Experience and Thoughts on EQing (1 Viewer)

[Doug_M_Fraser]

I have posted this here because I believe that the subject of EQ can be treated as an advanced topic if persued with rigour. I have been very interested in EQing my set up for some time now. I have lurked here for several years and occasionally posted. Now I would like to give back and present some activities and views on EQing.

My sub experience…..

I have a Stryke 15.2 sub driven with a Crown K2 amp. This link http://members.cox.net/frankcarter/T...e Sub Data.htm documents the Tom Nousaine’s test results for the Power 15.

Before EQing the sub, it sounded real bad. I purchased Doug Plumb's ETF http://www.etfacoustic.com/ software, swept my room and found I had a 30dB resonate peak at 38 Hz (plus a bunch of other anomalies). Thanks to Sonnie Parker and his BFD set up guide http://www.snapbug.ws/bfd.htm and Wayne A. Pflughaupt for his advice and guidance and many posting by Ken Bruce, I managed to tame the in room response and add a house curve. So all is now good in Sub Land.

With respect to EQing a sub here are some idea's that I think are important and haven't been discussed in detail on this forum (to the best of my knowledge) . The Tag Mclaren Audio site http://www.tagmclarenaudio.com has a great tutorial on using ETF to EQ a room. Go to the Club area. You need to log in, however, they don’t seem to send Spam. Go to the TMREQ project and look at “Room Resonances their Visualization in Measurement Systems”. If you take the time to follow through the diagrams, what is apparent is that one needs a very narrow bandwidth EQ to remove room resonances. The BFD is such a device! It is very clear that you need to accurately pick the centre frequency dead on. In addition, the bandwidth of the filter is very critical too. This needs to be done using test equipment and luckily, ETF is available in a demo version for free. Therefore I recommend that you are serious (or should I say anal - like me) that you reveiw the tutorial and understand why it is important to have the correct centre frequency and bandwidth.

As side note, the Tag tutorial refered to above presents an interesting procedure as to how to EQ your Sub. They suggest you should measure the response curve of your crossover and use it as a target curve for your subs in room response. This means injecting the ETF test signal before the crossover. Something that is easliy done if you use the crossover in your SSP (Surround Sound Processor). I originaly used the techniqe as described by Sonnie in his document, however I plan to give the Tag technique a try in the future. The beauty of the BFD is that you can save differant EQ curves for comparison.

It seems that in most cases, even a 31-band graphic EQ would not be up to the task as it has fixed centre frequencies and bandwidth. So my advice is get the BFD. It is inexpensive (about $120 more or less) and it works!

Now on to the other speakers….

A local audio store here in Vancouver, BC called Speaker City http://www.speakercity.ca/index3.html (not to be confused with Speaker City USA) had built some speakers using drivers from another local Vancouver company AVI http://www.avisound.com/ . I auditioned the speakers and fell in love with them. Their mid's and highs sounded very close to my B&W 802’s. Now Speaker City seems to have a soft spot in their heart for DIY guys. They sold me the drivers, crossovers and provided plans. I ended up building 7 speakers including dipole surrounds. Speaker City even redesigned the crossover to compensate for the two-tweeter design (for the dipole design) instead of one tweeter.

Now bitten by the speaker building bug I started to notice that some crossover designs try to equalize out bumps in the combined response curve of a speaker. If you go to here http://www.linkwitzlab.com/reference_earphones.htm you can see that Siegfried uses passive components to equalize earphones. In addition Siegfried invented the Linkwitz Transform circuit to boost the low frequencies for subs in a sealed enclosue. http://www.linkwitzlab.com/filters.htm#9 This is simply an application of very specific equalization.

This got me thinking, why not get a good parametric EQ and have a go at my other speakers. So at this point I have purchased two Behringer DEQ 2496 EQ’s http://www.zzounds.com/item--BEHDEQ2496. This is quite the piece of gear. First of all, it samples at 96Khz with 24 bit words. The DEQ2496's provide the following functionality, all done digitally: stereo 12 band parametric EQ’s, stereo 1/3 octave graphic EQ’s, a real time audio analyzer, auto configuration of the graphic EQ by using a calibrated mic http://www.zzounds.com/item--BEHECM8000 and a plethora of other functionality that I will never use.

