Jump to content

Bass horns operating below K-402


jazzmessengers

Recommended Posts

3 hours ago, Chris A said:

To my knowledge, the Haas research and perhaps later psychoacoustics research hasn't shown the effects of 1-3 millisecond delays in perception by the human hearing system (relative to the 20 to 30 ms delays characteristic for the full precedence effect) but it's quite clear to me that there is a regime change in terms of phantom center channel and soundstage perception in this range of 1-3 ms. 

 

My own informal testing has shown that delay becomes noticeable at around 1/2 millisecond. Perhaps not coincidentally, that corresponds almost exactly to the time it takes sound to traverse the distance between my ears. I cannot say with certainty that the perception of pure delays between channels correlates with reflections that are delayed by similar amounts, but I am pretty confident that such early reflections cannot be summarily dismissed as inconsequential.

Link to comment
Share on other sites

Griesinger has mentioned that the human hearing system is basically a 1 ms system (pg. 35):

 

Quote

“The human brain is a computer of great sophistication and complexity, with a clock frequency of 1kHz.”  (Hiroshi Riquimaroux)

Other research shows a regime change at 0.7-1 ms in terms of localization (i.e., the precedence effect):

 

gallery_26262_6_4037.jpg

What I mentioned above was that there is apparently something else happening at 1-3 ms that's not typically discussed (in terms of localization/imaging)...and that the complete absence of reflections actually decreases clarity/intelligibility, but too many "early reflections" dramatically decreases clarity.

 

It's not exactly what I'd call a black-or-white situation...there are some fine grained nuances involved.

 

Chris

  • Like 1
Link to comment
Share on other sites

Note that the period of a sound wave at 333 Hz (@room temp.) is 3 ms, and that of a 1 kHz wave is 1 ms. 

 

So I'm beginning to believe that "early reflections" also have a wavelength-distance component in terms of perceptions that is separate from the purely psychoacoustics phenomena discussed above.  There is apparently a change in imaging perception in small rooms at 1 to 3 ms--which s right in the middle of the midrange band and just below (in frequency) the 1 ms processing step size of the human hearing system. 

 

It's the midrange frequencies that are the problem when placing a loudspeaker near a room boundary.  And it's quite easy to control these reflections using standard 1" thick absorption pads in the nearby walls--which have no effect at bass frequencies--without having to resort to moving the loudspeakers away from the room boundaries.  Hence the advice to place absorption pads on adjacent walls next to the mouth of the K-402s and other midrange frequencies--such as the bass bin itself playing midrange frequencies above ~300 Hz.  This effectively does the same thing as placing the loudspeakers out away from the room boundaries--without the corresponding loss of mid-bass and bass frequencies that PWK talked about in his "corner speaker placement" article from July 1959.

 

Chris

  • Like 1
Link to comment
Share on other sites

3 minutes ago, Chris A said:

Note that the period of a sound wave at 333 Hz (@room temp.) is 3 ms, and that of a 1 kHz wave is 1 ms. 

 

So I'm beginning to believe that "early reflections" also have a wavelength-distance component in terms of perceptions that is separate from the purely psychoacoustics phenomena discussed above.

 

The explanation might be as simple as "comb filtering".

Link to comment
Share on other sites

Well, yes. 

 

Let me know if I'm going too fast here...

 

Greg, you mentioned the space between the eardrums being 0.5 ms (basically a physiological average measurement)--corresponding to a frequency of 500 Hz for one wavelength (or 500 Hz notch filtering, as the case may be).  The interaural direction-finding capability of the human head/ear lobes/ear drums, etc. probably varies between 700-800 Hz for small children to perhaps 350-400 Hz for larger adults with longer interaural spans between the ears.

 

Once you get below 500 Hz (say, 333 Hz - which corresponds to 3 ms of delay for one wavelength), the direction finding capability of humans falls off quite rapidly (as seen in Toole's book), i.e., the imaging/locating capability of the human hearing system is mostly formed above 300 Hz. 

 

All this is to say that the corresponding 1-3 ms "zero reflection zone" advice that you found above for best imaging, it's likely the midrange frequencies between about 300-1000 Hz are what really set the major part of our acoustic imaging/soundstage framework.  Early reflections in this part of the spectrum apparently confuses our built-in DF system, so control of those early reflections is most important for "hi-fi" reproduction. 

 

So let's not throw the baby out with the bathwater and move the loudspeakers away from all room boundaries.  That's not the best tradeoff that you can make, IMHO...

 

Chris

  • Like 1
Link to comment
Share on other sites

On 1/1/2018 at 8:30 PM, jazzmessengers said:

 

For front loaded horns these tend to become enormous (even say ones that load down to 50 Hz) unless you were thinking something like an array of multiple woofers horn loaded so the increased radiating area at the throat allows for a shorter length (ie pictures below if I didn't describe it correctly, I know these are more upper midbass type horns, just using them for illustration)?

