Room and Sub EQ 101: How to use parametric EQ to flatten your bass

Introduction

This main focus of this blog article is to show you how easy it is to EQ your room or sub using parametric EQ.

First we look briefly at the typical bass issues in our listening rooms and home theaters and reveal the Acoustic Frontiers targets for good bass. Next we discuss some common solutions to dealing with bass issues and highlight the importance and usefulness of having a parametric EQ solution in your acoustic toolbox. We explain what a parametric EQ is and then go into a couple of tutorials on using XTZ Room Analyzer II Pro and the Xilica DSP processors to EQ your room or sub.

 

An overview of bass issues and targets

Below the room’s transition frequency room modes and speaker boundary interference can cause severe peaks and dips in the frequency response and ringing in the time domain. These acoustic distortions are audible as one-note bass, boominess, poor dynamics and lack of articulation.

Frequency response measurement taken using XTZ Room Analyzer II Pro from a real customer’s room. Note the 25dB (!) bass swing.

 

From our Acoustical Standards for Two Channel Listening Rooms White Paper we have the following low frequency targets:

  • Bass response within +/-5dB at 1/3rd octave resolution and +/-10dB at 1/24th resolution
  • Bass decay time* less than 0.45s below 35Hz and less than 0.35s elsewhere (* defined as time for sound pressure level (SPL) to decay by 40dB)

The White Paper considered the difficulty of achieving flat(ish) bass with only two speakers, no subwoofers and no equalization. For home theaters, the frequency response target can be tightened significantly, as subs are de riguer and EQ is frequently used. We haven’t formalized our standards yet, but let’s say +/- 1.5dB at 1/3rd octave and +/-2.5dB at 1/24th resolution.

Suffice to say most people don’t have bass that meets our targets and EQ is one option that can help.

 

Options to control your bass

To get bass to meet targets below 300Hz there are only a limited number of ‘controls’ that can be used:

  • Positioning – speaker, sub, listener
  • Acoustic treatment
  • Multiple subwoofers
  • EQ

Every room needs some level of acoustic treatment to deal with the 100-300Hz issues. We don’t typically run our subs up this high so multi-sub approaches have no benefit here. The application of EQ to deal with bass issues above around 100Hz also starts to get difficult as the wavelengths of sound are much smaller than in the low bass* and so the frequency response changes more rapidly over shorter distances in the room. * e.g. at 20Hz the wavelength is 53ft, at 100Hz it is 11ft, at 300Hz it is 3.8ft.

Many of the room’s worst issues are from 20-100Hz, in the so called ‘sparsely populated modal region’, where room modes are few and far between and so all the more audible. Acoustic treatment effective in this area is expensive and large. Nearly all commercial off-the-shelf products lose effectiveness at 70Hz or so, regardless of what acoustic treatment manufacturers may report. The only real products effective down this low are panel or Helmholtz resonators such as the RPG Modex Plate. Custom acoustic treatment is effective, as is using the wall structure as a giant absorber, but those are topics for another day!

That’s why home theater enthusiasts are turning to room mode canceling arrangements of multiple subwoofers. The results of using multiple subs can be spectacular – near flat frequency response, little modal ringing and low seat-to-seat bass quality variability in multi-seat theaters. There are caveats though – multiple subs can add up cost wise, take up valuable floor real estate and have poor aesthetics. Even then there are often room issues that remain untouched. One good example is if all your subs are on the floor then there is no height mode cancellation going on, so you will still see height mode related peaks in the frequency response.

That leaves equalization. EQ, if properly applied, can further flatten the frequency response and reduce modal ringing compared to the results of using acoustic treatment and/or multiple subs. It is also a great substitute for either of these solutions, particularly when there is only one listening position that matters. The system can be EQ’d flat at that position without care or regard to what is happening elsewhere in the room. If you are solitary listener (as many of us are) and you have two alternative positions in which you listen to music then you can set up multiple EQ profiles, each with a set of EQ curves customized to each position.

