heh heh heh
It's all part of an evil plan to convince everyone they can save money and still get outstanding sound, and not have to get a divorce. (Well, that's optional.)
It won't solve ALL our problems. But yes, a good OB can help provide pretty good sound in even lousy rooms. And really good sound in good rooms. And superlative sound in great rooms. While a good conventional speaker may be somewhat limited to only the really good rooms and better category.
Oh but wait, I'm terribly sorry.... I'm not allowed to say that without providing the required 80 page white paper. I take it all back...😛
It's all part of an evil plan to convince everyone they can save money and still get outstanding sound, and not have to get a divorce. (Well, that's optional.)
It won't solve ALL our problems. But yes, a good OB can help provide pretty good sound in even lousy rooms. And really good sound in good rooms. And superlative sound in great rooms. While a good conventional speaker may be somewhat limited to only the really good rooms and better category.
Oh but wait, I'm terribly sorry.... I'm not allowed to say that without providing the required 80 page white paper. I take it all back...😛
Fun test
Interesting test. I scored 86.6% which is supposed to be very good according to their criteria. It shows if we can distinguish between complex notes and sequences and then remember them in the short term. Aside from being a memory test for music, how would you propose this relates to our ability to notice resonances?
Interesting test. I scored 86.6% which is supposed to be very good according to their criteria. It shows if we can distinguish between complex notes and sequences and then remember them in the short term. Aside from being a memory test for music, how would you propose this relates to our ability to notice resonances?
Below Schröder frequency 1/3 is too coarse. Ask any of the "veterans".
You didn't show any data for a boxed speaker, so there's nothing that could be compared.
Please let me know when you post that data. Or, even better, upload the FuzzMeasure file(s).
I'm not interested in averaged data. I'm interested how similar the response is across a certain area and if the response is equalizable (minimum phase behavior). Averaged data is - like 1/3 smoothing - just sugar-coating.
Hope you will back up your claim one day. Until then it's simple conjecture.
One single room would be a good start. Your data IS NOT a good indicator because a) just a single data point, b) smoothing and c) missing data for the case you compare against.
OBs sound lite, however this does not mean they provide an accurate reproduction for all sounds.
You may have come across the following two threads where I am trying to come up with a speaker system that mimics the actual sound source accurately.
For realistic reproduction, speakers employing the gentle cue are necessary. The gentle cue is a combination of techniques that uses horn based speakers, bass reflex and OB.
directly, it doesn't. indirectly it might show some people that they shouldn't be trusting their ears as much as they think they can.
i scored 66. my mother scored 100% and she hasn't ever held a musical instrument in her life. Also, Leviathan, guitarist for the demogroup Andromeda Software Development (ASD) got a perfect score. my father, a 30+ years violinist, something like 60%.
It's a weird test.
Not only Toole but virtually anybody researching low frequency sound reproduction. All recent AES papers concerned with low frequency reproduction stress the importance of high resolution data.
Not really.
I did in my reference mains with a cabinet using very minimal bracing, CLD panels for baffle and the 1-piece curved side/back, and carpet padding (or whatever Tannoy used to line the cabinets of the 12 DMT II's that became my mains; I just took the stuffing out of the factory cabs) lining the sides and baffle. The resulting cabinets are light (WITH a 12" Dual Concentric and heavy passive crossover with poly caps and air-core inductors, ~41lbs for a cabinet 26.5" high, 16.5" wide, and 9" deep), rigid, and just do their job. (Though in retrospect I should've gotten the baffle and top/bottom in either glossy cherry, or espresso-stained bamboo rather than similar grey paint to the cabs of the donor speakers. But that's an aesthetic, not acoustic, issue.)
The "R&D" consisted of about 15min of thinking about what I wanted to do, and e-mailing my thoughts to the gentleman who built the cabs for me (Nathan Funk in western Canada). And the cost was quite reasonable as well.
