1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  

Onboard filter preamp questions about Q (with pictures!)

Discussion in 'Pickups & Electronics [BG]' started by BunchyMutt, May 23, 2016.


  1. BunchyMutt

    BunchyMutt Supporting Member

    Jul 3, 2014
    Nashville, TN
    I was editing some compositions in GarageBand last night and stumbled on some flexibility to the eq that I hadn't known about: that the high/low pass filter slopes are adjustable by -___db/octave AND Q. Pretty cool. As I was playing with it, I had some questions and potentially self-discovered answers about filter based preamps like ACG/Wal/Alembic.

    First, when reading about HPFs like @fdeck's or @Azure Skies's Broughton, stats are usually given in db/octave, but these pictures are all at -48db/octave but just changing the Q value:

    Screen Shot 2016-05-23 at 8.27.56 AM.png
    Screen Shot 2016-05-23 at 8.28.17 AM.png
    Screen Shot 2016-05-23 at 8.28.38 AM.png

    So, first question - Why isn't Q written about more? Am I missing something? It seems like a powerful tool to have adjustable.


    Then, the interesting part is when Q got high enough that it started boosting frequencies, like so:
    Screen Shot 2016-05-23 at 8.29.08 AM.png
    Screen Shot 2016-05-23 at 8.29.34 AM.png
    Screen Shot 2016-05-23 at 8.29.54 AM.png
    You can see in the last one that when Q=100, it notches out.
    Second question: Is this what’s called an overshoot peak? Is that the same thing as resonance? @Skelf?


    Then I started thinking about applications of this.

    Third question: It'd be really useful to have a HPF and LPF with adjustable slope and Q values to go right before a power amp. Anything like this exist already?

    The above application doesn't really need a Q value higher than ~.75 (or whenever frequencies start getting boosted), but I could see an onboard preamp with that capability being really, really, really useful. Much more useful than bass and treble shelving controls. Now I'm thinking how great it'd be to have an onboard preamp that included:
    • volume
    • blend
    • a LPF at a fixed cut value like -24db/octave, and that goes from some value like 10KHz to 5KHz for example
    • Two semiparametric mid controls, high and low, that control boos/cut and center frequencies
    • a HPF at -24db/octave with a movable corner frequency like the LPF, AND an adjustable Q control that could boost frequencies)
    Essentially, something that could make this possible:
    Screen Shot 2016-05-23 at 10.59.39 AM.png

    Fourth and last question: any preamp like this exist?

    I feel like I'm maybe discovering something that everyone was talking about all along, but if that's the case, then maybe another TBer would also appreciate the visual representation of these.

    Thanks all q-who1.jpg
     
  2. Azure Skies

    Azure Skies Commercial User

    Mar 21, 2012
    Toronto
    Businessman and EE Importer/Exporter Broughton Audio
    A state variable filter can give you independent control over Q (resonance) and cutoff frequency for either LPF or HPF. The slope is fixed when it comes to analog, and can only be changed by switching stages in or out. A digital implementation is much easier to have that much control. The Q for a "standard" filter found on a mixer or like on my pedals use a Butterworth response (Q of 0.7) for a good trade-off between slope and overshoot. I believe the Wal pres use a state variable filter. I believe the Alembics use a standard single opamp configuration with an adjustable gain that affects the Q for each filter.
     
  3. Azure Skies

    Azure Skies Commercial User

    Mar 21, 2012
    Toronto
    Businessman and EE Importer/Exporter Broughton Audio
    Oh, and nice pic of "Q" :thumbsup:
     
  4. There are essentially three types of equalizers. Graphic EQs have only an amplitude control. Semi-parametric EQs have amplitude and frequency controls. The third kind of EQ is the fully-parametric EQ, which has amplitude, frequency, and bandwidth (Q) controls.

    On basses, you typically just have graphic EQ, but sometimes the mid control will be semi-parametric. A fully parametric mid control is very rare, because that level of control simply isn't needed on a bass. Fully parametric EQs are typically outboard components, with more than just one band. They have specific applications, and you typically want them to be of a higher quality than what you would find on a bass.
     
  5. BunchyMutt

    BunchyMutt Supporting Member

    Jul 3, 2014
    Nashville, TN
    I hope this doesn't sound rude, but I feel like either I'm missing your point or you're assuming I don't know about EQs to the level that I feel was necessary to write my post, so I'm feeling like I'm not following you. Which of my questions were you responding to? One reason why I started on this question is that the high and low pass filters on my Rane PE-17 are only -3db/octave, which I don't find steep enough.
     
