EQ frequency selection

chrisv · 8th August 2003

Hi fellow AE geeks,

Having built a few pre/eq's for my own little studio, I've always been interested in how respectable manufacturers select their frequencies for their EQ's.

I remember hearing from a respected local mastering engineer that in the early days of Neve the EQ's would be customised to whatever guitar amp was in fashion at the time, and certain engineers would request subtle changes to stock eq's when something new came out.

I also heard that Manley have in the past offered different frequencies for their Massive Passive but nobody has taken them up on the idea. Is this fact or fallacy?

Lastly, is sitting down with your newly designed EQ in a recording studio armed with a few different head/amp combinations and a few different mics commonplace? I would have thought it would be a given but in reality it is a very difficult thing to do (swapping caps and inductors here and there, especially when it's so subjective).

Tim, I'm not interested in your comments because I know the All-Blacks are going to win the tri nations and the world cup and for that I resent you.

(kidding)

although I suspect the NZ is going to be pretty hard to beat, particularly after our recent thrashing.

Cheers

Chris

fatty · 8th August 2003

hhhmm very interesting topic. i've wonderred about this myself. i'm toying with the idea of building a 2 channel pre - eq unit with a very simple 3 band eq section. 2 or 3 frequency choices on each band and maybe a wide - narrow Q switch. more of a tone shaping tool than a serious eq. but what frequencies would be most useful for general use??

good question. i look forward to responses from those more knowledgable folks around here.

gyraf · 8th August 2003

Well, I might not be "respectable", but our company manufactures some pro audio stuff, including EQ's.

You're right, EQ frequency selection is not all easy. The frequencies that may look right "on paper" very often seems to be less than sparkling when applied to real-life.

This is why base our design heavily on feedback from our team of "test-flight" engeineers - that on a regular basis uses prototype gear in real recording sessions.

Our range of eq's - the G6, G8, G14, as well as the eq section in the G2 recording channel - are all results of close cooperation with engeineers applying the units to real studio work - and trying to translate their feedback into a workable design.

This way of working means that you need to keep a short design cycle - when an engeineer wants to try a different set of frequencies (or any other change), the new version has to be made quickly for the comparision to make sense. Keep in mind, that it is often not possible to "extract" a piece of gear from the studio for long time while the recording session is in progress.

When it comes to deciding EQ frequencies, engeineer feedback often steers you to a completely different set of frequencies than you may have anticipated - and trying to make a compromise between different engeineer's needs is not always easy! But I know of no other way of working that will enable you to come up with designs that is as closely matched to the real-world engeineer's needs.

And yes, sometimes there is a real need for "costum" eq frequencies - like recently, when Denmarks Radio ordered a bunch of our G2 recording channels, not for music recording, but for recording dialouge on their movie productions. It turned out that the eq spectrum that was desireable for speech tweaking was quite different from the ones wished by our regular test crew - so we had to take a couple of design cycles with the movie staff.

To cut this short; in my opinion, eq frequency decisions (as well as decisions on several other "sonic" issues) can only be taken in close interaction with the intended users.

Jakob Erland
Gyraf Audio

Fatty, for a simple, yet very good sounding "minimal" eq, you should maybe try the Trident A-range monitor-side eq - it's just high and low, no choice of freq or Q, but very very useable.

hutch · 9th August 2003

For the Massive Passive we were very careful in choosing specific “magic” frequencies. For example, 47 Hz is far more useful than the more commonly used 50 Hz ……

Sorry, just teasing. I don’t believe in “magic” frequencies. I hope you don’t either, but there may be specific settings on one box that are special and we have seen that on both analog and digital EQs. These just don’t translate from one EQ to the next. Our choice of freqs was based on several considerations. We wanted 4 overlapping bands with stepped frequencies and wanted to cover the full spectrum evenly and wanted consistent step sizes.

Pre-Massivo, a friend at Waves emailed me some questions about Pultecs and rather than draw CAD curves, I simulated the curves with a Q10. They loved them and they became the basis for the Ren EQ shelves. The bells and HP & LP filters came from other unusual techniques. All these experiments were quickly achievable and could be evaluated by a bunch of beta-testers and we were getting positive feedback. So, a few months later, I had some goals to achieve with the Massive, though they were obviously different in technique and execution from the Ren, but perhaps conceptually pre-tested by the Waves beta team.

Once we sat down and modeled the curves we wanted with digital EQs, and modeled the circuits with SPICE to mimic those curves, and bread-boarded some circuits to verify the SPICE visual data, some things like frequency choices became clear. The number of frequencies depended most on switches that we wanted to use, which in our case was Grayhill gold contact sealed PCB mount types and the maximum number of positions was 11 or 12 and the 11’s have a nice 330 degree rotation. To keep Q or bandwidth consistent between steps we needed to wind custom inductors with at least 11 taps each. So we have an individual inductor tap (custom) and capacitor that determines each center frequency.