So with ETF and a pair of DEQ2496's I now have the ability to correct for very narrow band anomalies as well as wide band tone shaping for my 3 front channels. I only intend to EQ the three front channels, as my receiver, an Integra 9.1, does not provide access to the signal between the internal pre amp and power amp for the surround or back channels.

We are in the process of redoing our family room right now so I won’t get to the task for at least a month or so. When I do, I will post results, experience etc.

I welcome your thoughts on the above or topics associated with EQ subs, speakers etc.

Regards,
Doug

 

[Chris Tsutsui]

What can greatly simplify your quest for room equalization is acoustical room treatments.

Check out bass traps, and acoustic panels which can help act to reduce the effects the room has on your sound system.

I built four bass traps that are 4 feet high and 16" diameter. This helped reduce all the lower HZ dips in my frequency response by as much as 3 decibels not to mention flatten the overall response. My 3.5" thick fiberglass acoustic panels absorb as low as 80hz which is enough to rid of any slap echo and unwanted reflections.

You can use all the parametric equalization in the world and you won't be able to get rid of the DIPS in your response. This is because standing waves are canceling out eachother and creating nodes at the listening position. (At least dips arn't as noticeable as peaks)

I suggest that you first read the white papers at harman.com on bass equalization. This will coach you how to position your loudspeakers and subwoofer in a room in relation to the listening position to allow the flattest overall frequency response. Then once your positioning is as best as it can get, then try acoustical treatments.

Last, if its even necessary at this point, i'd go with parametric equalization. You ONLY want to reduce the peaks in your response to help flatten it. DO NOT attempt to raise the dips using an EQ, because it doesn't work that way. If a wave is canceling out with itself at a specific frequency. The more you increase it, it will still cancel out with itself. I think that your EQing should only target the subwoofer frequencies. If your mains do not provide a relatively flat response then maybe you should just tweak those or upgrade.

Perhaps try playing with Absorption and diffusing acoustical treatments if you can. I think they can make a significant improvement in sound systems which could even be greater than an EQ.

Again, I Urge you to play with position of BOTH the listening location, and the subwoofer before you decide to use the BFD to solve your huge peak problems in the room. I graphed data that showed differences when I moved the subwoofer only a foot. I also recorded huge differences by raising or lowering the listening position. (Remember that axial waves occur between floor & ceiling and wall to wall.

I'd sit down and make a floor plan of your room with L-W-H dimensions. Then calculate (using software/calculator/ or Excel) where all the possible 1st, 2nd, and 3rd degree standing waves occur. You could adjust the positioning of your sub along with your listening position to help reduce the audible effect of those standing waves.

I know this is really important if you have dual stereo subs in which you reap the benefits of not having to worry about as many standing waves. Then again, if some frequencies arn't affected due to dual subs, it may make others more noticeable. It's not easy getting a flat response in a room...

Good luck, and let me know what works out. Also, welcome to the wonderful world of audiophile sound. (Appreciating sound more than the average Joe)

 

[Doug_M_Fraser]

Chris:

I have read the papers and presentations done by Dr. Toole that are posted on the Harman site. I agree with your comments on cutting vs. boosting, room treatment etc.

I wish the world was perfect and I could have a dedicated HT room, however, I use our family room for the HT. Like alot of us, I suspect that the SAF is always a limiting factor in being able to adjust speaker placement and apply room treatments. My wife has just recently stopped commenting on how many speakers we have in the room. Even though the installation is very clean, wires in the walls with outlet plates, front speakers (book shelf) on articulating arms attached to the wall, subwoofer that is disguised as an end table etc, etc there is no way that I could introduce base traps, absorptive panels for 1st reflections etc. I would have to find a new place to live.

I must say, that even without room treatment that my experience in EQing the sub has been very dramatic. It has gone from sounding terrible to now very good. In staying consistent with your comments, the maximum boost that I use is 3dB on only 2 filters. This is done to gentle shape a house cure at the low end.

