[jsawyer09]

Hello. I've posted this over at AVS as well, but wanted to get as many viewpoints as possible, giving it exposure where I can in the event there's not many responses.

So, here goes: I wanted to get people's opinions on whether or not the following measurements are good/sufficient for a 2-channel only system; or can I expect to gain some improvements with a little more effort, time, patience and REW/possible DSP? And if so, how might I go about achieving this objective where the nulls and rises are concerned? It appears that with the Definitive Technology Mythos STs, people have reported, as have reviewers, a 'natural' rise in the area where I've stated below. So this may explain what I'm experiencing there (I guess I don't know for certain). Also, is the drop-off in the 25Hz - 20Hz region anything out of the ordinary; same question for the drop-off in the 16kHz - 20kHz range? I'm trying to get a grasp on what the accepted wisdom is regarding great, flat measurements in comparison to problematical regions. Here are the measurements I've taken:

64dB  70dB  73dB    74dB  76dB  72dB  72dB   66dB   65dB    72dB    72dB

20Hz  25Hz  31.5Hz  40Hz  50Hz  63Hz  80Hz  100Hz  125Hz  160Hz  200Hz

75dB    77dB    76dB    76dB    76dB  72dB   72dB  71dB    71dB

250Hz  315Hz  400Hz  500Hz  630Hz   1K    1.25K   1.6K     2K

73dB   77dB  80dB  84dB  85dB  80dB  74dB   70dB   67dB  60dB

2.5K   3.15K   4K      5K    6.3K    8K     10K    12.5K   16K   20K


I've cut and pasted the following post I placed yesterday for a little context on how I came about the measurements. It is here:


It all began with a friend of mine's purchase: a pair of Def Tech Mythos STs. I somewhat chronicled it in the speakers forum a few weeks back, trying to figure out how to cross-over his SVS SB12-NSD sub with his Mythoses, while running an Antimode 8033. I think since then, we've got it sitting pretty good; and running REW in his space shows that we've probably done about all we can do. It does sound very nice, though.

I've since picked-up a pair of the Mythos STs as well for a 2-channel-only system. Wonderful speaker, BTW (but I'll maybe talk about that on the Mythos forum)! I also have the same SVS sub as my friend, running the Antimode 8033 as well.

I have done measurements with the right speaker's gain dial at about 9:30, and the left speaker at about 10:30 (the R is about 4.5 ft. from a side wall; the L about 7 ft. from the other side wall, which then opens-up, or fires into/toward an archway where a dining/kitchen area with equal space as my listening room resides. So, there's quite a bit of cubic feet that needs to be pressurized. Now, the SVS sub is on the back wall beside my couch/listening position at nearfield. At the moment, I've set the phase at around 90-degrees (I tried 0 and 180 as well, but honestly could not tell much of a difference here), the gain around 1:00, and the crossover at 75Hz. After fiddling with measurements in order to calibrate the speakers/sub (once landing on the above) so they'd both be 75dB, I ran test tones using an SPL meter, C-weighted and set for slow for the bass, mids and highs from Stereophile's Test CD2.

Thus far, in this configuration, from 20Hz to 20kHz, things are pretty good overall (though I've no doubt they could be better...just not certain how to tackle it yet). Across the spectrum, it is admirably flat (within +/- 2dB) for the exception of the following: The bass starts to roll off at 25Hz, then drops like a brick before registering at 20Hz (down about -9dB). I'd like to get a better response from my bass, seeing that I essentially have '3 subs', but am uncertain how to tackle this, either. I have a nasty -10db dip for some reason between 100Hz - 125Hz; a +9dB rise between 4k - 6k; and then the response rolls off from 16k - 20k. All of this is with the Antimode 8033 in the chain.

BTW, my form of amplification isn't via an AVR; it is a Rega Elicit 2-channel Integrated without bass management capabilities.

Anything offered is greatly appreciated. Like I say, it sounds wonderful so far, but like everyone, I'm curious if it can be improved. Thanks!

[jsawyer09]

50 views and no responses? Same thing's going on over at AVS. So, is this that difficult a question? I'm not being facetious or condescending; I really am wondering why nobody is at least offering an opinion — experience with powered towers and sub integration or not.
I can try to do 40Hz, 60Hz then the 75Hz crossovers, and run the Antimode/measurements again, but thought I could be steered in the right direction first as how best to tame the aforementioned nulls and rise in said regions. Also, I really am not certain if the measurements could be considered good or problematic. Is +/-5dB swings considered relatively 'flat'?
I need to at least have an understanding of the readings if I'm to do this correctly.

[medium jim]

It is a bit difficult to read the number in relation to the hz.  It would be easier to comprehend it if you converted it to a simple plot graph.  There is a stickied thread, helpful links that has a great link on measurements.