 

I have no "as much horn loading as possible" dogma :) I'm an AES paper/Toole fan just trying to put something together for music use only that is objectively high fidelity.

 

OnXiow6.jpg

 

jVg0prK.jpg

My Quarger Pie horn sticks out only a foot more than a LaScala cabinet in the corner and it gets to 50 Hz. FLAT.

MiniMWMsCorner3.jpg

  • Like 1
Link to comment
Share on other sites

In the computer code "ICHIN" ("I Can Hear It Now") it's most often seen that human helicopter detection frequencies of around 500 Hz are found--based on things like acoustic propagation physics models, wind, and digitized terrain--which actually corresponds to experimental evidence of what actually occurs in the field with real people. Perhaps there is more to aural detection than just meeting threshold signal/noise detection levels after all...:emotion-21:

 

Chris

  • Like 1
Link to comment
Share on other sites

29 minutes ago, Chris A said:

Greg, you mentioned the space between the eardrums being 0.5 ms (basically a physiological average measurement)--corresponding to a frequency of 500 Hz for one wavelength (or 500 Hz notch filtering, as the case may be).  The interaural direction-finding capability of the human head/ear lobes/ear drums, etc. probably varies between 700-800 Hz for small children to perhaps 350-400 Hz for larger adults with longer interaural spans between the ears.

 

Careful, Chris. A half-millisecond corresponds to one wavelength at 2000 Hz, and a comb filter at multiples of 1000 Hz.

 

Greg

Link to comment
Share on other sites

3 hours ago, Chris A said:

I think that you may need to reexamine your comment:

In Toole's 3rd Ed., he references research that directly negates what you've said here, depending on your definition of "early".  So does Griesinger's latest presentations on clarity.

 

Chris

No. This is very well established. And also extremely easy to hear for oneself by experimenting.

Your link talks about clarity in a different matter (large room acoustics) and has zero relation to this.

 

It saddens me that I actually have to defend this. I think this one of the reasons why very knowledgeable people don't bother to discuss as forums any more. Getting through the forest of confusion is difficult.

Link to comment
Share on other sites

True that, but the quarter wavelength cancellations occur at multiples of 500, 1500 Hz etc., which is what I believe is what the human hearing system is using for soundstage perception (among other characteristics of the acoustic field). 

  • Like 1
Link to comment
Share on other sites

Hey jazz...

 

Have you considered an MTM alignment with the bass horn instead? Also, how much depth are you willing to go with?

 

I've been mulling over an 80Hz to 800Hz bass horn for a few years now and there's always something you have to give up if you want it to be a straight horn. The mouth area requirements for 80Hz make it real hard to get good vertical polars at the xover region - especially if crossing over to the K402. You would think you could undersize the mouth if the speaker is placed in a corner, but I haven't found a way to get the wavefront to align well with the room corner without folding the horn. If you don't align the wavefront, then you end up with early reflections from the corner that create large dips in frequency response at the lower frequencies (due to comb-filtering).

 

Going to an MTM alignment makes the mouth appear larger, and I've been experimenting with ways to use the vertical space to the sides of the K402 to augment loading into the room corner. I haven't found a great solution yet, but I'm still working on it. The interesting thing is that all methods to improve polar response cause the design to start looking closer and closer to Danley's synergy horns. I know Chris is really excited about the frequency response he gets out of his K402 with woofers bolted to it, but I'm not sure the K402 is providing much loading at the lower frequencies...it's been a while since I crunched numbers, but I want to say it was less than a  5% reduction in excursion for the same SPL (which isn't a large decrease in FMD). In other words, it's not much different than having normal direct radiators - the only real benefit is polar alignment. All that to say, I think it would make more sense to make a larger version of the K402 to provide the loading to an even lower frequency, and then port the woofers to reduce cone excursion as much as possible. Unfortunately this makes the xover to the subwoofers for 80Hz and below a lot more complicated due to all the phase happening. A larger K402 ain't exactly a walk in the park either...

 

Also, room modes start to dominate as high as 200Hz in some rooms - in which case I argue the location of the acoustic source is more important than polar consistency to the higher frequencies. (Thinking in light of the dual zoned bass array concepts). This has me returning to the MTM approach - but custom tailored to the dimensions of the room. Basically instead of avoiding early reflections at lower frequency, we can use the room as part of the wave guide....

 

 

Granted, these are all reasons why absolute perfection isn't possible - which we probably already knew. I think a split bass-horn in an MTM alignment can give very good results, and would be the best method to have the most control over the height of the tweeter. The only tradeoff is a notch in the far off-axis of the vertical polars, which shouldn't be very noticeable....certainly a lot less of an issue than the polar lobing that happens with the Jubilee LF. This approach will require a cabinet that is roughly 4ft deep if you want good loading all the way down to 80Hz...

