Hey AllenB, nice to see you provide some psychoacoustics-based arguments in favor of a high crossover point between mono sub and mains. I have some more of the same.
DISCLAIMER (short summary, read the rest for the details):
A) I'm not claiming that humans can't localize below 500 Hz, they can, you can easily localize sine waves at least down to 80 Hz
B) This doen't matter, because music is not made of single, unrelated sine waves
C) The rest of the music above will override conflicting directional cues derived from low frequencies
So here are my thoughts on stereo separation in the kick bass range:
Well, I know my opinion is somewhat controversial, and I'm going a bit out on a limb. (I'm sure some people will shoot my ideas down) Take it with a grain of salt, it's definitely not the gospel truth. Nevertheless it is based on real life experience.
I've been running stereo 4-way systems on 6 output DSP crossovers (namely the Behringer DCX) for a long time. A lot of people on a budget are using this trick. For full 4-way stereo operation you normally need a processor with 8 outputs. Basically you run sub and kick in mono, and mids and highs in stereo. This way you need only 6 outputs. Crossover frequencies were usually 80 Hz and 160 - 200 Hz.
In my experience this works well. I haven't noticed any detriment to the stereo separation with this approach. I believe the reason for this is the fact that above 200 Hz there are another 6,5 octaves of information containing localization cues.
(sorry about the psychoacoustics blurb to those familiar with this)
Human hearing works based on two mechanisms, interaural level difference (ILD) and interaural time difference (ITD). ILD is based on level differences caused by sound absorption and shadowing of high frequencies by the human head. ITD is based on the phase difference between the two ears, and is effective below 1500 Hz.
https://en.wikipedia.org/wiki/Interaural_time_difference
Here is a research report I found about ITD/ILD trading
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3003727/
one of the conclusions from this report:
"Rivalry experiments with opposing interaural level differences (ILDs) found that the trading ratio between ITD and ILD increasingly favored the ILD as coherence decreased."
this would confirm my observation, that in practice stereo imaging is not impaired by having the below 200 Hz range played back in mono, because the human brain, presented with incoherent (=conflicting) ITD and ILD, will favor the ILD cues for localization.
Finally, as a comic relief I present to you this funny video about another psychoacoustic effect that will seriously tease your brain:
https://auditoryneuroscience.com/McGurkEffect
DISCLAIMER (short summary, read the rest for the details):
A) I'm not claiming that humans can't localize below 500 Hz, they can, you can easily localize sine waves at least down to 80 Hz
B) This doen't matter, because music is not made of single, unrelated sine waves
C) The rest of the music above will override conflicting directional cues derived from low frequencies
So here are my thoughts on stereo separation in the kick bass range:
Well, I know my opinion is somewhat controversial, and I'm going a bit out on a limb. (I'm sure some people will shoot my ideas down) Take it with a grain of salt, it's definitely not the gospel truth. Nevertheless it is based on real life experience.
I've been running stereo 4-way systems on 6 output DSP crossovers (namely the Behringer DCX) for a long time. A lot of people on a budget are using this trick. For full 4-way stereo operation you normally need a processor with 8 outputs. Basically you run sub and kick in mono, and mids and highs in stereo. This way you need only 6 outputs. Crossover frequencies were usually 80 Hz and 160 - 200 Hz.
In my experience this works well. I haven't noticed any detriment to the stereo separation with this approach. I believe the reason for this is the fact that above 200 Hz there are another 6,5 octaves of information containing localization cues.
(sorry about the psychoacoustics blurb to those familiar with this)
Human hearing works based on two mechanisms, interaural level difference (ILD) and interaural time difference (ITD). ILD is based on level differences caused by sound absorption and shadowing of high frequencies by the human head. ITD is based on the phase difference between the two ears, and is effective below 1500 Hz.
https://en.wikipedia.org/wiki/Interaural_time_difference
Here is a research report I found about ITD/ILD trading
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3003727/
one of the conclusions from this report:
"Rivalry experiments with opposing interaural level differences (ILDs) found that the trading ratio between ITD and ILD increasingly favored the ILD as coherence decreased."
this would confirm my observation, that in practice stereo imaging is not impaired by having the below 200 Hz range played back in mono, because the human brain, presented with incoherent (=conflicting) ITD and ILD, will favor the ILD cues for localization.
Finally, as a comic relief I present to you this funny video about another psychoacoustic effect that will seriously tease your brain:
https://auditoryneuroscience.com/McGurkEffect
Last edited:
On axis is different because the sound sources arrive at the same time, no time difference = no comb filtering and no lobing. But at higher frequencies on axis is a very narrow spot.
The only reflections you have outside in this case are floor bounce, and as I already mentioned that is a somewhat unavoidable problem. But that's the only reflection at play here that can cause any timing issues on axis.
