to the degree that is the cause, an attempt to fix the amplitude will alter the phase response adversely. In the old days of live sound, the thumb rule was only cut lumps and peaks, never try to fill a deep pointy hole (likely a cancellation notch from a reflected signal).
Bottom line, I think for systems that are set up by measurement, a parametric equalizer is the way to go.
At the opposite extreme, for a system that will be tuned to taste by the album or CD, then simple tone controls might be more useful than say a 1/3 oct eq let alone a parametric.
In any case, the idea that tone or eq adjustments are somehow traitorous to the purity ideal, remember if you hear a tonal issue that you can make more pleasing , then all the better.
Don’t forget that particularly with older recordings, there were larger differences between the voicing on recordings.
Back then there were more different kinds of loudspeakers used in the recording and mastering stages, a recording mixed through very bright speakers will probably sound dull on most systems because the nature of the monitors is imposed on the mix’s spectrum.
With CD’s, the problem is often they are too bright and compressed reflecting the need for “portable music” which can be save with a minimum number of bits and dynamics compatible with being listened to while wing walking or driving Nascar.
Lastly, not a hifi story but a story from the mixing world where eq is ubiquitous. There is a story I heard at a Synaudcon meeting in the 80’s an then a number of times since. It is meant to be a reminder of how your mind and senses work blending your expectation into what you “perceive”.
An engineer was going over a final mixdown and applying the finishing touches.
He hears something in the piano that isn’t quite right. He spends 20 -30 minutes tweaking the eq and finally is satisfied.
He sits back and listens to the whole track twice and then, notices the “Bypass” button has been pressed the entire time. His adjustments did nothing to the sound, everything he heard, he was imagining. Ask anyone who mixes sound or works on complicated audio systems if something like this has ever happened to them and if they are honest, they will say yes. Moral of the story, check your work often, you are also human.
Best,
Tom Danley
Danley Sound Labs
These might be fun, they help associate "the sound" to the actual numeric pitch, making eq by ear faster.
http://sft.sourceforge.net/
http://www.audio-production-tips.com/audio-equalizer-ear-training.html
A background is needed on how things work to see how eq “fixes things”.
In a time invariant system, like a simple circuit made of analogue parts like R’s, L’s and C’s etc, one has a situation where if one alters the amplitude in any way with a filter, one finds there is always a corresponding phase shift that accompanies it. For example if you have a 2nd order (12 dB/oct) filter, one observes a 180 degree phase rotation going from way above to way below he filter corner F. A fourth order filter causes a 360 degree rotation, IE, 90 degrees per order or 5dB slope.
By Phase shift, I mean that the output is delayed or in special cases advanced relative to the input signal. Zero degrees shift means the output and input track in time while say a 180 degree shift means the output is delayed ½ cycle relative to the input.
Thus, if one wants to reproduce visually recognizable but a wide band signal like a square wave, one must have flat amplitude and no phase shift (time) for about 1/10 to 10X the square frequency.
Flat amplitude but having phase shift alters the shape just as not flat amplitude.
Since phase is also a view of time, when we adjust phase, we play with time as well.
So, we have a simple non-flat circuit made of R’s C’s and L’s which we will pretend we want to fix with eq. We measure it’s magnitude an phase response and say, boy, I can fix that with eq.
So you insert an eq in the line and by measuring the output and fiddling, you make it flat AND by doing so, one also fixes the phase shifts that went along with the amplitude flaws.
Naturally this part sounds a lot like the panacea coolaid pitch used to introduce audio gear.
I saved a long time to get my first EQ in the late 70’s and then built my fist third octave unit. Eventually the price fell to that of dirt like it does for everything electronic in the modern era.
So, the down side;
One is potentially adding a noise and distortion source, often “bypass” still goes through circuitry. This may or may not be an issue, if you can have it in circuit and in / bypass on your preamp and not hear a difference when set flat, then it’s not an issue.
The reason it has /got a “bad name” (aside from units suffering from the above issue).
An obvious use might be to “fix” ones loudspeakers and while if one measured them outdoors and limited the corrections to below say 300-500Hz and to no more than about + - 6 or 10 dB, than this too works well.
The problem is, in a room what one measures is both a mix of the normal signals BUT ALSO reflected signals which are delayed in time and add back into the sum at the mic position. This means a signal that is delayed and returns a half wavelength or an odd number of half wavelengths later arrives out of phase and cancels the original signal. This appears as a notch in the frequency response, if you have a high resolution measurement, it can be quite deep.
This kind of response alteration is NOT a time invariant effect and so to the degree that is the cause, an attempt to fix the amplitude will alter the phase response adversely. In the old days of live sound, the thumb rule was only cut lumps and peaks, never try to fill a deep pointy hole (likely a cancellation notch from a reflected signal).
Bottom line, I think for systems that are set up by measurement, a parametric equalizer is the way to go.
At the opposite extreme, for a system that will be tuned to taste by the album or CD, then simple tone controls might be more useful than say a 1/3 oct eq let alone a parametric.
In any case, the idea that tone or eq adjustments are somehow traitorous to the purity ideal, remember if you hear a tonal issue that you can make more pleasing , then all the better.
Don’t forget that particularly with older recordings, there were larger differences between the voicing on recordings.
Back then there were more different kinds of loudspeakers used in the recording and mastering stages, a recording mixed through very bright speakers will probably sound dull on most systems because the nature of the monitors is imposed on the mix’s spectrum.
With CD’s, the problem is often they are too bright and compressed reflecting the need for “portable music” which can be save with a minimum number of bits and dynamics compatible with being listened to while wing walking or driving Nascar.
Lastly, not a hifi story but a story from the mixing world where eq is ubiquitous. There is a story I heard at a Synaudcon meeting in the 80’s an then a number of times since. It is meant to be a reminder of how your mind and senses work blending your expectation into what you “perceive”.
An engineer was going over a final mixdown and applying the finishing touches.
He hears something in the piano that isn’t quite right. He spends 20 -30 minutes tweaking the eq and finally is satisfied.
He sits back and listens to the whole track twice and then, notices the “Bypass” button has been pressed the entire time. His adjustments did nothing to the sound, everything he heard, he was imagining. Ask anyone who mixes sound or works on complicated audio systems if something like this has ever happened to them and if they are honest, they will say yes. Moral of the story, check your work often, you are also human.
Best,
Tom Danley
Danley Sound Labs
These might be fun, they help associate "the sound" to the actual numeric pitch, making eq by ear faster.
http://sft.sourceforge.net/
http://www.audio-production-tips.com/audio-equalizer-ear-training.html