 

 

 

 

Link to comment
Share on other sites

59 minutes ago, Bjorn said:

Your link talks about clarity in a different matter (large room acoustics) and has zero relation to this. 

You tend to dismiss very quickly those things that you've not spent much time considering...and I think you again may be overlooking the fact that Griesinger's presentation is applicable to the present case. 

 

You also didn't consider Toole's references that I referred to.  Just because you might disagree on one subject with an author probably doesn't mean that you should disagree with all that he writes.  Also note that his books are quite useful in pulling together wide swaths of acoustics and psychoacoustics information that would be otherwise laborious to digest... See

 

Chris

  • Like 1
Link to comment
Share on other sites

1 hour ago, DrWho said:

but I'm not sure the K402 is providing much loading at the lower frequencies...it's been a while since I crunched numbers, but I want to say it was less than a  5% reduction in excursion for the same SPL (which isn't a large decrease in FMD). In other words, it's not much different than having normal direct radiators - the only real benefit is polar alignment. All that to say, I think it would make more sense to make a larger version of the K402 to provide the loading to an even lower frequency, and then port the woofers to reduce cone excursion as much as possible. Unfortunately this makes the xover to the subwoofers for 80Hz and below a lot more complicated due to all the phase happening. A larger K402 ain't exactly a walk in the park either...

Note that I did an analysis on a 1.5x- and 2x-sized K-402 MEH before I settled on the K-402-sized version.  There is a reason why I think that the present K-402-sized MEH (basically the same size as the Danley SH-96) is in the sweet spot in terms of size. 

 

When you do the measurements on horns with mouths as large as the K-402, things begin to occur in terms of room boundary gain that you might not have counted on.  Trying to do this with a full-length midbass horn will effectively take you out of the corner boundary gain regime due to the size of the horn itself not being able to be placed in proximity to a room corner. 

 

This is a problem that caused me to abandon the 1.5x or 2x horn--at least for the moment, and the measurements and listening experience has since validated that the K-402-sized horn is large enough-especially if used as a corner horn.  And getting such a large 1.5x or 2x horn through a doorway also becomes an issue without resorting to multiple-part horn construction, something that I've not been a big fan of...

 

Chris

  • Like 1
Link to comment
Share on other sites

By the way, and in case you're wondering, the K-402-MEH picked up 7 dB of gain at 40 Hz going from (elevated) mid-wall to the floor.  Going from elevated mid-wall to corner--over 12 dB at 40 Hz...

 

Even though the unit is elevated mid-wall, it has more than enough Sd (woofer area) to support flat EQing down to 40 Hz without incurring significant distortion (harmonic or modulation) and certainly not at home theater listening levels. This is probably due to the half space loading and its large mouth.  This was part of the surprise that I was alluding to in the K-402-MEH thread, but no one seemed very interested in that aspect at that time.  There is a trade off on wall position and bass extension, but it's not nearly as confining as you might think...at least as the result of my listening trials.  YMMV.

 

(BTW...it's not FM distortion which is the issue at these low frequencies--it's AM distortion.  See Kippel.)

 

Chris

  • Like 1
Link to comment
Share on other sites

7 hours ago, Bjorn said:

What would you choose a horn speaker like K-402 if you desire lateral reflections? A horn like K-402 minimizes side wall reflections with it's high DI. If you want lateral reflections, a horn is isn't the right choice and you might want to condsider a CBT speaker. 

 

A traditional CBT speaker has a wide horizontal dispersion, avoids the floor bounce and minimizes ceiling reflections. FIY: I'm working on a CBT design with Don Keele.

5a4ccbddeb1b8__MG_2364(Large).thumb.jpg.2ed22f8ab54341223d6561777dcc08d7.jpg

 

 

 

 

Primary reason is I am really not a fan of the way they look :)

 

Second is I've heard the K-402/Jubilee in a rather live sounding room and absolutely loved the way they sounded, so this is a safer choice and I can mess around with diffusion or absorption at the side walls when I get to that point. For now the midbass horn/bass solution is most important.

  • Like 1
Link to comment
Share on other sites

2 hours ago, DrWho said:

Also, room modes start to dominate as high as 200Hz in some rooms - in which case I argue the location of the acoustic source is more important than polar consistency to the higher frequencies.

I agree that the ability to hear polars becomes quite difficult at 200 Hz and below from my listening experiences in my home-sized listening room.  It's much more important to have good consistent controlled polars from 500 Hz down to 200 Hz,  and it's startling how much this affects speech intelligibility/clarity.

 

Chris

Link to comment
Share on other sites

2 hours ago, DrWho said:

Hey jazz...

 

Have you considered an MTM alignment with the bass horn instead? Also, how much depth are you willing to go with?