We at Acoustic Frontiers nearly always recommend that systems include some form of EQ capability to properly and most cost effectively deal with room issues in the low bass below 100Hz.

 

We’re talkin’ about Parametric EQ

In this article we are going to discuss parametric EQ, as in our experience it is still the best way to correct low frequency issues. Properly set up parametric EQ equals the frequency response of the best automated systems such as Audyssey XT32 and beats them in the time domain, with reduced modal ringing.

We have not measured all the automated systems out there but we are planning to investigate most if not all of the big players this year and write up our results on this blog. Trinnov, DSPeaker and Dirac Live are on the list.

A parametric EQ has three parameters:

  • Frequency, in Hz
  • Gain, in dB (can be + or -)
  • Q, which determines how wide or narrow the filter is. A high Q filter is narrow, a low Q filter is wide.

JL Audio have a nice little simulator, which while it is for car audio and only shows boosts, not cuts, gives a good overview of how parametric EQ works.

Equipment for setting up Parametric EQ

You need a couple of things to EQ your system…

1) An acoustic measurement system capable of generating parametric EQ values such as XTZ’s Room Analyzer II Pro (the Std version does not have this capability) or Room EQ Wizard

 

2) A way of implementing the parametric EQ. There are a couple of ways this can be done:

  • In an AV receiver or pre-processor. Rotel and Classe units have onboard parametric EQ. We have sold a lot of XTZs to people with these machines. I have heard some Yamaha and Pioneer models have this capability too.
  • In a room correction processor such as a DEQX
  • In music playback software on a PC or Mac music server such as JRiver and Amarra
  • In a standalone DSP such as a Xilica or MiniDSP. A standalone DSP typically has the advantage of being able to handle complex multi-sub setups with different delay and/or EQ applied to each sub.

 

The XX Tutorial: XTZ and Xilica – Part 1

We’re going to show you how easy it is to set up parametric EQ and the difference it can make to your system. For this tutorial we are using the XTZ Room Analyzer Pro II and the Xilica XP DSP box. The first example is applying a single band of EQ to deal with a room mode resonance in the right speaker.

Connect your XTZ Room Analyzer to your computer and right channel input to your amplifier. Once you’ve done that open the XTZ software and go to the ‘Room Analyzer‘ tab. Using the SPL meter function set to ‘C-Weighting‘ and ‘Slow-1s‘ set the volume control on your system so that the ‘LCS‘ (Level with C-weighting and Slow averaging) reads 75dB or thereabouts. This ensures that there is sufficient signal to noise ratio to get a good measurement. Now press the ‘Measure‘ button on the top right to take a measurement.

Screenshot from XTZ Room Analyzer II Standard showing how to set SPL levels and take a measurement

Now you should see a frequency response such as the one below. We’ve highlighted the peak we’re going to EQ which is at 72HZ, and has also been identified by XTZ as a Room Mode.

Here’s a better look at the Found Room Modes parametric EQ values:

In addition to the peak on the frequency response there is also ringing in the time domain. An easy way to spot room modes is that they take longer to decay than frequencies around them. To see the time domain or spectrogram graph in XTZ you double click on the top right window containing the spectrogram. This is what it looks like, with the 72Hz mode circled:

To deal with the frequency response bump and time domain ringing we set up a parametric EQ filter in our Xilica processor. The Xilica can be controlled by the XConsole software on a PC or Mac. This makes changing filter values easy and quick. Here is the screen with our EQ filter circled:

So with our EQ values in the Xilica let’s re-measure the room with the XTZ. Here’s the frequency response before vs. after. You can see that we’ve removed that bump at 72Hz:

And if we look at the time domain, we’ll see that the ringing at 72Hz has gone:

Part 1 of this tutorial has conclusively shown the ease of taking measurements and generating EQ filters with the XTZ Room Analyzer II Pro and implementing them on the Xilica XP series DSP processors. Onto Part 2!

