Not all graphic eq's work like this. For example, the 560 from API's midband is a very wide band for broad strokes. I would argue that this kind of graphic eq is good for certain applications, just like any tool. It's just a tool, and it's your knowledge of how to use it and what it does to your signal that helps you choose the right tool for the job at hand. Also, not all parametric eq's can get as tight of a q as the one you're using, which again comes down to the differences between eq builds and not the difference between graphic vs parametric. I would argue that how the result sounds is the only important factor. What you told it to do depends on what you hear while applying the tool.
Sure, that's why I feel it is important you know how your tools work by measuring them. And a 100% with you on using whatever gets the job done. Just wanted to point out that people often assume that just that one particular frequency of the GEQ is affected, when in fact it is not so.
@@gainmedialab Sure. But there's no such thing as an eq that can notch out 1000 Hz without impacting 999Hz or 1001 Hz. You're going to be shifting frequencies around it, even with the absolute narrowest q possible. It's good that people understand that as a fundamental of how audio processing works. But what application does this have on the ground?
I agree. I just want people to know what happens to the signal when you, for instance, try to shape an EQ curve to a target trace for your PA system and keep getting different results from what you are seeing on the GEQ set filters.
Thanks Aleš! I’ve been wanting a video exactly like this. So helpful to see what’s happening with GEQ - definitely helps to understand our audio tools better.
@@gainmedialab I've switched off of GEQ but I'll admit I still have one on my vocal buss for feedback surprises. It still feels like the fastest way to grab a frequency if I hear something getting ready to give me a problem. That's partially a GUI issue and partially a habit that needs breaking. LOL
if you narrow the Q of a parametric and cut 1k by 15db, it will also pull down the neighboring frequencies. so with parametric, you have maybe six, maybe four frequency ranges you can alter, each affecting the nearest frequencies to it. with a graphic, you have so much more control. if you wreck it to the point of phasing then you have problems on your board.
With a PEQ you can define the width of the Q and be much more precise. With a GEQ you don't really have a great representation of what it is doing and how the filters affect the signal. Just trying to shed some light on the fact that the GEQ filters are not notches, but can be much more wide than people suspect.
the best representation of what a graph is doing is *your ears*. a parametric is great on an individual channel - you certainly should be able to make any signal sound great with five points of adjustment. but the main mix for the entire house needs much more nuance than that.... in my humble opinion :)
Thanks for the informative video. I am old school sound engineer and used mainly analogue equipment because digital wasn't around at the time. I encountered on tour every night different equipment and remembered a cheap nasty Yamaha Eq on a system and just pulling a frequency I was getting feedback. It was so inaccurate. Anyway are well build graphic analogue EQs behaving similar???? Also use to work on same venue for many years using DB monitors. Always pulled maybe 2 or 3 freq out if needed at all. Nice powerful monitor. One night visiting tech brought his cheap Allen and Heath digi desk did some drastic EQ and the monitor never sounded so bad and hardly any volume. Anyway curious if analogue EQ are adding the same inaccuratecy. Thanks again. Raymond Steeg
All GEQs have the same behaviour, but with various Q widths. My point was that you actually can't be sure until you measure, but 9 out of 10 times the neighbouring filters will affect each other - the quality of the build and the desired character of the tool will determine how much.
Indeed. I remember the same. The Klarks you could count on. XTA great . BSS, Opal, smoother (probably wider) but also great. Ashleys and Ranes were sometimes ok, sometimes not. dBx caused a bit of concern occasionally and Behringer analog graphics were a bit of a joke.
This is great advice. It should be no secret by now that the built in graphic eqs on digital desks generally use exactly the same biquad filter math as their parametric counterparts. But the parametrics allow one further control of the Q and the exact frequency you want to hit. Super narrow notching on these eqs can sometimes create as many problems as they solve, but whatever works. To say that a true freq/phase plot of a typical venue looks like a seizure in graphic form would be an understatement. It's among the top reasons why we need high q, narrow filters for live sound.