Capacitors come in certain standard values, like 1.0, 1.5, 2.2 etc with lesser availability for in-between values like 1.2, 1.8, 2.7 etc so our step size is dependent on capacitor availability or we have to double up on capacitors to get intermediate freqs. The latter was not attractive because we use nice (pricey) Wima caps and the design called for hundreds. We could have chosen any arbitrary starting point because we wind our own inductors but one ends up with many unconventional freqs given a fixed step size. What became strikingly obvious was why some of the classic Eqs used frequency choices based on standard capacitor values. It was by far the most practical solution in building a L-C based EQ. It also seemed comforting that there was precedence in that some of my favorite Eqs (like Neves) used those freqs. Now we knew why. The operative word is “practicality”. There are no universal magic frequencies. Sorry.

Did we sit in a studio with a bucket of caps and inductors? No. Methinks that might be a mistake that could result in a bunch of random freqs and Qs. It is probably more important to maintain predictable and consistent step sizes and Qs. We did take a nearly complete unit to a few engineers for comments. Did we look at other Eqs or compare the Massivo to other EQs at various points in the design? Strangely enough, not even once. We didn’t play with any other EQ at all, hardware or software, once our design began and for a few months after. The EQ software simulations happened at a time when we were just beginning to think about our EQ, and the rest came from 20 years of studio experience and our own eccentric tastes. Every day designers have to make dozens of little choices, many almost arbitrary, and a product is the net result of a few big decisions and thousands of little ones usually forced by those big decisions. What were the big decisions for the Massive Passive? Should be obvious – We wanted to do a new LC design, which had not been attempted seriously in decades and we wanted to use tubes. We hoped to get similar sounds as some classic EQs yet make it more versatile and flexible than most of those, and we were not about to copy or clone anything in the process, pure original. In terms of a target market, we thought the tracking guys were most in need of a powerful EQ, and the mastering guys needed it mostly subtle, and mixers needed some valid alternative to standard EQs, because they had those already. An EQ box for anybody with other EQs.

Back to the original question – how would a typical designer choose EQ freqs. It depends on the maximum Q. With a high Q possibility, the best answer is continuously variable frequencies, and this suggests a typical parametric so that one could notch out a specific nasty. Difficult consideration with tubes. With a low Q design (where a given center freq can be considered to be boosting or cutting octaves wide swaths), then freq choice becomes almost arbitrary plus less freq choices could be fine. With a mid Q optimized design (like the Massive) then we want a good number and good range of freqs. In this case we end up with, about 4 freqs per octave based on standard capacitor values, and 44 freqs total. However, we did spend an inordinate amount of effort to make sure we achieved specific unusual curves and dynamic characteristics that we were targeting. And we had to bend the rules a bit at the very lowest and highest freqs just to have useful shelves at those extremes. Of course, a lot of listening and tweaking is mandatory, for musical signal processors of any type, to get it right.

Have we customized the Massive for certain engineers, specifically in regards to other frequency choices? Once, a mastering engineer wanted to duplicate the Pultec EQP 14K and 16K boost sound. It was sort of silly and we don’t recommend it and he now owns a standard one. Do we offer this as a service? Possibly, but would definitely tell the guy to use a standard one first for a few months and would expect that his mind would change on that issue. We have created a new Mastering version with +/- 11 dB boost/cut ranges on detented controls for a request from Scott Hull for his new room at Hit Factory. The only frequencies changed were on the HP and LP filters and made them more appropriate for mastering by bringing them closer to the edges of the spectrum. This is now a standard product and we have sold a number of them since. Many mastering engineers use the standard non-detented version.

We do have a new EQ on the drawing board based on the Massive and specifically meant to address a few rare complaints. The complaints being price, detents, or the subtle (deliberate) tube & transformer color. So the new EQ will be 1U, 2 bands (the lowest and highest Massive bands), stereo, (detents extra $) and solid state based on a new amp circuit that we have recently designed and dubbed the “Rapture Amp”. It is cleaner than most solid state and has some of the best characteristics of class A tube circuits. Minimal cosmetics to keep the price down and much of what people like about the original down but certainly about 40% of the Massive Passive features. In other words, not another compromised budget parametric, but a Mini-Massivo to suit demanding alternative tastes and surround applications and tighter budgets. Different yes, similar yes, beyond that we’ll let you judge or choose.

hutch
PS - I won't be able to reply for a week or so, family reunion in Canada.