Besides, minimizing the diffraction signature is more perceptually important. Unlike "box colorations," diffraction can be both measured, and heard in blind listening comparisons. And diffraction is a bigger problem in OB than in boxes, because a thin baffle necessarily has abrupt surface changes, and there's also the structure on the back of the midrange to consider.
Sort of.
I agree that in the upper part of the modal region, pairs of OB speakers properly placed tend to be better than box speakers properly placed. And pairs of OB speakers are markedly worse than a properly-implemented box solution (multiple subwoofers along with mains stout enough to play full range).
Also, as your measurements show, one can avoid the problems of room modes in the lower half of the modal region and in the 1st mode region simply by designing the system to not be good enough to excite those modes in the first place!
While I agree with that statement, our respective notions of "room friendly" differ markedly.
If you have to shove 'em out in the middle of the bloody room, as one does for most OB speakers - one can make a partial exception for the Gradient Revolutions here, due to the cardoid mid-loading and rotatable bass column - they're not in my view "room friendly."
I'm of the view that audio is best heard and not seen.
The best fit for that goal, in most rooms, is cabinets either built into, or resting on, the front wall. While it's hard to find commercial speakers that are a good fit for that (some of David Smith's old Snell designs seem well-designed for that use, but I can't recall any others), this is DIY so the first premise should be "design the speakers to fit the actual room in which they will be used." That includes, of course, a directivity pattern that's well-suited to the room. That is to say, relatively narrow if the room is very reflective, and perhaps wider if the room is relatively dead. And, it should go without saying, the power response should be consistent, perhaps with some off-axis holes but without off-axis flares.
Also, care should be taken to minimize diffraction, such as by moving equipment away from between the speakers, moving tables to the sides rather than between the listening position and the front loudspeakers, etc.
Any system with two speakers, no matter how good they are, is going to be low-fidelity in the model region and below. (See some measurements of full-range speakers here here.) Even with your EQ, they'll be OK at the primary listening spot - and worse most other places in the room.
That's why one needs multiple subwoofers for a high-fidelity system. Fortunately, such subs don't need to be that big, and careful room design can help in concealing them. I have three subs in my current listening room. One is fairly large (65L closed box holding an Aura NS12-794-4a) but the others are very small (15L closed box holding an Aura NS10-794-4a and 10L closed box holding a Peerless XLS10). None of them are visible.
Last edited:
Agreed.
Given that notches in bass response due to boundary cancellations and room modes can be very narrow and very deep, 1/3rd octave smoothed bass measurements are pretty much useless, as they will completely hide such notches.
As bass notes tend to be discrete frequencies it often happens that a particular note will fall into such a notch and go MIA, which clearly sounds different to what the 1/3rd octave smoothed measurement would indicate.
The sudden phase reversals that often occur around a deep notch are also usually non minimum phase resulting in a huge spike in group delay which may have some audible significance...again we don't want to just smooth this information out.
As for 1/3rd octave smoothing at higher frequencies, it's not particularly valid there either IMHO. If the only objective is to assess perceived tonal balance based on critical band smoothing, 1/3rd octave has some merit, however if we're using frequency response as a measure of sound quality, only narrowband un-smoothed data (1/24th octave or narrower) has any meaning.
It's a lack of narrowband smoothness at high frequencies that causes "harshness" for example, and 1/3rd octave smoothing completely obliterates this information...
Why does narrowband non-smoothness (narrow peaks and dips) that doesn't show up in 1/3rd octave averaging matter if its narrower than our critical bands ?
I'm not entirely sure, perhaps there has been research on this, but if I had to guess from my experiences in taming and correcting driver resonances both with damping and EQ, I would say that we notice two things - ringing from high Q resonances in the time domain, and frequency modulation (vibrato etc) being converted into amplitude modulation by the narrow steep slopes in the peaky response...
The only time I ever use 1/3rd octave smoothing is taking a power response measurement in room, for anything else I always look at the raw un-smoothed data, warts and all...
