What are the sonic DISadvantages to EQ'ing bass ( & only bass)? Please be specific?

caesar said:

Gentlemen,

Are there sonic downsides to EQ'ing the bass? Does one lose signal purity when EQ'ing the bass frequencies? Does the life get sucked out of the music.

Feel free to argue, just please answer my question!

Click to expand...

There are no downsides whatsoever if the crossover of your main speakers is not affected in any way. This is the case if you have separate subwoofers that run strictly in parallel to your main speakers, i.e. without crossover filtering away from them.

In fact, the ability to freely EQ bass is one of the reasons why I love my speaker set-up, mini-monitors that fall off in linearity below 50-60 Hz, and subwoofer. The (powerful) upper and mid-bass is provided solely by the main speakers, and the low bass by the subwoofer. I find the low-frequency content of recordings far too varied as to imagine being able to live with a truly full-range speaker without any bass EQ.

For example, the low bass on Joe Walsh's "But Seriously Folks" CD is a bit too fat at standard setting and needs to be reduced two notches from there on my system to sound good. On the other hand, the low bass on Led Zep's re-mastered "Physical Graffiti" CD is far too weak at standard setting *). Yet when turned up a few notches, it sounds great on my system, and is still tight and tuneful. Similarly, my favorite CD of Beethoven 9 (Dresden Philharmonic/Blomstedt) needs to have the low bass turned down to sound good. On the other hand, the low bass on the Collins CD of Maxwell Davies' First Symphony needs to be dialed way up -- and then it turns out to be of truly excellent and 'fast' quality with a lot of great 'blackness'.

__________________
*) On another forum I read a post by a guy who thought something was wrong with his bass based on precisely that CD and he wanted to change his system just because of that! Poor lad, I was thinking, why don't you simply turn up the bass? (Well, if you can, that is.)

The question of signal purity when adding components is a religious war.

The biggest downsides to EQ'ing the bass that immediately come to mind are:

1. EQ can be used to significantly increase the power at "null" frequencies, reducing system (amplifier and speaker driver) headroom. It must be used in moderation and preferably with other things like room treatment and careful subwoofer placement.
2. EQ'ing flat results in a sound many folk simply do not care for. Most of us like a rising bass response, and for that matter like frequency irregularities in the rest of the audio band. The "preference vs. reference" response is another religious war.

IMO etc. - Don

Not at all just the opposite. You want good pitch definition and tone. To me there is nothing worse than one note high q subs that are essentially boom boxes. If you have obvious bass peaks when you notch them out it may sound like "less Bass" because it is. What matters in the balance within the last 2 octaves. Many people including me don't run a flat response. The bass levels are up a bit compared the the midrange. I find flat to sound quite anemic. So even though the bass region is up the bands in that region are closely matched as far level.

Don't let EQ scare you off. Look for subs with built in DSP preferably with at least 3 parametric EQ's where you can vary frequency, Q and level of the individual filters.

Rob

I use a Vel 18 and a stereo 32 and my speakers have 8 6 inch woofers and the bass is about as well defined and musical as I think possible. I hate to sound like a broken record,but noise reduction has had a profound effect on bass reproduction,power and musicality. So you have another option and I would pursue that before any EQ option. Another point is impedance matching of the amplifier to the sub,which I don't quite understand how my method of NR effects the total system impedance,but it must as the bass quality improves more then I think pure NR would do.

DonH50 said:

It must be used in moderation and preferably with other things like room treatment and careful subwoofer placement.

Click to expand...

Sure, I'd never want to have a boomy bass. When in doubt between two settings, I choose the slightly lesser one.

Oddly enough, although my room sounded bad before acoustic treatment in the middle and high frequencies (I didn't know how bad until I heard the difference), the bass quality was always excellent and was the least affected by the room treatment (I have a wooden house, no concrete bass 'boom').

EQ should only be used to attenuate frequencies. Also, I've found better results using an analog EQ in the chain, such as in the tape loop of your pre. And like was said above, that it doesn't interfere with your main crossover units.
And forget graphic EQ's, use only a Parametric EQ. There you can change freq. and Q

Depends on what type of EQ. If done right, there are only upsides to using filters.

Bruce's comment is only partly true. You can use effectively EQ to boost minimum phase bass issues as long as the amp is up to the challenge.

Caesar,
Why don't you simply try some EQ and let us know what you think? You have a computer and a DAC, right?

There are many flavors available to demo and they are all free to try. You might enjoy that more than watching others fuss and fight over this stuff. That is, if you are serious about getting better bass.

caesar said:

Gentlemen,

Are there sonic downsides to EQ'ing the bass? Does one lose signal purity when EQ'ing the bass frequencies? Does the life get sucked out of the music.

Feel free to argue, just please answer my question!

Click to expand...

caesar said:

Hi Bruce,

Why should it matter if it's analog or digital? we are talking very low registers. Can one actually hear the difference here, or are we just "playing it safe"?

