Power response, acoustics and how a "flat on axis" design sounds in a room

[fjhuerta]

Hi all,

I've been playing around with one of my designs (the Monoliths): a 2.5 way with Peerless 830884 drivers and a SEAS 27TBFC on top. My parents have them, and they asked for a bit more top end sparkle (I assume due to their hearing loss).

Anyway, after further refining the design, I took a measurement at 1.8 meters, and it was dead flat on axis. Off axis (starting at 15°) a small "flare" centered at around 4 KHz was visible (no big deal, I thought).

But, when I took in-room measurements, the small flare was now a +4 dB hump, centered at that same frequency. The speaker was pretty good except for that.

Now, the room I'm using the speakers in is far from optimal. Hardwood floors, concrete walls covered with hardwood, tile floors around, bare walls, etc.

I did an experiment. I took even more in room measurements with an RTA. At 1.5 meters, on axis, the speakers were flat (except for some floor bounce, which cancelled some of the bass). As I kept walking backwards, the 4 KHz bump became more and more evident. Then I took the 1.5 meters on axis measurement, and noticed that walking to the center of the room, at that same distance, also showed the flare.

Here are some questions I need some help with. Some are dumb, some are really dumb, and some are puzzling.

1) Is the 4 KHz bump to be expected? Is it a function of the room?
2) How do off-axis flares reflect on an RTA in room? Would I still see the flares, even if the room was properly treated?
3) If the speakers are facing forwards, and I'm sitting in the center of a triangle between them, am I listening on-axis, or off axis? (Please don't laugh, I started thinking about this and I don't know. I'd suppose I'm off-axis, but I'm not sure).
4) Since my wife will never let me treat the room, would a bit of EQ help things?

Thanks a lot for your help!

[norcad]

Im no expert in this, and Im doing som testing myself to find some answers.

The "problem" you have at 4khz could be phase related. When you measure onaxis on the tweeter at a specific length, say 1.5m, the tweeter and midrange are in phase, but when you go further away from the speaker the length from the midrange vs tweeter will be different, and the drivers are not in phase anymore.
At high frequencies like this it is rather easy to threat the room. Som heavy curtains and perhaps a carpet will do the trick. Even a bookshelf at the right location can help remarkely well.

[Dennis H]

Javier, I see you were in this thread over at PE.

http://techtalk.parts-express.com/showthread.php?t=203023&page=2

I liked what DDF had to say about the 4K bump. It shows up at 45 degrees off axis because the tweeter still isn't very directive up there. He thinks 45 is the best measurement for how the speaker 'sounds' because that's about the axis for early reflections off the wall.

[fjhuerta]

Im no expert in this, and Im doing som testing myself to find some answers.

The "problem" you have at 4khz could be phase related. When you measure onaxis on the tweeter at a specific length, say 1.5m, the tweeter and midrange are in phase, but when you go further away from the speaker the length from the midrange vs tweeter will be different, and the drivers are not in phase anymore.
At high frequencies like this it is rather easy to threat the room. Som heavy curtains and perhaps a carpet will do the trick. Even a bookshelf at the right location can help remarkely well.

I don't think it's phase related. The crossover is set at 1.35 KHz. At 4 KHz, the midrange output is negligible.

I believe it has something to do with the baffle shape. I tried two different speakers in the room, and measured them (a pair of Energy C3s and the Miniliths), and both are very flat. So the room is not helping matters, but it's not responsible for the 4 KHz hump.

[fjhuerta]

Javier, I see you were in this thread over at PE.

http://techtalk.parts-express.com/showthread.php?t=203023&page=2

I liked what DDF had to say about the 4K bump. It shows up at 45 degrees off axis because the tweeter still isn't very directive up there. He thinks 45 is the best measurement for how the speaker 'sounds' because that's about the axis for early reflections off the wall.

You mean this paragraph?

I did the LR, Butterworth, Bessel etc power response calculations for myself in the '80s but found they never reflect the tonal balance I think I hear well enough. The 30 to 50 degree off axis response does, at least to my ears. I found it's usually best to use the lateral 45 degree response in most cases, as an indication of tonal balance in room. For a typical 2 way with the typical 1.5 to 2.5 kHz xover, a big bump in the response at 3 to 5 kHz is easily seen with high order xovers, as the tweeter comes in. With an overdamped (gentle transition), this is much improved.

Given that the first side wall reflection plays such a strong part in the tonal balance of the speaker due to its low delay, it makes sense to me that the 30 to 50 degree offset response is so important in smaller and livelier rooms.

That's interesting. I wonder why a big bump is there at said frequencies, since the transfer function of a higher order crossover seems to be smooth at that point.