**Tim Farrant** · 11th August 2003

Quote:

Originally posted by chrisv
Tim, I'm not interested in your comments because I know the All-Blacks are going to win the tri nations and the world cup and for that I resent you.

(kidding)

although I suspect the NZ is going to be pretty hard to beat, particularly after our recent thrashing.

HA !!

Yes looks like the Aussies are doomed but from past performances of the All Blacks I wouldn't be too sure - they have a habit of stuffing up at the last minute !!

As far as the original post question goes, I'm not quite sure I can add anything after Hutch !!

Cheers
Tim

fatty · 11th August 2003

thanks jakob for the tip on the trident eq.
where can i find a shcematic??

gyraf · 12th August 2003

Fatty,

I have a (not very clear, but useable) scan of the A monitor eq section.

Drop me a mail at

jakoberland(at)feedback(dot)dk

and I'll send it.

Jakob E.

Jay Kahrs · 16th August 2003

Hutch, great post! So now that we know how a Pultec and other inductor based EQ's work, how does a fullly variable parametric work?

chrisv · 16th August 2003

Yes Hutch, thanks for that great post!

I love hearing no-nonsense explanations that dispell myths, especially surrounding eq's.

I kinda know the answer to Jay's question, but my explanations suck so I'll leave it to Hutch or Tim or Jakob.

Alternatively, check out Steve Dove's articles on designing a professional mixing console. That guy's explanations are the bomb!

Chris

**Tim Farrant** · 20th August 2003

Quote:

Hutch, great post! So now that we know how a Pultec and other inductor based EQ's work, how does a fullly variable parametric work?

The most common type of swept parametric circuit would be the State Variable Filter (SVF). This is a very interesting circuit that roughly resembles an oscillator but is slugged enough to not oscillate. The SVF has three outputs, a high pass, low pass and bandpass. The bandpass output is used for a sweep type eq circuit.

The UREI 545 is a classic example of the SVF used for audio EQ.

There are other types of circuits such as the CAPS circuit that Steve Dove wrote about in his Studio Sound articles. Does anybody know of an equaliser built using the CAPS circuit ??

Cheers
Tim

Jay Kahrs · 26th August 2003

Quote:

Originally posted by The Buzz
This is a very interesting circuit that roughly resembles an oscillator but is slugged enough to not oscillate.

What does "slugged" mean? In stupid AE, muso terms please.

Thanks,

**Tim Farrant** · 26th August 2003

Quote:

Originally posted by Jay Kahrs
What does "slugged" mean? In stupid AE, muso terms please.

Thanks,

"Slugged" is another term for "damped" I suppose. Imagine a wet blanket on a fire, it smoulders but does not burn.

Tim

8th August 2003	#1
chrisv Gear interested Joined: Oct 2002 Location: Sydney Posts: 23 Thread Starter	EQ frequency selection Hi fellow AE geeks, Having built a few pre/eq's for my own little studio, I've always been interested in how respectable manufacturers select their frequencies for their EQ's. I remember hearing from a respected local mastering engineer that in the early days of Neve the EQ's would be customised to whatever guitar amp was in fashion at the time, and certain engineers would request subtle changes to stock eq's when something new came out. I also heard that Manley have in the past offered different frequencies for their Massive Passive but nobody has taken them up on the idea. Is this fact or fallacy? Lastly, is sitting down with your newly designed EQ in a recording studio armed with a few different head/amp combinations and a few different mics commonplace? I would have thought it would be a given but in reality it is a very difficult thing to do (swapping caps and inductors here and there, especially when it's so subjective). Tim, I'm not interested in your comments because I know the All-Blacks are going to win the tri nations and the world cup and for that I resent you. (kidding) although I suspect the NZ is going to be pretty hard to beat, particularly after our recent thrashing. Cheers Chris __________________ "Fashions change, the laws of physics don't." Steve Dove