Last edited:
Member
Joined 2009
Paid Member
I'm trying to keep up with this all but there's lots of stuff here that I can't relate to box colourations.
What I'm seeing is opinions that an OB, with lack of a box moves the 'issues' to the room. So there's no free lunch. And when you are listening from further away from the OB, where direct radiation from the drivers no longer dominates, you are going to experience a lot of reflected sound which won't have the same spectral content. There are benefits but the whole situation becomes more complex to optimize. StigErik 'solved' this problem with significant room treatments, absorbing as much of the reflected back wave as possible. This makes it an IB rather than an OB ?
So, what does a box need to look like so that it behaves more like an IB than a box. Simplest is to make it huge - the room comes with the speaker drivers. Is it enough to make it large, or does it have to be aperiodic - ruling out bass reflex ?
What I'm seeing is opinions that an OB, with lack of a box moves the 'issues' to the room. So there's no free lunch. And when you are listening from further away from the OB, where direct radiation from the drivers no longer dominates, you are going to experience a lot of reflected sound which won't have the same spectral content. There are benefits but the whole situation becomes more complex to optimize. StigErik 'solved' this problem with significant room treatments, absorbing as much of the reflected back wave as possible. This makes it an IB rather than an OB ?
So, what does a box need to look like so that it behaves more like an IB than a box. Simplest is to make it huge - the room comes with the speaker drivers. Is it enough to make it large, or does it have to be aperiodic - ruling out bass reflex ?
Member
Joined 2009
Paid Member
Last edited:
There is a lot of testosterone in the air.
My first question is, do "box colourations" equate with "sounding boxy", or is this a syllogistic fallacy?
Again, this seems to be begging the question. You did suggest that it might be simply monopole/dipole preferences in the OP. Has anyone done comparisons of boxed monopoles versus boxed dipoles? Or, perhaps more to the point, OBs compared to boxed dipoles?
Not exactly
Precisely because of this issue in the bass, several designers have attempted OB designs where the directivity is very uniform, from top to bottom, to further reduce the effect of the room reflection issue. One of the designs that meet this goal brilliantly is the Nao series from Music and Design. There's some serious design chops over there!
My own TimePulse Lambda OB design is also fairly good this way, although there are zones where the directivity is not quite as constant as the Nao units. Still it's good enough to be much better than most, quite acceptable even.... and there are other factors, such as being time-coherent over its entire coverage zone, which I think are equally or maybe even more important, and make it a formidable competitor. The in-room performance vs anechoic shows a great degree of uniformity from 100Hz up in most rooms.
So, no, OB systems don't necessarily have more issues with reflected sound, I would say they have less when it's a good design. When they offer more constant directivity top to bottom than even very good CD designs such as the Gedlees, IMO they become LESS complex to optimize because there are less resonant modes and reflections in the bass and mid-bass ranges.
And yes, there are clearly some here who disagree with this position, but I wonder if they've actually experimented with open baffles, especially ones with a more constant directivity. They demand proof better than what I've offered so far, and for that I think I will simply refer them to Linkwitz' work generally and more specifically that of John Kreskovsky at Music and Design. There are some really great white papers on his site.
IMO the issue is pretty well proven, but then again that's the beauty of opinions, no one is forced to accept them.
There is a point where the reflected content is huge with just about ANY speaker, and because of that it is good to have uniform directivity, whether from being totally omni or from having a restricted directivity pattern. Most conventional speakers have a lot of variability in their dispersion patterns, which makes them far worse in terms of the spectral content of the reflected sound. That's why Earl Geddes designed his speakers with waveguides to restrict the directivity and make it more constant. However after a certain point, the directivity in those designs start "fanning out" in the bass range, where the pattern looks nearly omnidirectional.
Precisely because of this issue in the bass, several designers have attempted OB designs where the directivity is very uniform, from top to bottom, to further reduce the effect of the room reflection issue. One of the designs that meet this goal brilliantly is the Nao series from Music and Design. There's some serious design chops over there!
