Axiom Home Page Forums General Discussion Advice from Axiom Owners Advice on Freq. Resp. (Wide Photo!!)

OK… now you'll all realize why I'm taking so long to post my big "impressions" article! Sorry about the scrolling, if I resize it to fit in the BBS, no one could read the text!

Below is a frequency response plot that I took a few hours ago. I wanted to see if I could actually chart the different responses from an M60 up against the wall, 12" out and 18" out. In this case, I also added a plot for 18" out with a sound blanket on the floor and wall alongside the speaker. This plot is for the M60 by itself; receiver set to "large" and sub off.

I've been mentioning in this forum for awhile how problematic my room is, and I've been doing some free-hand drawn plotting here and there with my RS meter and a disc of test tones that I made…. I probably have 30 of them at this point! Looking at this one though, I decided to draw it in Photoshop (instead of scanning a hand drawn one) and post it here for opinions.



This shows the enormous peaks and valleys that I'm dealing with…one of which is a 25db swing! The area where that valley is centered (60Hz) also reflects in plots from my Left speaker AND the sub… i.e., it's the room, not the M60.

There's a photo positioned in the upper-right corner of the graph to demonstrate the setup as far as where the speaker is located. Here's a photo showing the setup with the sound blanket:



FWIW, I've plotted the sub out of the corner position as well and find it plots close to the same for all locations, so I figure I might as well leave it in the corner for aesthetics and gain.

I tried the blanket because I thought many of the problems were reflections from the floor and wall to the right. As you can tell, the plot with the sound blanket helps a bit, but not much. Now I know that room plots are always a bit scary looking- they never are as smooth as one would hope. But this chart really looks to be off the…ummm…charts! In all honest, I'm not unhappy with the way things sound in comparison to my old speakers, but I have to believe it would be much better if I can smooth some of this out!

As I've mentioned here, I'm considering a BFD to help with my bass situation. If I change my M60s to "small" from my current "large" setting and enable the sub (with BFD) handle all the bass, maybe I can at least tame the peaks.

Any thoughts? Burn the house down and start again?

/runs away screaming!

I'm not seeing this! I don't even want to know what my room response is!

I'm not seeing the image, for some reason... ?

I added a couple of sentences...i.e. was editing the post. Try reloading the page?

Nope. Maybe my work firewall is the culprit. I'll look later. But if it makes you feel at all better, Mark, I spent several hours last night trying to tame my freq. curves. My sub went from 77 dB SPL to 113 db SPL between 25 Hz and 50 Hz!!!

With the Behringer Feedback Destroyer, I've been able to tame it a bit - now it's more like 80 dB to 95 dB. So it's going to be an ongoing project. But without seeing your frequency problems, I bet they're treatable with some EQing, if nothing else works.

The BFD is recommended for taming peaks rather thatn filling in holes, right? That 60Hz hole is going to be a b**** remedy.

Filling holes is hard. You hit a ceiling after a point and then you'll lose all headroom and bad things happen.

Of course, in theory, I suppose you could turn the levels on the sub way up (such as a doubling it, which I know the 3910 will let you do) so that the lower part of your valleys are near the ideal level and everything else is outrageously loud... then you could cut all that other stuff down.

Don't know if I'd recommend such a couse, or what attendant problems that might bring, though...

Well, maybe I should ask:
Is this really that unusual a plot? I know that many people get scary plots in their rooms.... maybe I should stop banging my head against the wall?

I EQ'd my QS-8s last night based on their plots and planned on doing the same for the M60s anyway. I guess I'm just concerned that trying to correct huge swings like these via EQ will introduce some funky phase problems, etc....

Considering I do like them anyway, I'd love to hear them in a good room!

I remember reading something a while ago that about how our brains do a good job of filtering out the specific signature a room gives to sound. I think Dr. Hsu said it.

To make an analogy, it's akin to wearing ski goggles. When you first put them on, everything looks decidely tinted. But then your eyes adjust over the period of a few minutes and everything looks normal again. When you take the goggles off at the end of the day, all the colors seem off again.

Have you tried to plot a graph with both speakers + sub? The sub might help smooth out some of the low frequencies.

I do mine with both mains + sub using the Rives Audio Test CD 2 (corrected for RS meter). All frequencies from 20Hz-10Khz are within +/- 6db at the main listening position. I do have some diy room treatments plus lots of plushy stuff in the room.