Click to expand...

I'm trying to appease the purists!!

^^^^ I think the degree of the disadvantage is related to the sensitivity of the speakers and the available amplifier power. In a powerful system, it may not be a problem at all.

I have seen nulls of 10 dB or more in almost every untreated room response plot I have seen posted. That implies ~10 times the midband power is needed to overcome the null. So in my mind it is a very valid point, but I do not know how practical for most folk. And, I really mean it when I say "I do not know". Perhaps it is very practical for many people. Even so, I prefer to maximize my system's headroom by addressing bass nulls by other means than EQ. Peaks should not exceed 6 dB or so and so are a little easier to deal with, but still...

Where on earth do people get the idea that they are conserving headroom by not applying any boost filters? It just isn’t so.

Let’s say your room generates an 8 dB peak at 60 Hz from your sub. Since that peak is the determining factor of what you hear from the sub, your sub / mains blend is based on that peak. What happens when you apply a cutting filter to eliminate it? Your sub level is now too low, so you have to increase it. Well – say “goodbye” to any headroom you thought you saved, because now you are pushing your sub harder than you were before.

People don’t consider what happens to electrical response when applying cutting filters, but with a little modeling using REW we can see that they leave peaks in response between the cuts, not to mention (at least in the case of the graph below) a big honkin’ boosted shelf on the back end.




Bottom line, there is no free lunch: The simple truth is, virtually any equalization taxes amplifier (and driver) headroom, so you have to have enough to spare going in.

There is also a lot of misinformation that floats around about whether or not nulls can be corrected with equalization. The short answer is “they can’t,” but people often confuse nulls with a mere trough or depression.

A true null is typically narrow and deep, as the graph below shows. Nulls can’t be corrected with equalization.




A trough is usually a rather broad depression in response and can typically be corrected with equalization, as the graphs below show.




Note that what appeared to be a possible null just north of 45 Hz smoothed out nicely with equalization. So it wasn’t a null at all, just a low spot situated between a couple of peaks.

Many misinformed people would apply a slew of cutting filters to bring everything down to the level of the 30 Hz trough, in the interest of “saving headroom.” However, it should be evident that’s merely going to lower the overall sub level considerably, requiring a significant boost in gain on the back end – again, no conservation of headroom in that. Not to mention, using the equalizer as a de facto volume control is poor form. A series of boost and cut filters would be the appropriate action.

Regards,
Wayne A. Pflughaupt

Excellent post, Wayne!

My assumptions were a true null and no re-normalization of gain after correction (the latter is at best optimistic as your post explains).

A null caused by standing waves at the listening position can in theory be infinitely deep. In practice I have seen perhaps 30-40 dB or so. Since the waves are actually cancelling at the listening position, EQ does not help; it is like multiplying by zero. No matter how much gain you apply the answer is still zero. What it is doing is increasing the signal to the amp and speaker, which will sound louder in other parts of the room, but at the listening position it's gas on a fire. Just wasted power.

We should remember that the depth of nulls depend a lot on the materials that are used in the boundaries of our rooms. My listening room main walls are built from two feet thick hard stone - before I build some large bass traps I could measure some narrow nulls about 45 dB deep. And yes, I can confirm nulls can not be equalized In single speaker configuration, but can effectively tamed using separate subs with equalization.

The best answer I've seen about whether one can effectively EQ a response dip is in the Room EQ Wizard manual. There's this idea of "minimum phase" in system theory that relates to whether the inverse of a system transfer function is causal and stable. If both are true, the system is minimum-phase. Given an amplitude response, there's a formula for computing the (unique) phase response of a minimum-phase system having that amplitude response. In the strict sense of the word, a system is either minimum-phase or it is not. But there's also this relaxed notion of "minimum-phase over a finite band of frequencies". In this view, if the phase response is numerically close to the minimum-phase response computed from the amplitude response, the system is said to be minimum-phase over that finite band of frequencies. If one takes such a system and equalizes it only over the frequency range for which it is minimum-phase using minimum-phase equalizers such as parametric EQ IIR filters, then fixing up the magnitude response will also fix up the time domain response caused by the non-flat response in the minimum-phase frequency region(s).

There's a fly in the ointment though. Because of delays due to physical distance in acoustical systems and the speed of sound, no such system is minimum-phase, even over a finite band of frequencies. But a constant delay due to the speed of sound and the distance the wave travels does not affect the ability of minimum-phase EQ to fix up the time- and frequency-domain responses of a system that would have been minimum-phase over a frequency band were it not for the propagation delay of the wave.

The Room EQ Wizard manual describes in ingenious solution to this problem involving the computation of the so-called "excess group delay" from the magnitude and phase response measurement. Here is the section of the manual that shows how to compute the excess group delay, which in turn can be used to determine whether it's "safe" to EQ a given band of frequencies.