9th August 2003	#4
hutch Gear interested Joined: Nov 2002 Location: California Posts: 9	For the Massive Passive we were very careful in choosing specific “magic” frequencies. For example, 47 Hz is far more useful than the more commonly used 50 Hz …… Sorry, just teasing. I don’t believe in “magic” frequencies. I hope you don’t either, but there may be specific settings on one box that are special and we have seen that on both analog and digital EQs. These just don’t translate from one EQ to the next. Our choice of freqs was based on several considerations. We wanted 4 overlapping bands with stepped frequencies and wanted to cover the full spectrum evenly and wanted consistent step sizes. Pre-Massivo, a friend at Waves emailed me some questions about Pultecs and rather than draw CAD curves, I simulated the curves with a Q10. They loved them and they became the basis for the Ren EQ shelves. The bells and HP & LP filters came from other unusual techniques. All these experiments were quickly achievable and could be evaluated by a bunch of beta-testers and we were getting positive feedback. So, a few months later, I had some goals to achieve with the Massive, though they were obviously different in technique and execution from the Ren, but perhaps conceptually pre-tested by the Waves beta team. Once we sat down and modeled the curves we wanted with digital EQs, and modeled the circuits with SPICE to mimic those curves, and bread-boarded some circuits to verify the SPICE visual data, some things like frequency choices became clear. The number of frequencies depended most on switches that we wanted to use, which in our case was Grayhill gold contact sealed PCB mount types and the maximum number of positions was 11 or 12 and the 11’s have a nice 330 degree rotation. To keep Q or bandwidth consistent between steps we needed to wind custom inductors with at least 11 taps each. So we have an individual inductor tap (custom) and capacitor that determines each center frequency. Capacitors come in certain standard values, like 1.0, 1.5, 2.2 etc with lesser availability for in-between values like 1.2, 1.8, 2.7 etc so our step size is dependent on capacitor availability or we have to double up on capacitors to get intermediate freqs. The latter was not attractive because we use nice (pricey) Wima caps and the design called for hundreds. We could have chosen any arbitrary starting point because we wind our own inductors but one ends up with many unconventional freqs given a fixed step size. What became strikingly obvious was why some of the classic Eqs used frequency choices based on standard capacitor values. It was by far the most practical solution in building a L-C based EQ. It also seemed comforting that there was precedence in that some of my favorite Eqs (like Neves) used those freqs. Now we knew why. The operative word is “practicality”. There are no universal magic frequencies. Sorry. Did we sit in a studio with a bucket of caps and inductors? No. Methinks that might be a mistake that could result in a bunch of random freqs and Qs. It is probably more important to maintain predictable and consistent step sizes and Qs. We did take a nearly complete unit to a few engineers for comments. Did we look at other Eqs or compare the Massivo to other EQs at various points in the design? Strangely enough, not even once. We didn’t play with any other EQ at all, hardware or software, once our design began and for a few months after. The EQ software simulations happened at a time when we were just beginning to think about our EQ, and the rest came from 20 years of studio experience and our own eccentric tastes. Every day designers have to make dozens of little choices, many almost arbitrary, and a product is the net result of a few big decisions and thousands of little ones usually forced by those big decisions. What were the big decisions for the Massive Passive? Should be obvious – We wanted to do a new LC design, which had not been attempted seriously in decades and we wanted to use tubes. We hoped to get similar sounds as some classic EQs yet make it more versatile and flexible than most of those, and we were not about to copy or clone anything in the process, pure original. In terms of a target market, we thought the tracking guys were most in need of a powerful EQ, and the mastering guys needed it mostly subtle, and mixers needed some valid alternative to standard EQs, because they had those already. An EQ box for anybody with other EQs. Back to the original question – how would a typical designer choose EQ freqs. It depends on the maximum Q. With a high Q possibility, the best answer is continuously variable frequencies, and this suggests a typical parametric so that one could notch out a specific nasty. Difficult consideration with tubes. With a low Q design (where a given center freq can be considered to be boosting or cutting octaves wide swaths), then freq choice becomes almost arbitrary plus less freq choices could be fine. With a mid Q optimized design (like the Massive) then we want a good number and good range of freqs. In this case we end up with, about 4 freqs per octave based on standard capacitor values, and 44 freqs total. However, we did spend an inordinate amount of effort to make sure we achieved specific unusual curves and dynamic characteristics that we were targeting. And we had to bend the rules a bit at the very lowest and highest freqs just to have useful shelves at those extremes. Of course, a lot of listening and tweaking is mandatory, for musical signal processors of any type, to get it right. Have we customized the Massive for certain engineers, specifically in regards to other frequency choices? Once, a mastering engineer wanted to duplicate the Pultec EQP 14K and 16K boost sound. It was sort of silly and we don’t recommend it and he now owns a standard one. Do we offer this as a service? Possibly, but would definitely tell the guy to use a standard one first for a few months and would expect that his mind would change on that issue. We have created a new Mastering version with +/- 11 dB boost/cut ranges on detented controls for a request from Scott Hull for his new room at Hit Factory. The only frequencies changed were on the HP and LP filters and made them more appropriate for mastering by bringing them closer to the edges of the spectrum. This is now a standard product and we have sold a number of them since. Many mastering engineers use the standard non-detented version. We do have a new EQ on the drawing board based on the Massive and specifically meant to address a few rare complaints. The complaints being price, detents, or the subtle (deliberate) tube & transformer color. So the new EQ will be 1U, 2 bands (the lowest and highest Massive bands), stereo, (detents extra $) and solid state based on a new amp circuit that we have recently designed and dubbed the “Rapture Amp”. It is cleaner than most solid state and has some of the best characteristics of class A tube circuits. Minimal cosmetics to keep the price down and much of what people like about the original down but certainly about 40% of the Massive Passive features. In other words, not another compromised budget parametric, but a Mini-Massivo to suit demanding alternative tastes and surround applications and tighter budgets. Different yes, similar yes, beyond that we’ll let you judge or choose. hutch PS - I won't be able to reply for a week or so, family reunion in Canada. __________________ Hutch Chief designer, Manley Labs Manley Labs

