On the the ceiling itself, between the loudspeaker and the listener area ? Or does the ceilling arm behind the listener too ?
What should be then the overture of such a horn ? 60°, 90°, more ? Or is this last related at home with the distance between the loudspeaker and the listener, midfield in living rooms for most ?
What is the average distance you are all talking about between the loudspeaker and the listener and also ceilling heigth, please ?Mine as illustration is 11.5' (3.5 m) and ceilling heigth is 8.9' (2.75 m). Is that matters ? I am not sur to understand if the distance for early reflexions are important whatever the frequency.
Last edited:
Only between the speakers and listeners.
Narrower directivity, high DI, is always better, especially in smaller rooms. But high constant DI gets ever harder requiring ever larger solutions. It becomes untenable at some point.
There timing and frequency influences of reflections is a highly complex subject.
My room is about 21 x 14 x 8 feet. The speakers and listener are in an equilateral triangle of about 8 feet.
Narrower directivity, high DI, is always better, especially in smaller rooms. But high constant DI gets ever harder requiring ever larger solutions. It becomes untenable at some point.
There timing and frequency influences of reflections is a highly complex subject.
My room is about 21 x 14 x 8 feet. The speakers and listener are in an equilateral triangle of about 8 feet.
@GM and @weltersys
Hi, I figured I draw your conversation here, since I am talking to both of you in different threads about the same stuff and you are both teaching yet I am confused.
The topic of Vent CSA I have overlooked some details in the past. When it comes to vent CSA the larger it is, particle velocity does down, but as long as we stay within a threshold, one should not expect consequences is what I think is true. Whether that number is 15, 20, 35, or 55 M/Sec depends on who you talk to. Either way, eventually Turbulence sets in, in an audible way, and as we increase from there, eventually, the air in the starts to compress, which eats energy.
As long as we stay within reasonable particle velocity, the other aspect is the passband of the vent. Larger vents contribute more SPL to combined total, and a wider spectrum is contributed, as well.
This seems to refer to the vent around its fundamental resonance. Vent CSA performance, away from the fundamental, in my recent sims, have shown me that, my ideas, of Radiation Mass area and efficiency, at the terminus, work at the fundamental resonance but away from there, the relationship is not well defined. IN slot pf my PPSL, according to the sim, as I shrink down the CSA, I lose output towards the fundamental while starting to gain some extension, but a whole bunch but more importantly is that LF efficiency does not fall off in the way I thought it would.
So in this front horn, I am wondering did I do myself a disservice by not placing the woofers even closer together, and is LF quality less to do with VV as long as under a certain particle velocity
Hi, I figured I draw your conversation here, since I am talking to both of you in different threads about the same stuff and you are both teaching yet I am confused.
The topic of Vent CSA I have overlooked some details in the past. When it comes to vent CSA the larger it is, particle velocity does down, but as long as we stay within a threshold, one should not expect consequences is what I think is true. Whether that number is 15, 20, 35, or 55 M/Sec depends on who you talk to. Either way, eventually Turbulence sets in, in an audible way, and as we increase from there, eventually, the air in the starts to compress, which eats energy.
As long as we stay within reasonable particle velocity, the other aspect is the passband of the vent. Larger vents contribute more SPL to combined total, and a wider spectrum is contributed, as well.
This seems to refer to the vent around its fundamental resonance. Vent CSA performance, away from the fundamental, in my recent sims, have shown me that, my ideas, of Radiation Mass area and efficiency, at the terminus, work at the fundamental resonance but away from there, the relationship is not well defined. IN slot pf my PPSL, according to the sim, as I shrink down the CSA, I lose output towards the fundamental while starting to gain some extension, but a whole bunch but more importantly is that LF efficiency does not fall off in the way I thought it would.
So in this front horn, I am wondering did I do myself a disservice by not placing the woofers even closer together, and is LF quality less to do with VV as long as under a certain particle velocity
With your sealed dual 18" box you may be imposing yourself to a sonic disservice crossing over so close to the peak/dip frequency caused by the volume of the "front horn" plenum slot. Placing the woofers closer together or further apart simply changes the frequencies of the peak/dip.