With respect to EQing higher frequencies this is going to be a big experiment for me.

The first point I want to experiment with is in the area of Psychoacoustics. At this point I neither advocate nor dismiss the effects or benefits. There is a body of work done by, Jens Blauert, a German, in the area of Psychophysics. An English version of his work can be found here: http://mitpress.mit.edu/catalog/item...ype=2&tid=7763

Mr. Blauert’s suggests that there are five critical frequency bands that effect the localization as to front/back and height perception. It seems that controlling these narrow bands by a few db will give more "presence” or a more diffuse, distant sound perception. I am going to give his ideas a try.

The second area I am going to look at, as I explained in my original post, is to smooth out the near field response of my speakers. It is clear to me that some cross over designers’ attempt to address this. At this point I don’t know how well my existing cross overs address this however I intend to find out.

Finally, the last point that I wanted to make was that from the Tag tutorial, for low frequencies, it is critical to accurately pick the centre freq and bandwidth of a resonant peak one is trying to EQ out. Until I saw this work I did not realize the importance of being able to have very narrow adjustable filters. This should provide guidance to those who ask the question can I use a 31 band graphic EQ for my sub? The answer is clearly – for an optimal results, a parametric EQ is required as is a tool (such as ETF) to measure the in room response.

I would be interested in hearing from anyone who has experimented with Psychoacoustic tuning of their system.

Regards,
Doug

 

[Wayne A. Pflughaupt]

Don’t sweat it, Doug. We have a knowledgeable acoustics consultant on this Forum, Terry Montlick, and he acknowledges that bass traps aren’t nearly as effective as equalization. I’ve helped more people than I can count equalize their subs on this Forum, and just about every one of them got the best performance with their sub in a suitable corner. This is why I typically tell people that moving a sub around the room looking for that magic spot is an exercise for people who don’t have an equalizer.

I’m particularly interested to hear what you find when you use the DEQs, because I’m leery of budget equalizers for the mains, especially digital ones. So keep us posted and let us know how it goes.

Regards,
Wayne A. Pflughaupt

 

[Doug_M_Fraser]

Wayne:

Nice to here from you. I value your experience and comments.

For the mains I run an Integra 9.1 with spec'ed power of 170 w into 6 ohms. My speakers are 4 ohms and are of average efficiency (I haven't measured them but I haven't perceived them to be power hungry). Having said that, I believe that I have a bit of headroom at moderate listening levels. We usually listen to movies at -15dB on a calibrated system (too loud for my wife but OK for the kid's and me). Additionally, I am from the school of not too much boost so I will heed your “use small amounts of boost” comment.

An interesting point in your post is to EQ the L&R together. It is possible to do this using the DEQ 2496 as I can tie the L&R EQ channels together in the Behringer DEQ 2496, similar to the BFD.

I have yet to play with the Behringer DEQ 2496 however, a review of its predecessor found here http://enjoythemusic.com/magazine/eq...ringer8024.htm is favourable. In addition, AudioXpress magazine http://www.audioxpress.com/magsdirx/ax/index.htm favourably reviewed the Behringer 8024 and recently the author of the review wrote in a letter to the editor extolling the virtues of the Behringer DEQ 2496.

For better or for worse, I am a believer in double blind tests and the doctrine as prescribed by Tom Nousaine, Ken Pohlman http://www.musicbooksplus.com/author_kenp.htm Dave Rich and Peter Aczel of the Audio Critic and various other publications. Essentially the doctrine reads that there is no difference in the sound between and expensive and a inexpensive amplifiers as long as they well designed and operate within their design constraints. It must be noted that good design does not have to be expensive. Any audible differences can be measured.

Given the above I am hoping that the Behringer DEQ 2496 has good design. I’ll let you know. Even if it doesn’t work out, its a hobby and I am having fun.

Regards,
Doug

 

[Andrew Steel]

Hi Wayne,
The graph looks great and I agree that if you can achieve that response with no treatment then great. I would like to see graphs on a grid of 2m x 1m (maybe 10 points)that approximates 2 couches in a theatre - much different to a single spot. Getting a response at 1 spot is not too hard or useful unless you always watch movies alone.