I use an EQ to cut or lower frequencies most of the time but what if we boost frequencies, does the same thing happen when you boost frequencies that are next to each other?
Okay, you make a lot of good points. Now, how can you have more than 5 adjustment points on the VEQ? Why cant you get a smaller Q setting on the GEQ? After all, this is all digital and SHOULD be more customizable.
But you never mention it the impact of phase correlation of the eq's and the impact it has on the phase. In theory you can use pinpointed PEQ to cut a frequency better that the GEQ, but can you pinpoint it exactly during showtime? That is why GEQ is useful. I tend to use PEQ when doing soundcheck, when playing safe. I use GEQ when fixing things during show I use TEQ on matrix busses to calibrate primary or secondary systems, As you can see, TEQ has the least impact on phase offsetting, which i think is the most important for a clear and pristine sound, and if you found yourself eq'ed too much, the system just sounds like through curtain, and is unable to punch lungs, drop guts and bite. Nice video though, i want to see you compare different models of digital mixers' eq's for phase correlation, do some mixers have better designed eq's, what about vs analog GEQ like the BSS maybe... Keep up the good work.
Thanks, I would like to do that as well, but currently only have access to this console. Once we start working again, I will hopefully do the same with other consoles. And I agree with you on the phase issue, but I thought it was just too much information to cram into one video. I might do a follow up addressing just phase and the time domain of these tools.
"PEQ to cut a frequency better that the GEQ, but can you pinpoint it exactly during showtime? That is why GEQ is useful. " How is it faster to pinpoint the ringing frequency with a GEQ than with a PEQ ?
Ok. So for FOH PA systems, if you need much manipulation even in your parametric EQ, pack up the rig and get one that presents a balanced response for the frequency range needed for the application(talking head events don’t need 30hz). So use your parametric on the PA, but I swear if you need more than one band any kind of huge notch then there are bigger problems, regardless of the room, save for very extreme circumstances. In other words, the PA should present as flat. Any issues or corrections should be done in the corresponding channels. The moment you shape a PA around a lead vocal mic, is the moment you should just send the band home. It’s not the fault of the snare drum, or the bass, or the keyboards, that a vocal mic has to be the most compromised input as it is usually the one with the highest input level to output needed discrepancy of any mic on stage. So when you shape around a vocal or even a great track, everything else suffers. Now the keys sound hollow, the snare is all highs and lows, with no meat, the bass has disappeared, but you can feel it. Ugh. The cymbals are all glassy. All because of a vocal mic. So shape the vocal mic with the parametric. Need more? Well, first off, sorry. But there are plenty of ways to put more eq into the chain. Regardless of whether I’m doing the symphony or death metal or the mayoral inauguration, the PA eq is all but flat. The input being offensive should be repaired. In the PA rig feed back should really not be an issue. Unlesss....... But for Monitors, it’s almost the opposite. Yes we should start with optimum frequency response from our wedge. Start with a coherent and cohesive product. Notch the mic a little, high pass as needed for sure in the card. Then the graph comes into its own. With thousands and thousands of hours of experience the graph becomes your buddy. Not a crutch. You learn how harmonics interact and how you can shape a wedge sound to beat feedback yet try to retain to meat and potatoes. Accuracy is crucial. So of course, obviously, a parametric would be great. But apart from a Lake Contour I/O , most parametric are four to six bands max. Trust me, I had an O1V. Graphs are needed for monitors. The tutorial about the accuracy of the perceived curve with parametric eqs is great. What’s most revealing is the harmonic interactions between frequency bands that can both confound you and make you smile. Try to use a graph to RTA match a room and you will be chasing your tail for hours, but more importantly you will learn about relationships and interaction. So yeah, I guess this means you’ve got a new 31 band parametric eq to insert?