  6. BunchyMutt

    BunchyMutt Supporting Member

    Jul 3, 2014
    Nashville, TN
    Many thanks for this answer. Great info about the slope being fixed and contingent on components as well as Q being adjustable through a couple different methods (state variable vs gain increase into op amps, if I'm reading that correctly).

    I'm confused about the Q of 0.7 being linked with overshoot though. Does overshoot mean when frequencies start getting boosted? If so, on the graphs I included, a Q of 0.7 doesn't start boosting frequencies, so how could there be overshoot?

    Thanks! Of all instruments, Q always seemed like he'd be a really cool, shade wearing bassist.
     
  7. Azure Skies

    Azure Skies Commercial User

    Mar 21, 2012
    Toronto
    Businessman and EE Importer/Exporter Broughton Audio
    The slope isn't -3dB / octave, it's -12. You are mixing that figure up with the -3dB point, which is the usual convention for determining the cutoff frequency. -12 dB / octave is a standard slope for a variable filter.

    Overshoot means the -3dB point is no longer -3dB, meaning you "overshot" your mark. Yes, eventually you would get a boost at the cutoff point. Or in the other direction, you would have a sub bessel curve, which starts cutting a bit earlier.
     
  8. BunchyMutt

    BunchyMutt Supporting Member

    Jul 3, 2014
    Nashville, TN
    Thanks for clarifying about the Rane filters + filters in general. I've started going down a rabbit hole of learning more about overshoot, ringing, Linkwitz/Riley, Bessel, Butterworth, Cheybyscheff, Cheybyscheff.....yeesh. Love it.
     
  9. Azure Skies

    Azure Skies Commercial User

    Mar 21, 2012
    Toronto
    Businessman and EE Importer/Exporter Broughton Audio
    Yeah, filters are excellent tone shapers. I prefer them to a conventional EQ. If you want to stay analog, state variable filters will be your best bet to keep things relatively simple while having plenty of control. Feel free to PM me if you have any other questions.
     
  10. BunchyMutt

    BunchyMutt Supporting Member

    Jul 3, 2014
    Nashville, TN
    Absolutely.
     
  11. BunchyMutt

    BunchyMutt Supporting Member

    Jul 3, 2014
    Nashville, TN
    @Skelf as I'm thinking through this, I'm realizing that I'm equating Q values above ~0.7 as boosting, just because that's what I'm seeing happen from the graph, but I have no clue if changing the Q value on an onboard HPF preamp is the same as simultaneous cutting frequencies below a certain point while also boosting a range above that. I could see that if it's just Q, it could get unmusical with a peak. Is this what happens with your "overshoot peak"?
     
  12. khutch

    khutch Praise Harp

    Aug 20, 2011
    suburban Chicago
    It is a powerful tool but it goes by several names so often when people are talking about it you won't recognize it. In fact the use of Q to describe the parameter you are playing with is somewhat non-standard. The hard core engineering term would be damping ratio or factor. Whatever name you use for it the parameter describes how quickly a filter transitions from its pass (flat) band to its reject (rolled off) band. Overdamped/low Q filters transition slowly so their "corners" are rounded. A critically damped filter has the quickest transition (sharpest corner) without any peaking. Underdamped/high Q filters have the response peaks that you note.

    Yes, pretty much. Overshoot is a term normally used to describe the time domain response of a filter but underdamped filters overshoot in the time domain and peak in the frequency domain. This is one way in which an electrical resonance can manifest itself but not the only way.

    I believe at least some Alembic filters have adjustable damping as do Alembic imitators. The humble passive tone control usually does this when rolled all the way down. Tonestyler tone controls seem to be built to be critically damped so that there is no peaking but DIY switched cap tone controls can be built to have critical damping or any desired degree of underdamping/peaking. In their intermediate settings passive tone controls tend to be overdamped which causes some loss of mids which is what Tonestyler/Stellartone is marketing their product as a solution for. State variable active filters can easily have adjustable damping but other active filter topologies can provide that too. A year or two ago some of us here in this very forum played around with Sallen-Key designs for low pass filters with peaks. I don't know if any of us actually made one save for myself whom I know did not! I do however have an underdamped switched cap tone control in one of my basses and I find a modest amount of peaking to be useful at times.

    Personally I have found little use for adjustable slopes and that must be a common opinion since I am not aware of any bass preamps that offer this. Maybe Alembics do and I just never noticed because I find it uninteresting. High slope filters are more useful in applications where you are trying to protect something from energy at harmful (to it) frequencies. I'm not sure there is a musical reason to need a filter with a steeper than second order response. The variable damping notion is much more useful, in my opinion.