Doug,
If you can't get a uniform response for the listening area with EQ and are forced to treat the room - there is some logic to eliminating the potential for interference rather than fixing the problems - prevention over cure. The only way I can suggest to make it acceptable is to dress it up. I have had to use stylish geometric prints over absorbers, ceiling absorbers and diffusers suspended and fitted with lights, tubes made to look like columns etc. I understand how hard it is to get the OK for placing treatment. I am working with a person who has over $500k invested in stereo and a house like it was from Miami Vice - minimalist and shiny It is hell trying to get him to stop saying 'lets use some nice silk curtains, would that work?" ugh!!!!

Anyhow, my philosophy is to use the least possible EQ and acoustic treatment but still get a great result. You have to be clever and inventive. My 2c worth I guess.

Andrew

 

[Chris Tsutsui]

I still don't see how "In an ideal situation" boosting the frequencies of your sub where nulls occur at the listening position will help.

The null is where the standing wave cancels itself out whether the wave amplitude is big or small. (Whether you boost or reduce the level) Of course rooms arn't perfect and there is furniture and objects which is why the cancellation won't be perfect. Maybe that's how the 40hz dip in that graph managed to rise in volume. I just wonder how that 40hz bass sounds... it could be insane in a certain locations of that room.

I'd also keep in mind that your recorded frequency response from the SPL meter may be reading something flat, but you could still have bass that "rings" in the room or just sounds wrong. This is the sound quality of the bass.

If I couldn't use treatments, then equalization and positioning is what i'd work with.

 

[Wayne A. Pflughaupt]

Andrew,

I should have mentioned that the graph is not mine – It’s Sonnie Parker’s. He’s nice enough to let me use as a reference.

Doug is using his family room for his set-up, and if his house was built in the last 25 years or so, that probably means the room opens up to other areas - like the dining room, kitchen, entryway, etc. A room like this has not much in common with a dedicated room, which is typically a much smaller space with symmetrical boundaries and full-length parallel walls.

I’ve found that these large “family room” spaces are easy to equalize for a wide listening area. At my place I EQ’d for the sweet spot and found that bass response was perfectly acceptable in all other seats, except for those against a wall near the sub.

This is a markedly different situation from what you find in a dedicated room. There, it’s very difficult to get ideal response in more than a single position, as you indicated.


Chris,

Speaking of the same family vs. dedicated rooms, the former has a significant advantage in that special acoustical treatments usually aren’t necessary, as “treatment” is adequately provided by the normal room furnishings: Upholstered seating, window treatments, etc. Things like wall hangings, bookcases, figurines, etc. all provide dispersion of sound waves. Reflections from the floor are eliminated if wall-to-wall carpet is used, and high cathedral or angled ceilings effectively render ceiling reflections moot.

You have to keep in mind that not all low points are truly nulls. Underrepresented areas can be caused by any number of things, like non-linear response from the combined driver/box system, or the room causing hot spots here and there (you can’t realistically call a low area between a pair of hot spots a “null”). Phase issues from the crossover can cause response irregularities (high and low spots) an octave or more in each direction from the crossover point, as can interactions between the mains and sub.

It’s easy to identify nulls because they can’t be equalized. As long as the reading of a low area changes in linear amplitude to applied equalization, then it is not a null. Nulls typically are very narrow, like 1/6-octave or less. For instance, notice the low spot centered at 40Hz in the graph I posted: It’s almost a full octave wide. So obviously, in Sonnie’s room there were no nulls in the sub’s operating range, since the low spots were wide and/or they responded to equalization.

Here’s a graph from another Forum member and BFD user:





By comparison, this graph shows a 1/6-octave dip at 111 Hz. If you follow this member’s “adventures in equalizing” at this thread, you can see from the numerous graphs posted that many dips and valleys in response came and went, but the null at 111Hz was always evident and did not respond to any equalizing.