Thanks for sharing your workflow. I do rely on PEQ in the matrix bus to shape the sound of the PA to a desired curve. Once my inputs for a band are set and I am happy with the sound of the show, I tend to do my EQ mostly on the output side of the console. Also, when doing monitors, I go for the PEQ first, use up my 6 bands, and only if I need more, would I use the GEQ. But as long as the job gets done well, I always go by the saying "if it sounds good, it is good". I just wanted to show that what you see on the GEQ as a filter setting usually does not correlate to what the actual signal is. I believe that knowing how your tools work makes you decide faster and with greater accuracy. Take care, thanks for sharing.
gainmedialab absolutely and thanks for illustrating the actual versus the perceived. It can be quite frustrating to pull a little 1k out of a wedge only to have 1.25 K start to ring. Oh, sound reinforcement, how you taunt me. Endlessly.
Honestly I feel at this point we're splitting hairs, especially when it comes to live shows, this level of detail doesn't really matter. Just like guitarists obsessing over the minutia of their tone live. The crowd doesn't care.
Look I understand your motive to make these kind of comparison videos, but I would say don't waste your and other's time by making these kinds of childish stuff cause the pros who have relied on graphic eq's for years aren't gonna change their mind by seeing the analyser cause the parametric one's won't give them the results that they would get with the graphics. So yeah consider another useful topic next time. Thanks.
Hi B.D., thanks for your comment. The purpose of this video was to show what a tool we use actually does. I am not saying you should not use a graphic EQ. I am saying, however, that you should know what it does and how it affects the signal. There is just one thing in your comment I respectfully don't agree with: you CAN get the same result with the parametric that you get with a graphic, but you can't get the result from a parametric with a graphic. Hope you keep successfully using whichever tool gets the job done for a long time. Stay healthy and mix great shows.
Are you loving your GEQ or shying away from it?
Not all graphic eq's work like this. For example, the 560 from API's midband is a very wide band for broad strokes.
I would argue that this kind of graphic eq is good for certain applications, just like any tool. It's just a tool, and it's your knowledge of how to use it and what it does to your signal that helps you choose the right tool for the job at hand.
Also, not all parametric eq's can get as tight of a q as the one you're using, which again comes down to the differences between eq builds and not the difference between graphic vs parametric.
I would argue that how the result sounds is the only important factor. What you told it to do depends on what you hear while applying the tool.
Sure, that's why I feel it is important you know how your tools work by measuring them. And a 100% with you on using whatever gets the job done. Just wanted to point out that people often assume that just that one particular frequency of the GEQ is affected, when in fact it is not so.
@@gainmedialab Sure.
But there's no such thing as an eq that can notch out 1000 Hz without impacting 999Hz or 1001 Hz. You're going to be shifting frequencies around it, even with the absolute narrowest q possible.
It's good that people understand that as a fundamental of how audio processing works. But what application does this have on the ground?
I agree. I just want people to know what happens to the signal when you, for instance, try to shape an EQ curve to a target trace for your PA system and keep getting different results from what you are seeing on the GEQ set filters.
yup i agree. TruEQ is better than the regular GEQ for speaker calibration, but fully parametric eq is the best eq
Thanks Aleš! I’ve been wanting a video exactly like this. So helpful to see what’s happening with GEQ - definitely helps to understand our audio tools better.
You are welcome! Glad you liked it.
Thanks Aleš! Definitely more going on there than we might think!
Absolutely. That is why you should always measure, then measure again, just to be sure you measured correctly ;) Do you use GEQ in your work?
@@gainmedialab I've switched off of GEQ but I'll admit I still have one on my vocal buss for feedback surprises. It still feels like the fastest way to grab a frequency if I hear something getting ready to give me a problem. That's partially a GUI issue and partially a habit that needs breaking. LOL
As long as it gets the job done and you know what your tools are doing... ;)
Parametric EQ with frequency bands is all i need im Ear And Visual guy it helps so much
Makes sense
if you narrow the Q of a parametric and cut 1k by 15db, it will also pull down the neighboring frequencies. so with parametric, you have maybe six, maybe four frequency ranges you can alter, each affecting the nearest frequencies to it. with a graphic, you have so much more control. if you wreck it to the point of phasing then you have problems on your board.