Wrong. As has been showed in very bright colors, when sound sources are within 1/4 wavelength of each other there is no destructive interference and no lobing. This is fact.
With dual sound sources far apart there will always be destructive interference and lobing. The farther apart they are the lower in frequency and the more evenly spaced the comb filtering becomes so theoretically the further apart the sound sources are the worse the effect will be. But a sub and mains can be pretty far apart without any problem if you crossover low enough and steep enough, then there are no lobing problems between the sub and the mains.
Context indeed. This Geddes quote appears to be primarily concerned with mains imaging in a small room, not subwoofer lobing (frequency response) outside with no reflective boundaries other than the ground.
Note this sentence in particular -
He specifically says don't worry about anything below 200 hz because multiple subs and eq in a small room will even out the response.
Besides, imaging has little (if anything) to do with subwoofers anyway. Imaging happens at much higher frequencies. This is why "fullrange" driver systems image so well, there's usually no bass and the range from 1 khz - about 8 khz usually has some nasty peaks. This is a recipe for terribly colored sound but fantastic imaging.
As the quote is talking about mains imaging in small rooms I have no idea how it applies to this discussion in any way.
Last edited:
It can.
I saw you use this term in your other thread. If my memory doesn't fail me you considered anything less to not be a real horn. Did you ever get to the bottom of what the differences are or how significantly they are perceived?
Read back to what was going on a couple of weeks ago. I was consistently complimenting weltersys on his experience (while disagreeing with some of his opinions) and he was consistently making personal attacks. I'm not going to apologize now for questioning what he's really saying.
Look at the first paragraph I've quoted here. You wrote it. Nobody can even figure out what he's saying, other than the fact that he's amused at everyone's incompetence. He keeps saying everyone is wrong but he won't say what is wrong, why it is wrong or how to fix it. He has repeatedly offered his services for hire.
This is diyaudio - projects for fanatics by fanatics. Not diyaudio - the place to shop for system design by designers that may or may not know what they are doing and won't talk publicly about their design decisions.
So what? You can't get the mains any closer to the sub(s) than the OP is already suggesting unless you put the mains INSIDE the sub. IIRC I was the first to suggest that, weltersys may have mentioned it once or twice early on, but since then no one has mentioned it since, and weltersys recently said the mains should be in the barn doors.
What is your solution to this issue? Quit complaining already and offer a solution.
I have no idea what other thread you are talking about.
Full size horns are quite different than massively undersized horns in a variety of ways. I have on many occasions pointed out exactly what the differences are.
If you have a specific question then ask it. The only direct answer I can give you is that I'm pretty sure you could easily perceive the difference between an undersized sub like the Keystone and a full size sub like OP is contemplating. If you want a more specific answer ask a more specific question.
When two sources are within 1/4 wavelength, lobing occurs at greater than 180 degrees, so out of the picture, but destructive interference in this case is proportional to the phase difference and not on spacing.
Just supporting what I claimed in #630..
It is relevant because OP is worried that too high a crossover point between mains and sub might negatively impact stereo imaging.
No. When two sound sources are within 1/4 wavelength they are sufficiently in phase so there is no lobing anywhere, the two sources act as one.
I still don't get it. The quote you pulled is all about imaging.
And he VERY SPECIFICALLY says don't worry about anything under 200 hz because in a small room multiple subs and eq will even it out. None of this is evenly remotely on topic in this discussion. And I'm sure if you asked Geddes about multiple subs outside he would say something very different than what that quote says.
Still doesn't make sense. The Geddes quote he pulled specifically says -
"We localize only on sounds > 700 Hz. so very early reflections < 700 Hz do not cause image problems. "
So what does that have to to with the discussion so far in which we've discussed lobing at subwoofer frequencies and crossover points between 80 - 300 hz?
The point is that the context of the quote is completely wrong for this conversation. He was trying to argue that the sentence "We localize only on sounds > 700 Hz" means that lobing at subwoofer frequencies doesn't matter. But that's not what the quote says at all. And localization is VERY different than standing in a huge bass null where there are whole notes missing.
Furthermore, he went on to say that the tone and pitch can walk across the soundstage with a steep 80 hz crossover, which directly contradicts the Geddes quote. So I still have no idea even what point he's trying to make.
Last edited:
The context was about crossing a tweeter at 700Hz (which I also do), although the crossing of a sub is quite dependent on a room (which this thread may not be about).
However...
At one stage I was correcting discrete room modes up to 800Hz using an auxilliary woofer between the mains. While I don't see it as a permanent solution there was little wrong with it, it worked well.
What I offer here is why I'm here. This is a DIY site. If you want more then ask nicely, and I'll consider it.
What can I say, we're on the same page 🙄
This is how you entered this discussion.
Originally Posted by just a guy View Post
So please educate us poor fools on what we are doing so wrong.