My own TimePulse Lambda OB design is also fairly good this way, although there are zones where the directivity is not quite as constant as the Nao units. Still it's good enough to be much better than most, quite acceptable even.... and there are other factors, such as being time-coherent over its entire coverage zone, which I think are equally or maybe even more important, and make it a formidable competitor. The in-room performance vs anechoic shows a great degree of uniformity from 100Hz up in most rooms.
So, no, OB systems don't necessarily have more issues with reflected sound, I would say they have less when it's a good design. When they offer more constant directivity top to bottom than even very good CD designs such as the Gedlees, IMO they become LESS complex to optimize because there are less resonant modes and reflections in the bass and mid-bass ranges.
And yes, there are clearly some here who disagree with this position, but I wonder if they've actually experimented with open baffles, especially ones with a more constant directivity. They demand proof better than what I've offered so far, and for that I think I will simply refer them to Linkwitz' work generally and more specifically that of John Kreskovsky at Music and Design. There are some really great white papers on his site.
IMO the issue is pretty well proven, but then again that's the beauty of opinions, no one is forced to accept them.
Last edited:
That's an interesting question. I'm not aware of any real boxed dipoles. (I think we can all safely reject "surround" speakers with out-of-phase drivers on opposite sides of a trapezoid as "dipoles.")
I'm not even sure it's possible, above the bass, to have a "boxed dipole," because the distance between the drivers on either side up high is going to be large compared to the wavelength. Maybe in practice it can be made to work, I don't know. I'm not familiar with any effort to try to make it work, though.
You're conflating two very different concepts here: "room reflections," and "room modes." Neither one of which has anything to do with this audiophool myth of "box coloration," which has been shown in at least two blind tests to be a total non-factor. Then again, I can understand why a wire marketer would defend any old audiophool myth s/he can get her/his hands on...
Furthermore, your assumption about the desired pattern for speakers in a domestic living room is unsupported by blind preference tests, which generally show a listeners prefer speakers with fairly uniform coverage (though holes, as opposed to fiares, in the power response seem perceptually benign) and a pattern that broadens down low.
My listening experience tells me that my preference, at least, is a consistent with the available data. And, for the record, my own experience with OB bass dates back about 15 years, when Doug Purl (a name familiar to fellow old Basslisters) persuaded me to try it. I did, found it wanting compared to modern best practices for bass (multiple subwoofers distributed through the room along with mains playing full-range, the subs playing as high as one can play them without causing localization issues), and moved on to the modern approach.
Though there are differences between OB and monopole bass due to the radiation pattern, and therefore people can reasonably have different preferences. Unlike this myth of "box coloration," which has no actual support. (Assuming fairly minimal competence in box design, and making an exception for planar drivers that are measurably adversely impacted by backwave reflections.)
I share your admiration of Dr. Kreskovsky's published work, and one speaker I would really like to hear is his NaO Note. (Even though I would never put speakers that require placement well into the room in my own home.) But one must also note that Dr. K doesn't pursue dipole bass radiation, but rather quasi-cardoid.
As soon as you're throwing a backwave - one contaminated by, inter alia, diffraction off of the midrange basket, motor structure, and baffle cutout, the end effect being that one will never get a true dipole short of using a symmetrical driver such as a BG planar anyway - out into the room, you are by definition creating "more issues with reflected sound" than a one gets with speaker that focuses its radiated energy where reflections are an actually relevant issue (midrange and up) in a slice of the forward hemisphere.
Am I saying that "room reflections" are a non-issue at, say, 500Hz? To put it simply, yes.
Alternately, as your own FuzzMeasure graphs show, your OB speakers simply have no bass to speak of. As one would expect by looking at them; a cheap short-throw 15 is not going to do much when left exposed in a room. They just don't have the volume displacement. Down low...I think it's an open question whether your OB speaker or a tiny bookshelf like the NHT Absolute Zero would win a bass-off. 😉
Last edited:
No bass?