BTW, the RS meter is not designed for frequencies above 10KHz.

Mark, when you are getting your readings, do you have the meter at your primary listening location? Or is it as pictured in the image? I ask because you may find that you get some notable differences as you move the meter throughout the room. I would think that the curves might look quite different if you set the SPL meter up on your couch for example.

Edit: Actually like mwc said (except without the sub), I think the curves I would be most concerned about if it were my setup, would be the curves taken with the SPL meter in my primary listening location and both M60s running.

Scrolled my way through, Mark. Another illustration of harsh audio realities: some worry about relatively insignificant things like players, amps and wire, while even with good recordings and flat(anechoically)speakers it's the listening room that makes most of the difference. Your graphs are a little bumpier than some I've seen, but nothing unusual. Are you going to run another set with 3805 equalization to note any differences?

In reply to:

---

Have you tried to plot a graph with both speakers + sub? The sub might help smooth out some of the low frequencies.

---

Mike:
No, I don't. I wasn't kidding when I said that I probably have 30 graphs, but I haven't done both R & L at the same time. I have done both M60s individually, with and without the sub & with receiver set to small and large in several positions.

In reply to:

---

I do have some diy room treatments plus lots of plushy stuff in the room.

---

At this point, I really wish I had an acoustically "dead" room. I really think that the problem is more small size, square size (especially!) and limited placement options.

In reply to:

---

when you are getting your readings, do you have the meter at your primary listening location? Or is it as pictured in the image?

---

Andrew:
Actually, this was the only test where I placed the meter in anything other than the "sweet spot"! This was more of an experiment to see what differences would be read in moving the M60 variable distances from the wall. In this one case only, I placed the meter there to hopefully "take the room out" of the readings a bit. Didn't work!

The other readings I've taken of all the speakers show similarly wide swings, though at different places.

In reply to:

---

Actually like mwc said (except without the sub), I think the curves I would be most concerned about if it were my setup, would be the curves taken with the SPL meter in my primary listening location and both M60s running.

---

I honestly would not have thought that a reading with both M60s running would have been that beneficial. As mentioned above, I did meter them separately and put them on the same graph (their irregularities were somewhat consistent). I would have thought that metering them together would not really accomplish much (in my room anyway) as the speakers are not "centered" within a wall in the room and I would think that they could "smooth each other out", but that doesn't really help me when you have an acoustic guitar on left and a different one on right… does it?

In reply to:

---

Are you going to run another set with 3805 equalization to note any differences?

---

John:
I didn't for that one particular test because, as mentioned above, it wasn't being metered from my primary listening position. With the other ones I've done, yes, I've re-metered after some EQ and re-plotted on the same graph for "before and after". There's a little difference, but nothing significant as far as the charts go and just tonight I've done some listening with the EQ Off and on my custom setting, and it's close, but I think I just prefer it "off".

Like I said, I can't really make very significant changes via EQ as the Denon 3805, while offering full parametric capability with it's "auto" function, actually isn't that flexible in manual mode. You have fixed center frequencies and no "Q" adjustments, so you're always fudging it a bit!

First comment -- wow, that's a lot of testing

I haven't plotted response as carefully as you but I remember getting readings with almost 20dB swings. Your room might be a bit worse but not much worse.

Other random comments :

How big is the room ? I'm guessing one dimension is about 11 feet ?

Sorry if I missed this, but was the graph you posted with the meter as shown in the picture or in the sweet spot ?

Was the blanket positioned to intercept the first reflection off the walls to your listening position or was it really close to the speaker ?

Have you had a chance to convert the peak & valley frequencies to distances and go looking for those distances ? I was just working the numbers in my head during a particularly dull meeting today and if I didn't screw up you're looking for 4-1/2 foot and 11 foot distances, either wall to wall or front of speaker to wall.

IMO you definitely did the right thing measuring the speakers independently. You can average the readings and plot what you would get with both speakers turned on, but you can't separate the plots if you measure together.

Now I can see your chart.

And I'm sorry to say that is f*ed up. Wow. I have hardwood floors + rug, too, but I've never seen freq. resp. like that in all of my testing.

Dude.

In reply to:

---

wow, that's a lot of testing

---

Yeah, it is. I'm a "detail" person anyway, but this whole thing's gotten away from me. Once I started testing and seeing results like this, I keep going back with different options to try. Along the lines of what John said, I don't "tweak" my systems once they're set up. I won't explore different cables, etc… but these swings sure seem big and they're certainly and directly affecting the sound.