The XX Tutorial: XTZ and Xilica – Part 2

Part 2 of this tutorial will show you how to use some of the more advanced functionality of XTZ Room Analyzer II Pro including the parametric EQ simulation mode which simulates EQ within the measurement software itself.

Let’s take a different measurement this time. This one was for a subwoofer; XTZ doesn’t suggest any EQ:

Now for some magic! Right click on your measurement and you’ll see some yellow lines appear. Now you can drag an area of the graph to reshape it:

Once you’ve got it looking the way you want to let go. XTZ will now tell you the parametric EQ values to apply to get the graph looking the way you wanted it. Note the value in the column Type has changed from Mode to EQ.

To get this EQ filter to be part of XTZ’s software based parametric EQ simulator press the Copy Room Modes button. You’ll see the EQ has now appeared in the Stimulus EQ table:

Now let’s re-run the measurement with the Stimulus EQ filter in the loop:

Whoops! You can see in the measurement that too much energy has been taken out by the simulated EQ filter. Luckily we can alter the values of Stimulus EQ by double clicking on the row in the table. The following dialog box comes up which allows you to modify the frequency, gain and Q values:

We’ve modified the gain down to -6dB from just over -9dB. Let’s rerun the measurement again with the Stimulus EQ in place and see what we get:

Well that’s much better! This second tutorial has shown you how to use XTZ Room Analyzer’s more advanced features such as Stimulus EQ.

Conclusion

Hopefully this article has show you how easy it is to EQ your room or sub using parametric EQ using the XTZ Room Analyzer II Pro and the Xilica DSP processors. You can buy both from our website, and if you have any questions on these products or you prefer professional consulting assistance for your room or sub EQ please contact us!

16 thoughts on “Room and Sub EQ 101: How to use parametric EQ to flatten your bass”

  1. Hi Nyal thank you

    With the dips and peaks is there a target for flatness e.g. dips 5db?

    When i measure @ like you suggested 85db the program appears to cut off then drops back to 70-75db, it actually use to be worse than this. IThe highest i can get on the graph is about 45 db-spl, is this enough to measure with?

    Thanks again

  2. Hi Nyal thank you

    With the dips and peaks is there a target for flatness e.g. dips 5db?

    When i measure @ like you suggested 85db the program appears to cut off then drops back to 70-75db, it actually use to be worse than this. IThe highest i can get on the graph is about 45 db-spl, is this enough to measure with?

    Thanks again

  3. Hi Nyal thank you

    With the dips and peaks is there a target for flatness e.g. dips 5db?

    When i measure @ like you suggested 85db the program appears to cut off then drops back to 70-75db, it actually use to be worse than this. IThe highest i can get on the graph is about 45 db-spl, is this enough to measure with?

    Thanks again

  4. Hi Nyal great tutorial.

    When setting up dual mono subs is it best to measure & eq the subs (mono) from 15-125hz (subs are Velodyne Digital Drive) or only to the crossover to be used by the receiver?

    Also Is it best to measure and eq with both L&R mains and subs together or separate?

    Will you be including the Velodyne DD+ eq in with your test?

    Thank you

  5. Thanks for this interesting post and looking forward to your article about Trinnov and Dirac Live as they both correct in the time domain. I am using Dirac Live for room/speaker eq in a two channel setup with very satisfying results, it delivers as promised and that is quite unique. The only con is the loss of gain which forced me to buy a more powerful amp :(.

  6. Hi Peter

    Thanks for reading and leaving a question!

    I would measure and EQ my subs to at least half if not a full octave above their crossover frequency. Above the crossover point the subs should follow a nice smooth slope. Any bumps or peaks in that or delayed rolloff indicates some room mode issues that will need EQing out to prevent sub localization and other bass issues.

    As to your second question…it does depend somewhat on how your bass management is set up. Assuming you are running your main speakers as small then I would likely look at the L, R and Sub as separate entities half to a full octave above the XO point. Around the XO point then I would look at the combined L/R/Sub measurement. Doing it that way is more in line with how your system is actually playing back bass frequencies.

    I hadn't thought about including the DD+ EQ but I can do that, as I am a Velodyne dealer.