With a PEQ you can define the width of the Q and be much more precise. With a GEQ you don't really have a great representation of what it is doing and how the filters affect the signal. Just trying to shed some light on the fact that the GEQ filters are not notches, but can be much more wide than people suspect.
the best representation of what a graph is doing is *your ears*.
a parametric is great on an individual channel - you certainly should be able to make any signal sound great with five points of adjustment. but the main mix for the entire house needs much more nuance than that.... in my humble opinion :)
Thanks for the informative video. I am old school sound engineer and used mainly analogue equipment because digital wasn't around at the time. I encountered on tour every night different equipment and remembered a cheap nasty Yamaha Eq on a system and just pulling a frequency I was getting feedback. It was so inaccurate. Anyway are well build graphic analogue EQs behaving similar???? Also use to work on same venue for many years using DB monitors. Always pulled maybe 2 or 3 freq out if needed at all. Nice powerful monitor. One night visiting tech brought his cheap Allen and Heath digi desk did some drastic EQ and the monitor never sounded so bad and hardly any volume. Anyway curious if analogue EQ are adding the same inaccuratecy.
Thanks again. Raymond Steeg
All GEQs have the same behaviour, but with various Q widths. My point was that you actually can't be sure until you measure, but 9 out of 10 times the neighbouring filters will affect each other - the quality of the build and the desired character of the tool will determine how much.
Indeed. I remember the same. The Klarks you could count on. XTA great . BSS, Opal, smoother (probably wider) but also great. Ashleys and Ranes were sometimes ok, sometimes not. dBx caused a bit of concern occasionally and Behringer analog graphics were a bit of a joke.
This is great advice. It should be no secret by now that the built in graphic eqs on digital desks generally use exactly the same biquad filter math as their parametric counterparts. But the parametrics allow one further control of the Q and the exact frequency you want to hit. Super narrow notching on these eqs can sometimes create as many problems as they solve, but whatever works. To say that a true freq/phase plot of a typical venue looks like a seizure in graphic form would be an understatement. It's among the top reasons why we need high q, narrow filters for live sound.
I use an EQ to cut or lower frequencies most of the time but what if we boost frequencies, does the same thing happen when you boost frequencies that are next to each other?
Sure. The change in phase is there no matter if you cut or boost.
Okay, you make a lot of good points. Now, how can you have more than 5 adjustment points on the VEQ? Why cant you get a smaller Q setting on the GEQ? After all, this is all digital and SHOULD be more customizable.
But you never mention it the impact of phase correlation of the eq's and the impact it has on the phase. In theory you can use pinpointed PEQ to cut a frequency better that the GEQ, but can you pinpoint it exactly during showtime? That is why GEQ is useful.
I tend to use PEQ when doing soundcheck, when playing safe.
I use GEQ when fixing things during show
I use TEQ on matrix busses to calibrate primary or secondary systems,
As you can see, TEQ has the least impact on phase offsetting, which i think is the most important for a clear and pristine sound, and if you found yourself eq'ed too much, the system just sounds like through curtain, and is unable to punch lungs, drop guts and bite.
Nice video though, i want to see you compare different models of digital mixers' eq's for phase correlation, do some mixers have better designed eq's, what about vs analog GEQ like the BSS maybe...
Keep up the good work.
Thanks, I would like to do that as well, but currently only have access to this console. Once we start working again, I will hopefully do the same with other consoles. And I agree with you on the phase issue, but I thought it was just too much information to cram into one video. I might do a follow up addressing just phase and the time domain of these tools.
@@gainmedialab looking forward to see new videos related to phase offset of EQ, very underestimated fact nowadays.
"PEQ to cut a frequency better that the GEQ, but can you pinpoint it exactly during showtime? That is why GEQ is useful. "
How is it faster to pinpoint the ringing frequency with a GEQ than with a PEQ ?