I'm sure that one if the bass hounds will chime in and give you some good advice.

Jim

[James Romeyn]

Medium Jim is correct.  It's way too much time and effort to decipher for free anyway.  Graph it and you'll get more replies.   

I looked only below 100 Hz, where you have a 12 dB window.  With one sub in my room the response window is 13 dB and to call this unacceptable (for me) is gross understatement.  As I posted countless times, the FR window is only an indicator of the most audible and most unacceptable problem: bass timing error.  Where one bass or unison bass section was recorded, at times during playback your room modes create two bass or two bass sections, one playing incorrect and usually disharmonious note from the other.  This destroys bass transient information and bass pitch, which destroys the rhythm of the music.  There is a commonly understood fact in orchestra and ensemble: the rest of the musicians can play no faster than the bass section.  The bass section defines the quality of the performance more than any other section.  This is the failure in improper bass and explains why many listeners prefer limited bass cutoff.  It is, IMO, better to lack deep bass than to have deep, powerful improper bass.     

There's only two methods I know of to achieve what I call acceptable bass.  My definition of acceptable is a 6 dB response window (+/-3 dB) with no EQ and no bass treatment, -1.5 dB @ 20 Hz, 113 dB maximum @ 20 Hz in 3k CF room. 

One method is a Distributed Array the other is a Double Array.  The later is much more complex, is twice or thrice as costly, and ideally requires permanent in-wall install on two walls and at elevated height, so I disregard it. 

Contrary to your note, you don't have three subs when two are built in the main speaker.  Reason being these "subs" are permanently attached to the mains and can not be separated.  A Distributed Array treats the room like what it is, not what we want it to be: rooms behave differently below vs. above the Schroeder frequency (about 150 Hz).  You have one sub + two towers, each with what are (to me) useless bass reproduction below 80 Hz.

This is one person's opinion only and I'm happy to report about 99.9% of members here have no experience with and no interest in three (four is better) subs, even though mine are only 1cf each.  But there are a few other members who believe in Distributed Array as much as I do.

Again, if you have a 12 dB window after EQ and a lot of work, that is fail.  Move your mic around the room and I bet you have hot spots mear the sweet spot with 20 dB window (almost inevitable when EQ is employed).  EQ makes worse elsewhere what it improves at the sweet spot.

My advice: get three more subs and follow my sub setup advice.  You will employ no sub EQ unless you have parametric function (likely not) in which case call me for advice.  High-pass the mains @ 80 Hz.  With your mains high-passed @ 80 Hz, you won't need towers of course.  Towers don't hurt, but you have much wider speaker choices with mains high-passed @ 80 Hz, including good stand mounts.           

[SteveFord]

[neekomax]

@SteveFord...

You are a nice dude. 

[SteveFord]

I try.
The software was from some Kid's Corner on an educational site which made me laugh.
How tough can it be?

[neekomax]

I try.
The software was from some Kid's Corner on an educational site which made me laugh.
How tough can it be?

Just a little too tough for the OP, apparently, but not so tough that you couldn't pull it off 

Anyway, what do we think of those measurements? I'm agnostic save for the sizzle at around 5.5Khz. Ouch. Bass looks good, no?

[*Scotty*]

In as much as this is framed by the OP as question sub-woofer integration, I will say that the apparent suckout at 100Hz and 125Hz is troublesome. Not knowing the OP's room dimensions I can't tell if it is related to the sub-woofer crossover frequency of 75Hz and its slope or if the room and the subs position relative to the room boundaries is contributing to the problem.
Scotty

[medium jim]

Scotty beat me to the suckout.  For some reason, the 100-125hz is where most will have issues, which is outside of where subs do there thing.

Jim

[jsawyer09]

SteveFord, you are a nice dude. I appreciate your efforts. I didn't post because I haven't been here since trying to improve some of my peaks and nulls. Can't say I appreciate the sarcasm much, but whatever.

Aside from the 100Hz & 125Hz suckout, and 5K & 6.3K peaks, are these pretty respectable measurements?

For what it's worth, I was able to experiment with phase adjustment (variable), subwoofer placement and mains placement a bit. After adjusting the variable phase and crossover (which are now set at approximately 120-degrees and 80Hz exactly), I was able to bring the dB up from 64dB to 72dB @ 20Hz, and to 78dB from 70dB & 73db, respectively, @ 25Hz and 31.5Hz. I thought this was a nice improvement.

However, everything else remained pretty much the same, except I was able to lower the peak some at 4K and 8K (both now only up +4dB, respectively), but not much 5K or 6.3K. That's where I might be able to move the speakers a tad bit more to see where that lands me. I also could not raise that suckout, unfortunately, from 100Hz - 125Hz, either. I was pleased with what I could do with the bass when all is said and done.