    Thanks Nyal

  7. Hi Scott

    It should be fun to measure and listen to all these room correction algorithms! I'm especially interested in how well they deal with bass EQ, sub/mains integration and also what effect the EQ above the transition frequency (where the speaker's response dominates) has on the speaker's balance and voicing.

    Nyal

  8. Hi Nyal thank you

    1. My crossover in the receiver is set to 80hz currently (looking at maybe 60hz) with the mains set to small and the crossover in the Velodyne DDs disabled (Off in Velodynes words). So i would basically measure and EQ from 15-120hz?

    2. The receiver is set to small for the mains (and all other speakers). Would this be the correct procedure?
    (a) Measure/EQ the subs (mono) as above seperately
    (b) Measure/EQ the mains (both together, mono) down to 40hz
    (c) Measure/EQ altogther (both subs and both mains) with the crossover set back to 80hz

    3. With the review of the Velodyne DD+ it would be intersesting to see the diiference in manual EQ between the DD+ and DD.
    Maybe you could demonstrate the benefits of combining the XTZ with the settings of the Velodyne.
    e.g 1. Using the subsonic filters to adjust roll off so not having to use as many filters in the PEQ for better decay
    2. Using the presets and 1 and 3 position measuring (in the XTZ) to setup for different seating arrangements
    3. Also halving the Q that the XTZ suggests and also using the manual way (as you suggested) seems to work extremely well with Velodyne DDs. There was no adsjustment made to either the frequency or gain just the Q. When i re-mesured it was extremely close to the predicted measurement.

    Thanks again

  9. Hi Peter

    1. Yes. Generally I only advise using EQ to deal with low bass issues (<100Hz) anyway, and to use acoustic treatment to deal frequencies 100Hz and above.
    2. Yes that procedure would work. Before you EQ make sure the sub and mains are properly integrated (i.e. time aligned). Step (c) should be more of an adjustment to the sub or mains filters already established in steps (a) and (b) rather than creating all new ones.
    3. Good idea. Honestly I did not find the 'auto-EQ' in the Velodyne that great. Better results were had doing things manually :) Unfortunately I don't have a DD to play with, and they are now discontinued so it is unlikely I will be testing one of those.

    Nyal

  10. Hi Nyal thank you

    1. I noticed that you reduced the gain suggested by XTZ, is this typical?

    2. When adjusting for the trial on the Veloodyne DD+ can you see if halving the Q with the same frequency and gain suggested, and then re-measure and see if the predicted is the same as the re-measure?

    3. How about the DSP Antimode 2.0?

    Thanks again

  11. Hi Nyal

    I noticed that there are dips and peaks, not flat.

    1. In the bass region <250Hz which is more important good decay without tails or flat response or both?
    2. Is there a specific db to drive the modes? My SPl shows 70-75db but in the frewquency graph it only shows 40-50.

  12. Hi Peter

    The graphs shown were for the purposes of tutorial and with a limited number of bands. It's quite normall to have minor peaks and dips after EQ.

    In respect to your more detailed questions:
    1. Both frequency response and time decay are important.
    2. Not sure why you are getting such low SPLs on the measurement graph. You may want to try measuring at 85dB?

    Thanks – Nyal

  13. Hi Peter

    For HT at 1/12th octave I'd look for the whole frequency range being within a 5dB window. At 1/3rd octave within a 3dB window.

    Nyal

  14. Hi Nyal hope your Christmas was enjoyable and look forward to your blogs in the new year.

    1. When i am setting up dual mono Velodyne DD subs, when i EQ using the XTZ Room Analyzer for each sub the prediction is very close to the re-measurement, but when i do the same with the combined DD sub response the re-measurement is alot different from the prediction?

    Thank you
    Peter

  15. Hi Nyal sorry just looking over my post something wasn’t clear.

    When i say combined i meant both DDs playing mono and the exact same filter values (FC, Q and Gain) were entered into each DD.

    Thank you
    Peter

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>