To me, all your PPSL has done for you is maybe a few dB boost over a very narrow band of freqs, but in return you net a null at about 300 Hz which you can do nothing about. Art's point was that a CD cannot go low enough to accomodate this null no matter how close together you get them.
I just use them direct radiating up to the CD CD (constant directivity compression driver,) Like I recommended before.
If you look at the graphs that you show over three decades, one can exclude the lowest one from the point of view of speaker design since <100 Hz is room dominated anyway and there is not much that the speaker design can do to influence that. Thus you have basically two decades to cover with a design. Ideally split this and you have a very logical two way design for >100 hz. < 100 Hz should be multiple subs anyways. With a huge hole at 300-400 Hz you cannot do this with a two-way.
I just use them direct radiating up to the CD CD (constant directivity compression driver,) Like I recommended before.
If you look at the graphs that you show over three decades, one can exclude the lowest one from the point of view of speaker design since <100 Hz is room dominated anyway and there is not much that the speaker design can do to influence that. Thus you have basically two decades to cover with a design. Ideally split this and you have a very logical two way design for >100 hz. < 100 Hz should be multiple subs anyways. With a huge hole at 300-400 Hz you cannot do this with a two-way.
So you are saying the vibration cancellation is more of a gimmick in my situation, considering the weight?
(1m outside pointed at the sky from the ground, few filters to level)
What are you worried about? THD/IMD? This is the loudest measurement I have. How loud do you speculate I need? If 95db rms, then 110db peak, minus -6db at the xo point and another -3db with the addition of another channel. 101db 200hz/1m. This is all about parasitic THD correct?
If it is about FR then, doesn't the above show that they can cross?
Yes, any "X" shows a cross
The PPSL 18" "drops off the cliff" above 160Hz, the HF driver's response is "off the cliff" below 300Hz.
Most would want drivers to have a reasonably flat overlap area near the crossover point, your choice has near an octave of compromised (shite) driver response around an acoustic crossover placed between those frequencies.
Considering you had/have high quality 15"s that could cover that range and more than an octave above and below, just seems a bad choice.
But that's "DIY" for you, you only need to make yourself happy
Art
I am sure that there is some vibration cancellation, but the "moving mass" is not the total mass, and anyway I'd be more worried about a huge null at 300 Hz.
Yea, I'd hate to try and design a decent Xover in that situation - doable, at best, but likely a disaster.
I discussed it here -> https://www.diyaudio.com/community/...w-distortion-with-a-2-way.334757/post-7253619
I tried things that were more in line with what is technically acceptable, for example, I tried crossing over at higher points using the 15". I came to the conclusion that at 1m it is impossible to crossover between two drivers where the CTC isn't low enough without consequence of imaging. As I lowered the XO it just sounded more and more natural. At 1m, if I separate vocals on the baffle, I can localize the different sources and it sounds unnatural. HF Axis takes priority in imaging and thus vocals should live there or if crossed to another driver, it should not be perceivable. Any deviation from that should be minimal. I tried to conceptualize why I never had issue with my little 2ways. CTC in a 1"+4" two way is low ... At 1meter, illusions fall apart much more readily. If crossing at 600hz.... I could tell that the High and Lower section was located on two different Axis. I can't easily perceive the split on the 1"+4" 2 ways I'm listening to right now, at just under 1meter. I think at 1meter the voice probably can't be stretched too much on the baffle without risk of loosing good imaging. The only way to improve things if your CTC isn't low, is to lower the XO hopefully until you reach a point to where the localization is no longer an issue, or, possibly MTM though I have not really experimented with MTM. My last system I listened to at 1m as well, 1"+4"+12".... crossed at 130hz and 2.1khz. The CTC was 12" for the midrange and 12" woofer, I had to lower the XO to 130hz before I felt localization was no longer an issue. So Now my Top to Bottom spacing has gone from 15.5" to 18.5" with the Cynosure, but when I aimed to recreate said previous experience, the same loss of localization of the XO point, judged by the singularity of vocals, 200hz was enough... even though CTC and top to bottom spacing is higher for the Cynosure versus the previous 3way. The Vocal image is very good and it takes something like Barry White to break the illusion. At 1m, thats not impossible, good vocal imaging that is... but the headroom I have at 1m, in concert with the imaging, its life like. At these Distances, I'd only expect Something like the Tom Danely Hyperion or Genelec 8381 to compare.