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread starter	Forum	Replies	Last Post
EQ Frequency Chart.	Chrizcol	So much gear, so little time!	56	27th November 2011 01:48 PM
Frequency Selective Side Chain Compressor	swinxx	So much gear, so little time!	6	27th November 2011 10:10 AM
plug-in development suggestion -> frequency selective with trigger function	robomix	So much gear, so little time!	3	28th October 2011 03:19 AM
EQ with Selectable High Pass?	Lunatic	Low End Theory	9	21st February 2007 06:05 AM

8th August 2003	#2
fatty Gear addict Joined: Jul 2002 Location: Melbourne Posts: 348	hhhmm very interesting topic. i've wonderred about this myself. i'm toying with the idea of building a 2 channel pre - eq unit with a very simple 3 band eq section. 2 or 3 frequency choices on each band and maybe a wide - narrow Q switch. more of a tone shaping tool than a serious eq. but what frequencies would be most useful for general use?? good question. i look forward to responses from those more knowledgable folks around here.

8th August 2003	#3
gyraf Lives for gear Joined: Jun 2002 Location: Denmark Posts: 667	Well, I might not be "respectable", but our company manufactures some pro audio stuff, including EQ's. You're right, EQ frequency selection is not all easy. The frequencies that may look right "on paper" very often seems to be less than sparkling when applied to real-life. This is why base our design heavily on feedback from our team of "test-flight" engeineers - that on a regular basis uses prototype gear in real recording sessions. Our range of eq's - the G6, G8, G14, as well as the eq section in the G2 recording channel - are all results of close cooperation with engeineers applying the units to real studio work - and trying to translate their feedback into a workable design. This way of working means that you need to keep a short design cycle - when an engeineer wants to try a different set of frequencies (or any other change), the new version has to be made quickly for the comparision to make sense. Keep in mind, that it is often not possible to "extract" a piece of gear from the studio for long time while the recording session is in progress. When it comes to deciding EQ frequencies, engeineer feedback often steers you to a completely different set of frequencies than you may have anticipated - and trying to make a compromise between different engeineer's needs is not always easy! But I know of no other way of working that will enable you to come up with designs that is as closely matched to the real-world engeineer's needs. And yes, sometimes there is a real need for "costum" eq frequencies - like recently, when Denmarks Radio ordered a bunch of our G2 recording channels, not for music recording, but for recording dialouge on their movie productions. It turned out that the eq spectrum that was desireable for speech tweaking was quite different from the ones wished by our regular test crew - so we had to take a couple of design cycles with the movie staff. To cut this short; in my opinion, eq frequency decisions (as well as decisions on several other "sonic" issues) can only be taken in close interaction with the intended users. Jakob Erland Gyraf Audio Fatty, for a simple, yet very good sounding "minimal" eq, you should maybe try the Trident A-range monitor-side eq - it's just high and low, no choice of freq or Q, but very very useable.

11th August 2003	#6
fatty Gear addict Joined: Jul 2002 Location: Melbourne Posts: 348	thanks jakob for the tip on the trident eq. where can i find a shcematic??

12th August 2003	#7
gyraf Lives for gear Joined: Jun 2002 Location: Denmark Posts: 667	Fatty, I have a (not very clear, but useable) scan of the A monitor eq section. Drop me a mail at jakoberland(at)feedback(dot)dk and I'll send it. Jakob E.

16th August 2003	#9
chrisv Gear interested Joined: Oct 2002 Location: Sydney Posts: 23 Thread Starter	Yes Hutch, thanks for that great post! I love hearing no-nonsense explanations that dispell myths, especially surrounding eq's. I kinda know the answer to Jay's question, but my explanations suck so I'll leave it to Hutch or Tim or Jakob. Alternatively, check out Steve Dove's articles on designing a professional mixing console. That guy's explanations are the bomb! Chris