Wrong is a relative term. We are insensitive to small timing differences in the low octaves, reflections proper, diffraction and direction, compared to more critical higher bands, except where the response is affected.
You very specifically said I was "relatively" wrong. Then you backed that up with a series of Geddes quotes that have nothing at all to do with anything that was being discussed. Then you made statements that directly contradicted the Geddes quotes.
I still am not sure what you think I was wrong about.
If you want to tell me I'm wrong, tell me exactly what it was that I said that was wrong, tell me why it's wrong and offer a solution.
If you want to continue to post cryptic hints that I'm wrong and quotes that have nothing to do with the discussion I can continue to point out how I'm not wrong and that your posts aren't even on topic, but this isn't a very efficient way to communicate.
This seems like an outright admission that your initial post and everything following it are not even remotely on topic.
Apparently not, you said "When two sources are within 1/4 wavelength, lobing occurs at greater than 180 degrees" I said there no lobing anywhere.
Last edited:
I thought I said you were relatively correct. Same logic applies to interference. Destructive interference is not necessarily 'destructive'.
Maybe I just walked into the wrong room, but I guess I'd like to see the issue of interference put into the right light. There is no benefit in saying it is always bad [and I gave you this one when I quoted about crossovers of non-coincident drivers not being perfect].
My attempt at enlightenment has failed. Let's move on.
All I could hear was ba ba, which is what she was actually saying. In this world of youtube videos where the sound and video is very often not perfectly synced I'm used to the mouth not moving when it's supposed to, maybe this is why I couldn't hear any sound but ba ba no matter how hard I tried. Makes no difference with eyes open or closed I hear the same thing.
As for the rest, I don't think a crossover anywhere near 300 hz will work for the variety of reasons i've already given.
@ just a guy
Hi, are you sure about F Low = (Qts x fs) / 2 ? Because F Low = fs / (Qts x 2) seems more like it !
Hi, are you sure about F Low = (Qts x fs) / 2 ? Because F Low = fs / (Qts x 2) seems more like it !
Jag,
You are correct, the experience of as many as 40,000 people at a single event using only 21,225 liters of bass horns using 80 drivers is not the same as an "experience" of a single 32,000 liter bass horn simulation using 5 or 10 drivers directed at a hot tub in Seattle 🙂.
Being in the business of providing portable sound systems, we could not transport a single 32,000 liter bass horn made of concrete.
We provided sound systems in "blocks", each block consisted 8 x15", 8 x12", 4x2" exit & 4x1" exit drivers powered with 4/kW/8kW peak power.
We hauled truck integer sized bass horns, the 4x15" L4 was 45" x 45" x 32", a 37.5 cubic foot (1061.25 liter) cabinet. Each "block" had two L4 (or four L2), a minimum of 2122.5 liters of bass horns.
Typical shows used two to four "blocks", but we would use as many as ten or more blocks on large events, 21,225 liters of bass horns using eighty 15" bass drivers. Bass output increased quite a bit when we reduced the FLH L4 size by 33% and converted the straight horn to a manifold conical horn exit with bass reflex ports.
That was in the mid 1980s, recent systems provided by the "Brothers" make the output level of that system, or a 32000 liter bass horn rather "cute" by comparison.
Cheers,
Art
Last edited:
Yes I am. When have you used a stack with a mouth size equivalent to OP's 17 foot diameter mouth size? Even with your 21225 liters of bass horn you very likely never had it set up with a clustered mouth size anywhere near OP's horn. And your sound system blocks sound like a nightmare for higher frequencies, line array would be better, or even better yet point source.
OP's full size bass horn is larger (and has a larger mouth size) and tuned at least an octave lower than anything you have likely worked with professionally.
And I notice you didn't bother to answer any questions on clarifying your recommendations for OP's system at all. Instead we get a story of a completely different type of system (wholly mediocre by today's standards in every way). I'm sure your glory days were interesting but you've never worked with a system even remotely like OP's and it would be far more enlightening if you spoke about recommendations instead of rehashing the past.
Pretty sure it's correct. The only reference I could find in a quick search is here - BD-Design - Bass Horn Design
I skimmed a bunch of Keele horn design papers but couldn't find the Fl formula - the Fh formula is always there though. The likely reason for this is because Fl is rarely a problem, while Fh is a major issue with all horns that are not subwoofers.
Fl and Fh are the mass corner roll off points of the driver. You probably won't ever have a problem with Fl so don't worry about it. Fh is more important, both for design considerations and to understand the ancient 1970s horn driver spec recommendations.
When's the last time you did? Or when's the first.
If I was going to Mars and had the choice of using an engineer that had already been to the moon, or a guy who had been to Mars a thousand times, on a Pc, guess who gets my money?
You may have much to offer guy but the I'm right I'm right I'm right blog is insufferable. I would like to see how this project turns out but this is a joke.
The last word is yours, I'm out.
Barry.