I did this messing around in REW
I wouldn't mind a slightly larger horn, a complete roll over, could do the trick, I'd love to see measurements of the horn above on the Axi. 142hz isn't saying much.@maravedis
After realising that 90% of my listening if of pre-1930's recordings I made up this - which gives lots of pleasure!
142hz hybrid-horn with Axi2050 and a sealed bass 18FH500.
The horn driven by a 300b amp with Nickel OPT and is loaded from 200hz and is pretty flat to about 15khz. I use a WARM tube EQ and adjust freely for each recording. The bass is driven by a Behringer NX3000D which adds the delay to meet the treble horn.
This also functions as a centre speaker in a 3 way stereo system.
Last edited:
I can't get my laptop to start up in windows where my other measurements are but I came across this, This is at 1m, in untreated room, IIRC these are 1db increments, Whats the issue at 200hz?
I believe that a better voicing and crossover is achievable, as this was sort of thrown together, using the amps dsp. FIR software would do better, right?
I believe that a better voicing and crossover is achievable, as this was sort of thrown together, using the amps dsp. FIR software would do better, right?
I just thought of something.
I also did tests by running a sine, up and down the crossband. FR is not going to be perfect because of the room but I listened for some type of abnormality that would suggest some type of failure in the crossing and I don't remember finding it. I think I found the opposite.
I also did tests by running a sine, up and down the crossband. FR is not going to be perfect because of the room but I listened for some type of abnormality that would suggest some type of failure in the crossing and I don't remember finding it. I think I found the opposite.
do you mean the phase trace? Likely nothing.
The phase seems to be plotted so that there is +/-180deg visible, so the phase wraps around the plot when its more / less than that. The phase trace has undulation to it and it just the plot that looks like it's all over the place as the undulation makes it wrap multiple times back and forth. If you smoothed the graph there would be just one wrap around the 200Hz mark. Or if you plotted the phase like +/-900deg I bet it looks the same around 200Hz as above, smoothly changing curvature.
Last edited:
The XO point is at 200hz.... I am trying to analyze the "flaws"... For a 1m measurement in an untreated small living room, rough draft Xo/voicing at the most...... probably 100ms window... hows it look?
Last edited:
Hi, not sure how usable information that is, but looks fine for what it is.
1m is likely too short measurement distance for such big speaker. VituixCAD measurement manual suggests 1m or 3x of device width measurement distance in order to reduce near field effects that are not present in actual listening distance at far field.
In general, since the system is big and crossover is so low, it is hard to get good data around the crossover indoors due to reflections messing up.
You should use some kind of tricks to reduce reflections, arrayed mics or something, see "klippel scanner on shoe string" thread. Or do measurements outside, group plane or high up in the air.
Big speakers are prone to measurement errors, as some issues could be windowed out from measurements, or the measurements contain anomalies from the environment that aren't really there.
1m is likely too short measurement distance for such big speaker. VituixCAD measurement manual suggests 1m or 3x of device width measurement distance in order to reduce near field effects that are not present in actual listening distance at far field.
In general, since the system is big and crossover is so low, it is hard to get good data around the crossover indoors due to reflections messing up.
You should use some kind of tricks to reduce reflections, arrayed mics or something, see "klippel scanner on shoe string" thread. Or do measurements outside, group plane or high up in the air.
Big speakers are prone to measurement errors, as some issues could be windowed out from measurements, or the measurements contain anomalies from the environment that aren't really there.