It's funny…. I mentioned in my original post that I am happy with my Axioms. If I'm happy with them now, how ecstatic could I be if I could tame this room a little? Further, sometimes I think "Well, John likes his Axioms, and Ray likes his Axioms, and Peter and Jack and Tom, yadda, yadda yadda…. BUT….. what if Tom "settled" on his sound and because of his room, isn't getting ANYwhere near the quality that Ray is getting? Since we each hear differently, and we're not hearing each others' systems, how do we know? How do I know that it couldn't just "come together" and I wouldn't love my Axioms 3X as much with a little tweaking?

In reply to:

---

How big is the room ? I'm guessing one dimension is about 11 feet ?

---

The room is 13' square, with an 8.5' foot ceiling.

In reply to:

---

was the graph you posted with the meter as shown in the picture or in the sweet spot ?

---

That particular graph was metered as shown in the photo… NOT in the sweetspot as all my others' were…

In reply to:

---

Was the blanket positioned to intercept the first reflection off the walls to your listening position or was it really close to the speaker ?

---

It was placed to diffuse that first reflection off both the floor and the wall….to that particular meter position.

In reply to:

---

Have you had a chance to convert the peak & valley frequencies to distances and go looking for those distances ?

---

No, I haven't. I don't know how to do this. Can you recommend an article? I come across "acoustics" articles pretty often, but not ones that explain the hard math and how to do the calculations….

In reply to:

---

you definitely did the right thing measuring the speakers independently. You can average the readings and plot what you would get with both speakers turned on, but you can't separate the plots if you measure together.

---

Good point; thanks. Yes, to me, the most important plot I've got is the one with the M60s metered independently, with the sub on, charted on the same graph. I can read them independently, or see the very obvious similar characteristics between them.

In reply to:

---

I'm sorry to say that is f*ed up. Wow. I have hardwood floors + rug, too, but I've never seen freq. resp. like that in all of my testing.

Dude.

---

Gee, now I'm depressed. Wait till you see how drunk I get at your Superbowl party!

OK, I've thought it over and worked through the issues several times. I believe your only option is to package all of your HT equipment in it's original boxes and send it to me as fast as you can! I'll perform an audio exorcism and return it to you. This sometimes takes several years, so be patient. Now, where is that packing tape?

While I recover from that spew, let me ask you something. If you had never started down th epath of reducing the audio to lines, squiggles and charts, would you have just enjoyed the audio? I mean, how does it really SOUND to you? Don't get me wrong, science and technology are good things, but why go looking for stuff to fret over if the speakers sound OK to you? STEP AWAY FROM THE METER WHILE YOU ARE STILL ABLE!

In reply to:

---

how does it really SOUND to you? Don't get me wrong, science and technology are good things, but why go looking for stuff to fret over if the speakers sound OK to you?

---

If I have to be completely honest, I'm happy. But not thrilled. And many of you are thrilled with your Axioms....so I wonder if I'm missing something.

The biggest variable between my system and others' is the room. I'm really not obsessing over this, I just want to be as constantly amazed and blown away with my sound as I am with my DLPs' picture everytime I turn that on.

I just think there's more there, and if I could tame the Wild Jungle Room I think I could be happier... and stop all this "setup" stuff that I've been working on for the last couple of months, and just get down to enjoying music and movies. I miss it. The "setting up" isn't fun after this long a time...

>>Can you recommend an article? I come across "acoustics" articles pretty often, but not ones that explain the hard math and how to do the calculations….

I'll look for something good. I learned what little I know from a Philips / DeForest "build your own speakers" book about 30 years ago which had a couple of really good chapters on room acoustics. I have not seen anything as comprehensive yet accessible since.

Here's what I remember. Speed of sound is about 1130 feet per second. For any given frequency, the wavelength is <speed of sound> / <frequency>. If the round-trip distance between the walls is 1 wavelength, or a multiple, you get a peak. If the round trip distance between the walls is 1/2 wavelength, or 3/2, or 5/2 etc..., you get a null (valley).

Let's look at your 13' dimension, or 26' round trip bouncing back and forth between the walls. One wavelength round trip gives you a peak -- 1130/26 is 43 Hz (43 hz signal has a 26 foot wavelength), so you should expect peaks at 43, 86, 129, 172 etc.