One thing seems to be happening throughout my very empirical measurements - the smoother the drivers, and the less steep the crossovers, the better power response (at least in my room), apparently. I have to make a 2nd order speaker to test this theory, but I've seen that poly / paper speakers are always smoother than my limited range of metal cone designs.

[fbov]

Javier,
The answer is in Toole's book. He spent 40 years, post-doctoral, investigating how speakers sound, first for NRC Canada, then for Harman International. He talks at length about what matters to human ears, including the importance of off-axis reflections to human perception. A very good read that any audiophile who intends to place speakers in a room should understand, before they blame the speakers for what room placement has done to their sound.

Have fun,
Frank

[wettou]

Javier, The answer is in Toole's book. Frank

Frank what book is that?

[fjhuerta]

Frank what book is that?

+1!!!

Is this the one?

http://www.amazon.com/Sound-Reproduction-Acoustics-Psychoacoustics-Loudspeakers/dp/0240520092/ref=sr_1_7?ie=UTF8&s=books&qid=1234807652&sr=1-7

[---k---]

Yes, that is the one.

[wettou]

Yes, that is the one.

Cool I am buying it today

[fbov]

Yes, that's it, and if you buy it today, plan on a couple weeks reading.

It's written like an introductory text book with lots of data. It's a challenge to understand just what an experiment is evaluating, but once you do, his conclusions are consistent, supportable and unbiased to the extent that he tells you what he's using it for, typically in home audio applications. Very real world, up to and including his home and his own ears - you can feel the disappointment as recalls finding out he was too old to be an audio test subject.

Have fun,
Frank

[Dennis H]

If you like Toole, he had a big influence on the Revel design philosophy.

[jkrutke]

My one comment for this thread is that hardwood floors hurt audio quality bad. That floor reflection is there with ripples much higher in frequency than you would normally deal with in a carpeted room. In a well carpeted room you often only get a minor droop in the 300-500 hz range with the higher frequency ripples well damped out.

Throw in close side walls and things get worse and even more unpredictable as to where the peaks and valleys show up for different listening (or mic) positions.

A hardwood floor may actually be one situation where a MTM with a relatively high crossover point benefits the in-room response. vertical directivity is your answer, and for side walls, maybe you can find some classy looking wall hanging options that look good enough to not mess up the decor. Also, if you are close to side walls, toeing in the speakers to a focal point a couple feet in front of the listener sometimes helps.

As a side note, I pretty much design for carpeted rooms. I really should add that note to all my posted projects, but I keep forgetting. Make no mistake though, I think hardwood floors are beautiful and worth the audio problems. If my house had that in the listening room, most of my designs would be a little different in tonality and focus on directivity more.

[jnickrand]

Javier,
The answer is in Toole's book. He spent 40 years, post-doctoral, investigating how speakers sound, first for NRC Canada, then for Harman International. He talks at length about what matters to human ears, including the importance of off-axis reflections to human perception. A very good read that any audiophile who intends to place speakers in a room should understand, before they blame the speakers for what room placement has done to their sound.

Have fun,
Frank

If the answers are in Toole's book, and you have the book, perhaps you can answer the questions for both Javier and the rest of us who are curious as to what the answers are.

[fbov]

This would assume that one can absorb a lifetime's worth of scientific investigation and accoustic expertise by reading a book. I'm not finished my first read through, and then I need to practice what I've learned before I can do anything for someone else.

Javier, however, is far more knowledgable than I to start with, and he's in the actual room. I'm hoping to learn from his experience.

HAve fun,
Frank

[Paul W]

Javier,
My money is on the DDF explaination Dennis pointed out.

A horizontal off-axis flare above crossover is normal when transitioning from the woofer to tweeter in a 2-way. The relatively large woofer becomes very directional just below crossover, the relatively small tweeter does not become substantially directional until well above crossover...that's what causes the flare. It is also normal for off-axis response to significantly influence in-room measurements in a lively room such as this one.

Dave's (DDF) solution for this is to use slow slopes to further overlap the directional characteristics of the two drivers. Tweeter waveguides, small mid 3-way's, etc, also minimize the abrupt change in directivity. You say you can't treat the room but you can you orient the speakers to minimize side and front wall reflections???
Paul

[brianpowers27]

Power Response

I think what you are seeing is the power response(Directionality) of that system.

Below 4khz the system sags both due to the directionality of the woofer, phase interaction between the woofer and tweeter and or room interactions. Above 6khz the system likely sags due to the directionality of the tweeter. It is very easy to see this high frequency tweeter droop when using >=1" dome tweeters.

There is a thread at the PE forum regarding the use of super tweeters. This can be a design option that is used to compensate for the high directionality of larger domed tweeters.

I am starting to agree with the concept of sampling the system at 45 degrees off axis. This does paint a clearer overall picture, in many cases.