Ok. So for FOH PA systems, if you need much manipulation even in your parametric EQ, pack up the rig and get one that presents a balanced response for the frequency range needed for the application(talking head events don’t need 30hz). So use your parametric on the PA, but I swear if you need more than one band any kind of huge notch then there are bigger problems, regardless of the room, save for very extreme circumstances. In other words, the PA should present as flat. Any issues or corrections should be done in the corresponding channels. The moment you shape a PA around a lead vocal mic, is the moment you should just send the band home. It’s not the fault of the snare drum, or the bass, or the keyboards, that a vocal mic has to be the most compromised input as it is usually the one with the highest input level to output needed discrepancy of any mic on stage. So when you shape around a vocal or even a great track, everything else suffers. Now the keys sound hollow, the snare is all highs and lows, with no meat, the bass has disappeared, but you can feel it. Ugh. The cymbals are all glassy. All because of a vocal mic. So shape the vocal mic with the parametric. Need more? Well, first off, sorry. But there are plenty of ways to put more eq into the chain. Regardless of whether I’m doing the symphony or death metal or the mayoral inauguration, the PA eq is all but flat. The input being offensive should be repaired. In the PA rig feed back should really not be an issue. Unlesss.......
But for Monitors, it’s almost the opposite. Yes we should start with optimum frequency response from our wedge. Start with a coherent and cohesive product. Notch the mic a little, high pass as needed for sure in the card. Then the graph comes into its own. With thousands and thousands of hours of experience the graph becomes your buddy. Not a crutch. You learn how harmonics interact and how you can shape a wedge sound to beat feedback yet try to retain to meat and potatoes. Accuracy is crucial. So of course, obviously, a parametric would be great. But apart from a Lake Contour I/O , most parametric are four to six bands max. Trust me, I had an O1V. Graphs are needed for monitors. The tutorial about the accuracy of the perceived curve with parametric eqs is great. What’s most revealing is the harmonic interactions between frequency bands that can both confound you and make you smile. Try to use a graph to RTA match a room and you will be chasing your tail for hours, but more importantly you will learn about relationships and interaction. So yeah, I guess this means you’ve got a new 31 band parametric eq to insert?
Thanks for sharing your workflow. I do rely on PEQ in the matrix bus to shape the sound of the PA to a desired curve. Once my inputs for a band are set and I am happy with the sound of the show, I tend to do my EQ mostly on the output side of the console. Also, when doing monitors, I go for the PEQ first, use up my 6 bands, and only if I need more, would I use the GEQ. But as long as the job gets done well, I always go by the saying "if it sounds good, it is good". I just wanted to show that what you see on the GEQ as a filter setting usually does not correlate to what the actual signal is. I believe that knowing how your tools work makes you decide faster and with greater accuracy. Take care, thanks for sharing.
gainmedialab absolutely and thanks for illustrating the actual versus the perceived. It can be quite frustrating to pull a little 1k out of a wedge only to have 1.25 K start to ring. Oh, sound reinforcement, how you taunt me. Endlessly.
Honestly I feel at this point we're splitting hairs, especially when it comes to live shows, this level of detail doesn't really matter. Just like guitarists obsessing over the minutia of their tone live. The crowd doesn't care.
The point I was making is to know what your tools are doing. How you use them is then entirely up to you.
you cant get better music without hardware equalizer.
Look I understand your motive to make these kind of comparison videos, but I would say don't waste your and other's time by making these kinds of childish stuff cause the pros who have relied on graphic eq's for years aren't gonna change their mind by seeing the analyser cause the parametric one's won't give them the results that they would get with the graphics. So yeah consider another useful topic next time. Thanks.
Hi B.D., thanks for your comment. The purpose of this video was to show what a tool we use actually does. I am not saying you should not use a graphic EQ. I am saying, however, that you should know what it does and how it affects the signal. There is just one thing in your comment I respectfully don't agree with: you CAN get the same result with the parametric that you get with a graphic, but you can't get the result from a parametric with a graphic. Hope you keep successfully using whichever tool gets the job done for a long time. Stay healthy and mix great shows.