I’ve seen similar results in 1/6-octave readings and graphs posted by numerous Forum members over the past few years. Typically all low spots equalize out nicely, except for the occasional null, and all report that their subs sounded remarkably better, even with some filters boosted.

Regards,
Wayne A. Pflughaupt

 

[Doug_M_Fraser]

Wayne:

You are right, our family room is quite long and is an intergal part of the eating and kitchen area - sorta a great room concept before great rooms were popular. There are two doors, one into the formal dining area and the other into a hall.

EQ has worked wonders for me at the low end. Mind you, we also have some overstuffed couches with carpet on the floor.

I know the Wayne has advocated a house curve (which I have) at the low end. Continuing on this theme has anyone every done any sound tone shaping at higher frequencies and experimented with Psychoacoustics?

Doug

 

[Rob Formica]

Great thread... very much reflects what I'm trying to accomplish, but am still several steps behind Doug.

I think using a living room or great room with some serious SAF for our listening area is quite a commun limitation. Although my place dates from the 50's, the smallish living area opens to a hall and a multitude of doors. I wouldn't be surprised that I could do some room tuning by simply opening and closing different combinations (like in the other referenced thread) but haven't tried the experiment yet. Room treatments other than usable furniture is currently out of the question... but I am planning an eventual dedicated room. That doesn't stop me from learning and maximizing my current asymmetrical arrangement.

I have also been playing with ETF, and it really seems to be a great tool but I have to settle a interconnect RFI hum I seem to be picking up. It is not a ground loop as simply pluging in a single end of the RCA will generate the noise. My options are getting a better quality interconnect or a USB soundcard with a line-in for a portable we have. Are you guys using a calibrated mic or just the RS SPL meter? My preliminary tests with the RS SPL do show this software to be great in telling the difference between speaker / room responses. Both nulls and ringing resonances are easy to differentiate from other non-critical peaks and dips. Just purchased a BFD1124P but have not yet tried to program it... I am hoping the ETF output will help me do it more precisely.

Given at which stage I'm at, I obviously haven't considered EQ my mains or considered the psychoacoustic tuning. Doug, depending on how revealing you mains are... I would be worried about the A/D and then D/A conversion. I actually did try a blind A/B test of different CD players (to prove to my girlfriend that I wasn't insane) and could consistently tell the two units apart. Unlike some other audio fictions, not all DACs sound the same.

I am curious by possible gains from your experiments... and hope you keep us posted...
Rob

 

[Doug_M_Fraser]

Rob:

Given that you are using ETF, I strongly recommend that you go to the Tag Mclaren Audio site (follow the link in the first post of this thread) and review the equalization project that they have posted. It steps you though how the use ETF effectively. Until I did that, I really didn't understand what ETF was telling me. One set of results showed that I had significant comb filtering. From the Tag site I found out ones ear simple integrates this filtering out. It is well worth the read.

The BFD is a great tool for what you are trying to do. Be sure you get the centre freq and bandwidth dead on. The Tag tutorial explains how to do this and effects of being off.

WRT psychoacoustic I see that ETF offers psychoacoustic software modules as an add on. Has anyone used this and what were your thoughts?

WRT to AD and DA the DEQ 2496 samples at 96KHz and a 24 bit words. I hope it is implemented well and there are no audible artifacts. I will post my experience when I finally get the time to conduct the tests.

Doug

 

[Rob Formica]

Thanks for the heads up... I will have to add that article on my "to read" list as it looks quite promising. I'm currently re-reading the ETF documentation and sample room now that I've used it a little while. I had never used a RTA before, so the learning slope was quite steep on the software but the paybacks are definitely worth it. Once you understand the terms and principals, it can be a real time saver. I don't have the psychoacoustic add on so I only know what the documentation says about it... hope some others with more experience with it will tune in.

The BFD on the other hand arrived this week, so I haven't even touched it yet... but I did start reading the "comprehensive BFD setup guide".

So many tools (toys) I don't seem to have time for them all...
Rob

PS: I recently built a Stryke AV15 6ft3 vented sub to replace my Energy, so it's what got me started on the whole tweaking thing...