Along with that I have a personal request: do you think you could drop the dismissive tone?
Dissing people as merely being wire marketers only shows uninformed prejudice, and takes away from what could be a valid technical discussion. Can we keep it respectful, please? Thank you for that.
Now, on to the technical stuff.
I get your point about distributed subs down low, and I agree it's a great way. I've worked with many variants, including the Geddes method. And, as I mentioned in one of my previous posts, my preferred cost-no-object approach is to use 4 distributed subs, two in front, two in back and adjust the arrival times of the front to match the two main channels, then to further delay the rear ones in such a way as to be exactly synchronous (at the back wall) with the pressure wave generated by the front woofers - but run them in reverse phase so as to absorb said pressure wave. Kreskovsky has discussed this technique using 2 subs, one fron and one rear, I prefer the results with 4. And yes, it is useful to run them up fairly high if the subs lend themselves to it, maybe as much as 120Hz or even 150Hz.
Now, your affirmation that my speaker doesn't go down low? It does a pretty good job actually, again look at the anechoic response, flat to 39Hz. And yes, I know the limited excursion of the 15" I've chosen is "only" 5mm. I'll tell you honestly, prior to building it, I too expected the system to sound dynamically limited in the bass.....and yet.... this is most definitely not the case, even with very large scale symphonic music. I guess there's just not that much deep bass in the majority of music! These things can play really LOUD and still stay very clean and impactful, no sign of dynamic compression at all.
If cost were no object would I recommend a subwoofer array with it? YES! but only for that small subset of the musical repertoire that really requires the low low bass underpinnings to make it work.
Otherwise it works brilliantly, and I am the first to admit surprise that it can carry itself as well as it does in the bass, and I mean at very loud levels, playing back full-on drumkits at amazing SPL in a moderate size room.
As for your challenge, I'd love to see your expression if only you could compare with the NHT - or practically anything else for that matter.
Respectfully submitted.
Pallas, some of your technical points are well worth discussing.
Along with that I have a personal request: do you think you could drop the dismissive tone?
Dissing people as merely being wire marketers only shows uninformed prejudice, and takes away from what could be a valid technical discussion. Can we keep it respectful, please? Thank you for that.
Now, on to the technical stuff.
I get your point about distributed subs down low, and I agree it's a great way. I've worked with many variants, including the Geddes method. And, as I mentioned in one of my previous posts, my preferred cost-no-object approach is to use 4 distributed subs, two in front, two in back and adjust the arrival times of the front to match the two main channels, then to further delay the rear ones in such a way as to be exactly synchronous (at the back wall) with the pressure wave generated by the front woofers - but run them in reverse phase so as to absorb said pressure wave. Kreskovsky has discussed this technique using 2 subs, one fron and one rear, I prefer the results with 4. And yes, it is useful to run them up fairly high if the subs lend themselves to it, maybe as much as 120Hz or even 150Hz.
Now, your affirmation that my speaker doesn't go down low? It does a pretty good job actually, again look at the anechoic response, flat to 39Hz. And yes, I know the limited excursion of the 15" I've chosen is "only" 5mm. I'll tell you honestly, prior to building it, I too expected the system to sound dynamically limited in the bass.....and yet.... this is most definitely not the case, even with very large scale symphonic music. I guess there's just not that much deep bass in the majority of music! These things can play really LOUD and still stay very clean and impactful, no sign of dynamic compression at all.
If cost were no object would I recommend a subwoofer array with it? YES! but only for that small subset of the musical repertoire that really requires the low low bass underpinnings to make it work.
Otherwise it works brilliantly, and I am the first to admit surprise that it can carry itself as well as it does in the bass, and I mean at very loud levels, playing back full-on drumkits at amazing SPL in a moderate size room.
As for your challenge, I'd love to see your expression if only you could compare with the NHT - or practically anything else for that matter.
Respectfully submitted.
Attachments
Last edited: