Is it implicit that the speaker is clean to begin with? As NLD is sometimes used to spread/smooth frequency dependent issues, it may also make them worse.
I've got some cabinets around 28"W that can clearly demonstrate the audibility of... something to do with edges.At the outside edge, they have a shallow recess, and then an edge that protrudes a couple mm past the baffle. I suppose there is both diffraction and reflection happening. Just listening to one speaker, the sound seems to be coming from inside the cabinet. If you put strips of 2" insulation down the edges, different sounds suddenly have different perceived depth along a line through the cabinet. It's not subtle. The difference in response measurement is pretty subtle.
It could just be something "special" about my non-ideal cabinets, though. It was my first time messing with something so wide, and the first time I noticed such an obvious change from edge differences.
I encourage you to try your experiment blind. And report back the results.
I encourage YOU to try it blind with a wide speaker and report back the results. After doing it non-blind with one speaker, I did do it blind with two speakers A/B'd, which of course is to say not really blind and not entirely valid either. I wasn't interested enough to recruit someone to help do it with a single speaker.
After doing it non-blind with one speaker, I did do it blind with two speakers A/B'd, which of course is to say not really blind and not entirely valid either. I wasn't interested enough to recruit someone to help do it with a single speaker.
I want to be clear that there are nonlinearities that are audible, like inductance modulation in a woofer going all the way up to a compression driver (800 Hz or more). My point is that these things are all well know and solutions are available - so use them. Once this is done NLD is no longer an issue and we need to move on to the more intractable problems like directivity control.
I have never said that NLD is not audible in any loudspeaker - that is absurd.
IME front wall absorption (or absence) works well and sounds good. But you'll give up a lot of mid-bass energy. Maybe low bass, too, but that's not as noticeable as the midbass.
A combination of diffusion and absorption works best to my ear. It's part of the old LEDE debate.
One reason that I do not do quick and dirty experiments like this is because they do not prove anything. It just gets down to "I like this" - "no I like that" - it's an endless discussion.
When I do tests, I do simple test that are pure in what they are manipulating and they are done with experimental certainty (i.e. > 95% reliability). So while the results may not tell the whole story or be conclusive in and of themselves, the results are incontrovertible and any theory must be consistent with them.
Hence, signal modifications like diffraction ARE audible at high enough levels in both the effect and/or SPL. NLD is not. So when someone says that such and such sounds "distorted" I know that it is not likely to be NLD, but it "could be" diffraction.
I do not think that anyone can exclude diffraction as a significant effect.
After reading Toole, I don't know if there is much benefit to doing anything anyway. I know people have experienced an audible effect. It could also be one of those things that we can measure, and because we can get a marginally improved measurement, it gives us a sense of control over something, which we like.
I'd much rather spend my time on other things, such as the crossover, phase overlap, smoothness of response. These things have a highly audible effect on the sound.
I'd much rather spend my time on other things, such as the crossover, phase overlap, smoothness of response. These things have a highly audible effect on the sound.
Last edited:
Yes, and I do the same. But does that midbass energy which has manifested in an arbitrary farfield location have the need (or even the ability) to be a fully productive contribution anyway?
+1, and this is also what SL says.. he actually put recently a reflection calculator on his website where you can see the effect for each surface/distance.
Last edited:
Whether controlled Q is important or not depends on the goal. If you looking for a speaker to cover several seats, high DI in the horizontal plane wouldn't make sense. Neither, if the listener desires lateral contribution for a more spacious sound. A horn cannot offer that.
And personally I would much rather have a speaker dispersing wide with a uniform polar then a horn that has a collapsing polar that becomes narrower and narrower. In my case I'm going to treat the room anyway, so a uniform response is the most important aspect. The only designs I know of that can offer a uniform polar down to schroeder, withouth becoming too large and end up as PA speaker, is a corner horn and CBT. One with high Q, the other would low Q. Both have their strenghts.
And personally I would much rather have a speaker dispersing wide with a uniform polar then a horn that has a collapsing polar that becomes narrower and narrower. In my case I'm going to treat the room anyway, so a uniform response is the most important aspect. The only designs I know of that can offer a uniform polar down to schroeder, withouth becoming too large and end up as PA speaker, is a corner horn and CBT. One with high Q, the other would low Q. Both have their strenghts.
Depends on the goal, room and what room treatment that can be done.
I don't however want a speaker that sends all the energy to the frontwall. It makes no sense to me. It might measure well in an anechoic chamber, but in a living room that frontwall contribution is impossible to stop completely. A baffle is there for a reason.
Besides, a dipole with several drivers will not measure well in the vertical plane. Neither does it have great dynamics.
Diffraction messes up the frequency response and listening window consistency of the response. So even if you think the temporal aspects of diffraction are inaudible, it makes sense to do something about it.
Horses for courses. In general I'd say that if you have a large room with good acoustics (the right amount of broadband damping and diffusiveness, no flutters and significant resonances) and you have the ability to place the speakers far into the room, a conventional wide-dispersion speaker will be one of your best options. Getting good sound at mid an high frequencies is not that difficult in such a situation. Although I previously thought otherwise, I now agree with Toole that lateral reflections may not be that bad and might actually be beneficial, as long as they are sufficiently delayed.
If the room has less than ideal acoustic properties, if it is smaller or if loudspeakers have to be placed close to boundaries, some sort of controlled directivity is required. In my experience imaging becomes softer (which might be preferred by some) when lateral reflections are present, but as long as they are symmetrical, imaging as a whole does not suffer much - strong symmetrical reflections make for a larger sweet-spot at least. Asymmetrical lateral reflections do more harm than good in my experience, so in asymmetrical setups lateral directivity generally is a good thing. A strong and early front-wall reflection leads to coloration, same goes for the back-wall. The front-wall reflection can be dealt with by placing the speaker far from the wall or with strong forward directivity (preferably from the Schroeder-frequency up). The wall behind the listener needs to be far away or requires a wide-band damper.
Based on the research available I think ceiling- and floor reflections have a potentially negative effect (I like the Archimedes study, Soren Bech et al.). Those are difficult and/or expensive to deal with (wideband dampers, line-array, CBT) and I'm not sure if it is generally worthwhile.
I agree Keyser to a large degree, but I also think it depends on what controlled directivity you have. I'm not so certain anymore that a speaker that for example only has controlled directivity down to about 1200-1000 Hz and becomes wide below that is better in a small room then a speaker with wide disperson but uniform down or close down to schroeder. I used to think so, but experiences have made me question this.
I was surprised when I compared a waveguide speaker and CBT in a narrow room. I suspected the waveguide to do better since I had no sidewall treatment. But It actually didn't and sounded all together more colored. I believe the reason is because it only had controlled directivity down to a certain frequency, mainly in the treble, and below it didn't and thus you have quite strong coloration from the reflected energy. While the CBT, though dispersing wide at all frequencies, sounded very natural and not colored. It didn't portray a super precise image of course, but it didn't sounded colored the way the waveguide did. Maybe the ceiling reflections also contributed to the result in this room, but I did use some vertical treatment for the earliest reflections. I will do the comparison again later, this experience was convincing though and difference was not subtle to my ears.
So I believe it's very important to have a uniform polar over a large frequency area and not only in the treble/upper midrange. If you have that (preferable down to schroeder), which you can achieve with either a large free standing horn or a corner horn, I agree with you that it's best opion for a room that isn't ideal acoustically.
I was surprised when I compared a waveguide speaker and CBT in a narrow room. I suspected the waveguide to do better since I had no sidewall treatment. But It actually didn't and sounded all together more colored. I believe the reason is because it only had controlled directivity down to a certain frequency, mainly in the treble, and below it didn't and thus you have quite strong coloration from the reflected energy. While the CBT, though dispersing wide at all frequencies, sounded very natural and not colored. It didn't portray a super precise image of course, but it didn't sounded colored the way the waveguide did. Maybe the ceiling reflections also contributed to the result in this room, but I did use some vertical treatment for the earliest reflections. I will do the comparison again later, this experience was convincing though and difference was not subtle to my ears.
So I believe it's very important to have a uniform polar over a large frequency area and not only in the treble/upper midrange. If you have that (preferable down to schroeder), which you can achieve with either a large free standing horn or a corner horn, I agree with you that it's best opion for a room that isn't ideal acoustically.
Griesinger makes a strong case for why reflections below about 700 Hz are not important. There is no justification to say that directivity needs to go to Schroeder. It is not hard to control a high DI down to 700-800 Hz (which is not "treble"). Below that is difficult, but unimportant. Ceiling and floor reflections do need to be dealt with but those are best done in the room itself.
I don't think that people here realize that a direct radiating driver can never be CD once its size is comparable to a wavelength. I keep hearing people talk about CD dipoles - the Orion is not CD - not even close. It is not CD where it is most critical.
A lot of prescriptions for success listed above, but they mostly seem to be based on "experience" and not on science. I take them with a grain of salt.
Also, a CBT is not CD - it only has controlled directivity in the vertical plane, not the horizontal one. I know, nothing is really CD in a pure sense, but some designs come much closer than others.
I think that we all understand that lateral reflections enhance spaciousness but degrade imaging (unless they are > 10-20 ms.), but I'd rather have both and that can only be done by killing the early lateral reflections and enhancing the later ones. This can only be done with a high DI CD loudspeaker.
I don't think that people here realize that a direct radiating driver can never be CD once its size is comparable to a wavelength. I keep hearing people talk about CD dipoles - the Orion is not CD - not even close. It is not CD where it is most critical.
A lot of prescriptions for success listed above, but they mostly seem to be based on "experience" and not on science. I take them with a grain of salt.
Also, a CBT is not CD - it only has controlled directivity in the vertical plane, not the horizontal one. I know, nothing is really CD in a pure sense, but some designs come much closer than others.
I think that we all understand that lateral reflections enhance spaciousness but degrade imaging (unless they are > 10-20 ms.), but I'd rather have both and that can only be done by killing the early lateral reflections and enhancing the later ones. This can only be done with a high DI CD loudspeaker.
Last edited:
I haven't read the Griesinger study, but from what I've heard from other he was basical concluding that we are less sensitive to reflections below a certain point. And that's something we already knew and is not the same as unimportant. My subjective experience tells me otherwise too. It does matter.
The floor reflection is not easily dealt with using treatment. Practically impossible to deal with it in a good way for most. Like earlier mentioned; any floor treatment typically exhibits a such a high frequency profile as to not be effective as broadband absorption for a traditional speaker. It will function as a filter. That's a compromise.
Having both spaciousness and good imaging doesn't require a high DI CD speaker. That can be accomplished with other designs too with correct treatment. It may also be easier then with a horn that is crossed in front of the listener because the latter will send much of the energy in a ping-pong matter between the sidewalls before it reaches the rear and thus comes back at a much lower level. Speakers that are pointed to the rear of the room will easier attain a stronger return. If it disperses wide, redirection can be used. Of course, having a horn speaker with high DI that is pointing to the back of the room and not crossed in front of the listener makes it even easier.
The floor reflection is not easily dealt with using treatment. Practically impossible to deal with it in a good way for most. Like earlier mentioned; any floor treatment typically exhibits a such a high frequency profile as to not be effective as broadband absorption for a traditional speaker. It will function as a filter. That's a compromise.
Having both spaciousness and good imaging doesn't require a high DI CD speaker. That can be accomplished with other designs too with correct treatment. It may also be easier then with a horn that is crossed in front of the listener because the latter will send much of the energy in a ping-pong matter between the sidewalls before it reaches the rear and thus comes back at a much lower level. Speakers that are pointed to the rear of the room will easier attain a stronger return. If it disperses wide, redirection can be used. Of course, having a horn speaker with high DI that is pointing to the back of the room and not crossed in front of the listener makes it even easier.
"less sensitive", "unimportant", pretty much semantic to me, but that's OK I'll go with your wording - "less sensitive". The implication is the same either way.
What aspect of any of this is not a compromise? Its all about compromise and making the better ones. But to do this you have to know what aspects are the more important. There is nothing "ideal" about loudspeakers in small rooms.
Again, its all compromises and practicality. In a small room I do not see any other practical solutions - certainly none as efficient as a high DI CD design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.