 

I've been mulling over an 80Hz to 800Hz bass horn for a few years now and there's always something you have to give up if you want it to be a straight horn. The mouth area requirements for 80Hz make it real hard to get good vertical polars at the xover region - especially if crossing over to the K402. You would think you could undersize the mouth if the speaker is placed in a corner, but I haven't found a way to get the wavefront to align well with the room corner without folding the horn. If you don't align the wavefront, then you end up with early reflections from the corner that create large dips in frequency response at the lower frequencies (due to comb-filtering).

 

Going to an MTM alignment makes the mouth appear larger, and I've been experimenting with ways to use the vertical space to the sides of the K402 to augment loading into the room corner. I haven't found a great solution yet, but I'm still working on it. The interesting thing is that all methods to improve polar response cause the design to start looking closer and closer to Danley's synergy horns. I know Chris is really excited about the frequency response he gets out of his K402 with woofers bolted to it, but I'm not sure the K402 is providing much loading at the lower frequencies...it's been a while since I crunched numbers, but I want to say it was less than a  5% reduction in excursion for the same SPL (which isn't a large decrease in FMD). In other words, it's not much different than having normal direct radiators - the only real benefit is polar alignment. All that to say, I think it would make more sense to make a larger version of the K402 to provide the loading to an even lower frequency, and then port the woofers to reduce cone excursion as much as possible. Unfortunately this makes the xover to the subwoofers for 80Hz and below a lot more complicated due to all the phase happening. A larger K402 ain't exactly a walk in the park either...

 

Also, room modes start to dominate as high as 200Hz in some rooms - in which case I argue the location of the acoustic source is more important than polar consistency to the higher frequencies. (Thinking in light of the dual zoned bass array concepts). This has me returning to the MTM approach - but custom tailored to the dimensions of the room. Basically instead of avoiding early reflections at lower frequency, we can use the room as part of the wave guide....

 

 

Granted, these are all reasons why absolute perfection isn't possible - which we probably already knew. I think a split bass-horn in an MTM alignment can give very good results, and would be the best method to have the most control over the height of the tweeter. The only tradeoff is a notch in the far off-axis of the vertical polars, which shouldn't be very noticeable....certainly a lot less of an issue than the polar lobing that happens with the Jubilee LF. This approach will require a cabinet that is roughly 4ft deep if you want good loading all the way down to 80Hz...

 

 

 

 

 

I haven't considered this, I haven't looked into any K-402 replacement, right now it's basically what I'm "building around". I have a decent amount of space that I can work with, for instance 4 foot depth is quite manageable, I wouldn't really want to go deeper than that.

Link to comment
Share on other sites

22 hours ago, Bjorn said:

 

It saddens me that I actually have to defend this. I think this one of the reasons why very knowledgeable people don't bother to discuss as forums any more.

 

 

@Bjorn, Forum comments come from many different perspectives. Variety of opinions helps keep things intellectually stimulating for the readers. Your projects are interesting. Please do continue sharing.

  • Like 1
Link to comment
Share on other sites

On 1/3/2018 at 12:31 PM, Chris A said:

True that, but the quarter wavelength cancellations occur at multiples of 500, 1500 Hz etc., which is what I believe is what the human hearing system is using for soundstage perception (among other characteristics of the acoustic field). 

If the comb-filtering itself was the mechanism, then how is it that we can tell the difference between left and right? The resultant comb-filter would be the same....

 

Talking about time-arrival differences in the frequency domain is possible if you focus on the phase.....but it's not very helpful. Sure, you can point out the phase mechanism, but it's really simpler to just analyze the time-arrivals in the time-domain. This is because the mechanism on both ends is occurring in the time domain, so these things are more naturally described and understood that way. For example, you could create a matrix transfer function to force the world to be flat when visually represented, but it would just obfuscate the reality of the world actually being round. It just complicates the math unnecessarily. The fourier transform between the time and frequency domains would be a similar way of trying to over-complicate something that is otherwise quite straightforward and elegant. I personally prefer simple maths....I'm not smart enough to visualize crazy complicated calculus descriptions.

 

The most advanced radar systems in the world are just now starting to mimic the incredibly elegant spectral smearing mechanism that our ears use for localization / spacialization. There is all sorts of great research on how our ears work and hear things. I was just at an AES meeting where they were talking about how this frequency dependent time smearing is the reason that frequency correction in hearing aids doesn't work for the "cocktail party setting". We need that localization function to focus our hearing to what we want to hear (like the person talking to us). The point is that comb-filtering is a horrible way to understand this - and you can't compare the comb-filtering of spacial reflections to the multi-dimensional localization mechanism of our ears. The processing on either end is different (1/4 wave reflection versus ear time-arrival) - even if you might see similarities in the frequency domain.

 

Nowhere near as in depth, but here's an entertaining video highlighting how our pinnae are used for localization:

 

(the part where he puts clay on his kid's ears)

 

 

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...