BTW, I settled on the 120-degree phase and 80Hz settings, not only because they really have sounded best based on all of the listening and measuring I've done recently; but because I crudely calculated my room's dimensions (12' x 21' x 8'), speaker position from LP (approx. 9 ft.) and the SVS sub's position beside the couch/from my LP (approx. 5 ft.). So, that would mean, if I'm understanding it correctly, that:

First, at 80Hz, 80/1130 = 14.125; approx. 5 ft. is the difference between distances of mains/sub/LP; 5/14.125 x 360 = 127.43 (almost 120).

At these settings (for the exception of the suckouts and peaks, which with a little more tweaking and speaker positioning, I'll hopefully achieve better results), I feel like things are pretty flat across the spectrum; and the sound is very, very good.

Anyway, thanks for most everyone's responses; and thanks again Steve for your efforts.

[medium jim]

This thread belongs in the acoustics circle.

Thanks,
Jim

[genjamon]

I was able to deal with several bass suck-out regions at once when I implemented a second subwoofer.  Based on this limited application of multiplying subs to eliminate peaks/nulls, I can definitely see how the multiple distributed subs approach could be quite successful.  Two subs and a DEQ2496 was enough for me to get +/- 3db, a swing of about 6db between 20-100 hz for my seating position - though still pretty damned uneven for the rest of the room.  It took a lot of playing with the DEQ and REW as well as some location play for the subs, but I'm satisfied.

[Ethan Winer]

This thread belongs in the acoustics circle.

That must be why I missed it originally.

The only comment I'll make for now is that spot checking bass frequencies at third-octave intervals is insufficient to tell what's really going on. Proper room measuring software such as the freeware Room EQ Wizard gives much more information. Not only higher resolution, but it also shows modal ringing, reverb decay times, and even individual reflections. The graph below shows the response and ringing in a typical home-sized room. As you can see, you can easily have a null, a peak, and another null all within one third-octave band! This article explains more about using such software:

Room Measuring Primer

--Ethan

[James Romeyn]

I was able to deal with several bass suck-out regions at once when I implemented a second subwoofer.  Based on this limited application of multiplying subs to eliminate peaks/nulls, I can definitely see how the multiple distributed subs approach could be quite successful.  Two subs and a DEQ2496 was enough for me to get +/- 3db, a swing of about 6db between 20-100 hz for my seating position - though still pretty damned uneven for the rest of the room.  It took a lot of playing with the DEQ and REW as well as some location play for the subs, but I'm satisfied.

Your 6 dB bass window is very good.  My window is 6 dB with no EQ and no bass treatments.  If you're completely satisfied and have no further curiosity on the subject, great.  If curious, and if time permits (estimate one hour), readers might benefit knowing if my setup instructions further improve your specific system with two subs.  I suspect yes. 

Preliminary questions: What is main speaker natural bass f3?  What type bass loading?  (sealed, reflex, OB, etc.) Do you employ hp filter on main speakers?  If yes: what pole and slope? 


 

[Alex Reynolds]

+1 with Ethan's suggestion above. REW or other similar programs can give a much better analysis of the room (plus, REW is free anyways). We did a video showing how to set up REW and take a measurement here: http://gikacoustics.com/room-eq-wizard-tutorial/

The ~120 Hz null is likely due to SBIR. Also, a 12dB window is not too bad a response if the measurements are not smoothed at all. Really, you should be looking at the decay of those frequencies as it will give you a larger insight to what is going on and how to treat the room. This article goes a bit indepth but covers a lot of ground on decay times: http://gikacoustics.com/understanding-decay-times/

[jsawyer09]

Thank you gents for your replies. This is quite helpful. REW will be the next step, indeed. It has been a while since I measured in my room (I have since changed speakers, added a sub, switched-out furniture/area rug and listening position over the course of a year or more), so I made crude measurements to see where I'm at. Overall, I don't feel the majority of the measurements are all that bad; the +/-6 dB swing from 'reference' as measured at 75 dB is very consistent. I was actually quite shocked, given that I am forced to listen along the long wall at this point. 

Alex, I feel you may be right about the speaker boundary interference for the nulls. At first, I instinctively blamed floor bounce, but there's an area rug that fills in the floor between the speakers and the couch. However, there are 5 bay windows behind my setup/speakers that jut outward, with a flat sort of built-in shelf that houses books, pictures, etc. The windows have wooden blinds on each of them. I typically have them closed during listening sessions, but I'm certain that when I measured and re-measured all of this, I had two or three of the blinds open. Perhaps this is the culprit, or no? And the peaks, well, not certain how to address those yet. Maybe a combination of speaker re-arranging by a few inches here and there, combined with subwoofer? Like I say, I feel like the measurements are respectable; and I hate to alter the results too much to the point that I then have to address wild variations in those. However, I'm also realistic about how much time this can take to get it right, so I'm willing to experiment.   