One-half wavelength (3/2, 5/2 etc..) gives you a null. If 26 feet is a half wavelength, then 52 feet is a full wavelength, corresponding to 22 hz. You should get nulls at 22 (1/2 wavelength), 66 (3/2), 110 (5/2), 154 (7/2).

OK, let's look at the chart. Yep, peaks around 43 and 86, and a hole around 66 Hz. Things rarely work out so cleanly, unless you have a square room

We need to do the same calculations for the 8.5 foot dimension but the 13 foot dimension will dominate since your room is square. The same concept can also be applied in a bunch of other ways as JohnK reminded me -- eg. if your speaker is 4 feet from a wall, or 8 feet round trip, you will get peaks at 141, 282, 423... and nulls at 70, 210, 350...

Even worse, if you have a reflection off wall, ceiling, floor then you're going to get peaks and nulls at frequencies where the difference in direct & reflecting path lengths is one or 1/2 wavelength.

Makes you want to just cower in the middle of the room with a big drink and good headphones, doesn't it ? The good news is that you only have to worry about a the lowest frequencies since at higher frequencies there are so many multiples that they all start to cancel each other out.

I have never played with bass traps, but from what reading I have done they sound like the best solution. Bass traps are basically custom-crafted big, flat boxes which absorb low frequencies to damp out the response peaks resulting from room modes. In theory they should be able to have some effect on the nulls (aka nodes) but nobody ever seems to talk about that.

Here's a first link to get you started. Google on "bass traps" and have fun. EQ will also work but with the drawbacks already mentioned -- you can't eq out a null, only a peak, and you need different eq for different locations in the room. Bass traps (in theory) treat the whole room.

http://www.realtraps.com/art_tuning.htm

Go get that drink now

Mark, I rarely agree with pmbuko, but I ran away screaming. I don't think you should ever go back into that room again.

So much for testing speakers in anechoic chambers!!

Try painting the door red.

That having been said, I don't think your room would look like a place of human habitation if you did what you must to damp those evil harmonics. I'm looking at what happens at 60, 250, 800 and 2.6K Hz, and wonder if you've been playing with any singularities?

Also interesting are those wild peaks at 50, 100, and 3K - 8K

I don't know how you'd "tune" that room - damp it at 50, 100 and 3K-8K, and pump it up at the other frequencies.

Maybe you should get an equalizer.

Ahh, here we go. Build your own bass traps. Electronic Musician was such a great magazine.

"There are many different types of bass trap designs, and this one is as effective as any. It is particularly well-suited for small rooms because the traps protrude very little--the thickest unit is only four inches deep. A two-foot width was chosen because it yields two traps from each sheet of plywood with no waste."

http://www.ethanwiner.com/basstrap.html

I don't think these would be so bad. They're big but they don't look like seven headed aliens or anything. Cubic footage is about the same as an EP600 and I don't think you would tell FedEx to take one of THOSE back if it showed up at your door.

I would be building a couple of these puppies right now if I wasn't moving in six months to a new house where the room dimensions were all run through a mode calculator before signing off the drawings

Might have to build them anyways. My room isn't much better than Mark's -- not square but pretty much 2:1 so same effect just a little less evil.

OK, I'm feeling a bit guilty because some of you (John, in particular) are taking time to help me through this, but basing much of it on a "test" graph, not one from the actual listening position. Remember that first graph was something of a test to try to analyze a difference in speaker distance from the wall and was measured with something more akin to a "close-mic" method. Some of the details that apply to that particular graph may not be indicative of what I'm getting from the sweetspot.

So…….

This morning I plotted a new graph (from the sweetspot) for the Left and Right M60s, with my sub engaged at a 100Hz crossover, but varying whether the receiver was set to "Large" or "Small".





I think it's interesting what you can see from a graph like this.

***You can tell that the right speaker, being close to a wall (in comparison to the "Left" being away from a sidewall) really interacts in the bass when set to "large".

***It's interesting that no matter what the setting, there are certain areas (like 44Hz and 70-80Hz) that have similar patterns whether the sub is handling them or the sub WITH the M60.

***I also found it interesting that the graphs for "small" and "large" settings, crossed over at 100Hz, don’t actually meet up until 200Hz. A more gradual slope on the upper end of the crossover than I would have thought.

I really have to go get some editing done for a client coming in this afternoon, but here's a promise: This weekend, I absolutely, positively will finish that gigantic article and put it all to bed. Except the MX-700. I may not get to that just 'cuz it drives Ray up the wall!

Thanks again to everyone- John, I printed your "math" post and will be reading it more carefully later today! Regarding the bass traps, I'm familiar with them, but I might have a WAF to consider there. I've read that a couple of large bags of towels, blankets, etc. in a corner of the room can work wonders- maybe I'll play with that a bit this weekend! Funny thing is, I have a corner behind the sectional where those would not even be visible!

That's a neat and beautiful installation, except, of course, for the speaker grills, which belong in the attic.

I am curious to see if having a auto calibrating system like the Pioneer Elite and Yamaha may correct the peaks and throughs. Maybe you can borrow a friends receiver which has those functions.
That being said I was wondering if those systems actually tame the peaks and level off the freq response for your rooms.

I remember the days when I was chastised by the long standing "experts" of this forum about using systems like my MCACC of the Elite or YPAO of the Yamaha that adjust the "supposedly" flat frequency curve these speaker inherit. Surely flat in an anocholic chamber but not in our regular rooms.
MarkSJohnson's listening room is pretty much not much different that the rest of your rooms. Maybe even most of you are actually NOT getting this supposedly "flat" frequency response from the Axioms. For those individuals who own and have used those Auto calibrating receivers get a nice flat response in thier HT/music environment.
It is sad that I normally don't join in any audio debates anymore due to closed minded "experts" that have not looked further into the benefits of new audio technology. As the saying goes 'Its hard to teach an old dog new tricks'
I applaud 2x4 for challenging their dogmas.

I'll tell you what, Saturn:
I'll run that test in the morning and post the results.

I have a Denon 3805 that has an auto-calibration function, and I've never liked the sound of the EQs that it provided (though it's a quick and easy way to let it set everything as a starting point).

It will be interesting to post an original plot overlaid with the plots of it's 3 or 4 auto EQ settings!

Great, Mark. Despite your lack of enthusiasm about the previous results, it should be interesting to see what the curves look like.

I miss ya!

Well, one thing is obvious from this graph. If you could set your speakers to "medium" the sound would be great

Seriously, it looks like the sub is not receiving signal when you are set to "large", ie the bass rolls off where the M60 rolls off, doesn't go any deeper. Is that how your receiver is set up, or is the M60 putting out a higher bass level than the sub ?

EDIT -- I don't suppose you could post a crude sketch of the room layout (doors etc..) so we could move your equipment around for you ? After staring at the graph for a few minutes the first thing that jumps out is that (as you mentioned) minimizing interaction with the walls seems to be a particularly good thing in your room.

First thing I would try is sticking the left M60 on the far side of the equipment alcove (nice setup there by the way) and shifting everything to the left a bit. Don't have it exactly centered on the wall but pretty close.

If that doesn't work for some reason is there another wall you could use ? At least all the walls are the same size...

In case you are starting to despair, you are doing exactly the right things here. A massive amount of testing and thinking, coupled with a significant amount of reading and discussing, resulting in some small, non-obvious change which in hindsight you can't believe you didn't try before

In reply to:

---

I'll tell you what, Saturn:
I'll run that test in the morning and post the results.

I have a Denon 3805 that has an auto-calibration function, and I've never liked the sound of the EQs that it provided (though it's a quick and easy way to let it set everything as a starting point).

It will be interesting to post an original plot overlaid with the plots of it's 3 or 4 auto EQ settings!

---

Mark: Maybe you do not like the sound of speakers that have a flat frequency response(after calibration).

I find it refreshing that you actually have laid out a real scenerio setup and quantified that the frequency response was NOT flat.

mwc: Hugs and kisses. I see that you have upped to the 1.6. Even after my Ushers .. I will be going back to planers. The midrange of planers still "to me" can't be beat by any speaker. When I sell my Ushers I will be getting the 1.6 or higher. Yes it does not have a flat frequency response. And yes it has a 60-80hz accentuated bass hump. But they sound phenominal.

To the self proclaimed "experts":
Hey boys...all those years you guys have been preaching to me and others that your Axioms has a flat frequency (yes in a controlled environment in Ottawa) has actually been bouncing all over the place. I'm not not an Axiom hater. I own and have used many of the models and love them. I just find lots of bias towards individuals who jump in and says the recent audition of Dynaudio, Magnepan, Revel speakers sound better than your Axioms. You get offended and give them a hard time. Yes they cost more two, three time more ... but its just what they observed.