    Yeah, see above.

    I don't think these subtleties are widely appreciated on TB because they tend to be over the heads of most musicians and the people who do make tone controls like this have buried the facts under impenetrable layers of marketing speak. Some people who have tried underdamped/peaked low pass filters hate them, perhaps because they had too much peak, others like them. I'm not sure I see any need for a peaked HPF.
     
  13. I am going to interject but not as a bass player but as a sound engineer.
    Yes those questions are critical especially when your bass needs to fit in a mix.
    A few reminders:
    The wider the Q the less you want to add or substract some eq. In other hand you can start by sweeping frequencies with a excessively boosted eq with a narrow Q in order to identify the bad sounding frequencies. Once you identify them, you create a notch to get rid of them
    Every knowlegeable engineer with use a Hi-pass filter with a steep slope on the bass when mixing. Like a 48db/octave filter at let say at 40Hz. That way the bass sounds punchy and tight. I will also use a low pass filter but the frequency here is genre dependant.
    An engineer will use filter hi-pass filter on almost every track.


    Let me add some useful info for your knowledge.
    In a band context and a mix, matching the kick and bass frequencies is important as they work together as a unit. If the drummer uses a large kick (24-26") with a lot of lows it is better to use a bright and punchy bass sound. If the drummer uses a smaller kick (20-22") with more of a punchy tone, then I would use a rounder and fatter bass sound.
    Sometimes engineers use what is called side-chain compression where everytime the kick hits, a few dbs of compression will slightly lower the volume of the bass at the same time. It helps after mixing during the mastering process to reduce the low frequency energy going into the limiter. As the result, the mastering engineer can make the overall music being perceived as louder.
     
    Last edited: May 26, 2016
  14. John East

    John East Commercial User

    Jan 10, 2002
    Oxford UK
    Owner of East Tone UK Ltd
    Since you pointed a question to Skelf, Alan - ACG, he suggested that I should help with answers, since I designed his electronics.

    > So, first question - Why isn't Q written about more? Am I missing something? It seems like a powerful tool to have adjustable.

    As already pointed out, the Q values in conventional analogue on-board bass EQs (none that are I know of) isn't variable due to the circuit topologies used, would require additional knobs of course.

    Many of the higher end analogue studio EQs do have variable Q. It can be achieved in various ways such as with the state variable filter, and the use of positive feedback with other types of filters. Q is sometimes called width or bandwidth. This makes sense with mid EQ which is giving a peak/dip response. The top of the peak stays put according to the boost/cut knob, and the Q variation determines how wide the spectrum of audio is, that is under control. The higher the Q value, the sharper the point = lesser amounts of the spectrum.

    The ACG filter style bass EQs do have variable Q but called overshoot peak for easier understanding since with this type of design, increasing the Q starts to create a resonant peak. Much like the graphs illustrate, the filter curves go from a gentle roll-over to where they increase the gain & resonance at the filter frequency. The top of the point goes up to a max 10dBs above the flat response line in the ACG.

    > Second question: Is this what’s called an overshoot peak? Is that the same thing as resonance?

    Yes, it is effectively a resonant peak.

    > Third question: It'd be really useful to have a HPF and LPF with adjustable slope and Q values to go right before a power amp. Anything like this exist already?

    Alembic SF-2, dual state variable filter rack unit might do what you want but only goes up to around 6kHz I believe. (Interestingly I designed an amp in the mid 70s with 9 band graphic with HPF & LPF filters fixed at 12dB/Octave, see here Zoot Horn )


    > The above application doesn't really need a Q value higher than ~.75 (or whenever frequencies start getting boosted), but I could see an on-board preamp with that capability being really, really, really useful. Much more useful than bass and treble shelving controls. Now I'm thinking how great it'd be to have an onboard preamp that included:
    • volume
    • blend
    • a LPF at a fixed cut value like -24db/octave, and that goes from some value like 10KHz to 5KHz for example
    • Two semiparametric mid controls, high and low, that control boos/cut and center frequencies
    • a HPF at -24db/octave with a movable corner frequency like the LPF, AND an adjustable Q control that could boost frequencies)
    > Fourth and last question: any preamp like this exist?

    Maybe a mix of the SF-2 and something else!
     
    Last edited: May 28, 2016
    GentProvocateur and deepestend like this.
  15. Primary

    Primary TB Assistant

    Here are some related products that TB members are talking about. Clicking on a product will take you to TB’s partner, Primary, where you can find links to TB discussions about these products.

     
    Mar 4, 2021

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.