[genjamon]

According to this blog post from the designer of my main speakers (http://tektondesign.wordpress.com/2011/05/28/tekton-lore-perfect-bass/), the anechoic -3 db point is 38hz, -10 db is 31 hz.  I run them full-range, no HP filter for bass.  I run amp outputs to the first sub, which sums the channels, and I run both channels to the second sub in a daisy-chain configuration.  So, both subs are summing to mono.  I have speakers along the short wall, about 3 feet from rear wall and 2 feet from sidewalls.  This is a good balance of soundstage depth and room bass reinforcement for these speakers.  Room has a large opening to dining room at right rear of listening position, and also opens to a hallway at rear right. 

The first sub is along front wall about 1/4 the total wall length, the other sub is on rear wall directly behind listening position.  I found the best combination where I run the rear sub up to 70-80 hz range, and the front wall sub with the low-pass filter at 40hz.  I believe both of them use first order low pass filters.  Basically, I use that front wall sub to fill in the low bass, while the rear sub handles the rest of the bass.  I use the DEQ2496 to EQ the subwoofers only - and it mainly only affects the rear wall subwoofer, as I need the deep bass reinforcement, but I do cut the frequencies down to about 30hz, and then below 30 hz the subs are run without signal cut.  This boosts those low low frequencies that the main speakers don't handle.  I found just using the EQ'd rear wall subwoofer on its own to take care of a lot of the unevenness, especially the peakiness.  However, I still had a couple suckout regions - and one particularly egregious (I think in the 50-60 hz region, but can't remember exactly).  When I added in the front subwoofer again, filtered for low bass, this suckout at 50-60 hz disappeared.  My theory is that the second subwoofer offset from the first eliminated a null at the listening position through wave interference at a multiple of what that front wall sub is focusing on - the 20-30 hz band.

Anyway, worked out pretty well by doing the best I could with one subwoofer first, then adding the other for a targeted weak region, which happened to disrupt nulls at higher frequencies as well.

 

Your 6 dB bass window is very good.  My window is 6 dB with no EQ and no bass treatments.  If you're completely satisfied and have no further curiosity on the subject, great.  If curious, and if time permits (estimate one hour), readers might benefit knowing if my setup instructions further improve your specific system with two subs.  I suspect yes. 

Preliminary questions: What is main speaker natural bass f3?  What type bass loading?  (sealed, reflex, OB, etc.) Do you employ hp filter on main speakers?  If yes: what pole and slope?

[Alex Reynolds]

Thank you gents for your replies. This is quite helpful. REW will be the next step, indeed. It has been a while since I measured in my room (I have since changed speakers, added a sub, switched-out furniture/area rug and listening position over the course of a year or more), so I made crude measurements to see where I'm at. Overall, I don't feel the majority of the measurements are all that bad; the +/-6 dB swing from 'reference' as measured at 75 dB is very consistent. I was actually quite shocked, given that I am forced to listen along the long wall at this point. 

Alex, I feel you may be right about the speaker boundary interference for the nulls. At first, I instinctively blamed floor bounce, but there's an area rug that fills in the floor between the speakers and the couch. However, there are 5 bay windows behind my setup/speakers that jut outward, with a flat sort of built-in shelf that houses books, pictures, etc. The windows have wooden blinds on each of them. I typically have them closed during listening sessions, but I'm certain that when I measured and re-measured all of this, I had two or three of the blinds open. Perhaps this is the culprit, or no? And the peaks, well, not certain how to address those yet. Maybe a combination of speaker re-arranging by a few inches here and there, combined with subwoofer? Like I say, I feel like the measurements are respectable; and I hate to alter the results too much to the point that I then have to address wild variations in those. However, I'm also realistic about how much time this can take to get it right, so I'm willing to experiment.

I doubt that the blinds would make a difference that low in frequency.
I wouldn't put too much weight into the SPL measurements as they don't tell a whole lot about the room. While the level might be accurate, we have no idea of how long each note resonates, and SPL will also register certain harmonics of the notes played and add it to the level, even though it might not necessarily be the note that you were playing. +/-6dB is a good response, but it is a good response when you're talking about full range measurement. There might very well be variations of that inbetween your frequencies of measurements.
Sorry, not trying to rip it apart, the room could very well be +/-6dB with measurement too, just want to say the SPL measurements don't really give you enough information.

[genjamon]

Good point, Alex. And I wouldn't doubt that my measurements for my room have obscured those very issues to some degree.