5460(1.3/22.86)=310.49868766404199475066
5460 is static in this formula
1.3 is the KA I am investigating
22.86 is the diameter of the driver divided by 2, in cm
310.49868766404199475066 is the frequency where a 1.3KA would start, as we go higher KA increases, lower, KA decresases. Look on the Geddes paper for the KA charts showing the anticipated dispersion width. Rule of thumb is to crossover with a 2=KA or lower....its a generic rule, with intelligent design the rule is bent. I personally think for a near-field 3 way for myself, the lower I can get ka, is the better. For you, 2 is probably fine, and most will tell you to try and match the polar width (KA) at crossover of each driver in the crossover.
5460 is static in this formula
1.3 is the KA I am investigating
22.86 is the diameter of the driver divided by 2, in cm
310.49868766404199475066 is the frequency where a 1.3KA would start, as we go higher KA increases, lower, KA decresases. Look on the Geddes paper for the KA charts showing the anticipated dispersion width. Rule of thumb is to crossover with a 2=KA or lower....its a generic rule, with intelligent design the rule is bent. I personally think for a near-field 3 way for myself, the lower I can get ka, is the better. For you, 2 is probably fine, and most will tell you to try and match the polar width (KA) at crossover of each driver in the crossover.
Hi,
I'm still reading papers and experimenting with hornresp because I wanted to use the 12PE32 with the conical midbass horn. But volvotreter is using a 15" driver in that horn. I mixed it up with his tractrix midbass horn. So far I couldn't find another 15" which performs well in the conical horn. And the 12" goes down to ~100-120Hz. I wanted to cross at 80Hz. This would mean the death of this idea and a win for the 15" woofer without horn. Or another compromise...
Geddes paper on directivity is on point and understandable.
I'm still reading papers and experimenting with hornresp because I wanted to use the 12PE32 with the conical midbass horn. But volvotreter is using a 15" driver in that horn. I mixed it up with his tractrix midbass horn. So far I couldn't find another 15" which performs well in the conical horn. And the 12" goes down to ~100-120Hz. I wanted to cross at 80Hz. This would mean the death of this idea and a win for the 15" woofer without horn. Or another compromise...
Geddes paper on directivity is on point and understandable.
Last edited:
krivium mentioned that there is a lot of punch in that octave 80-160hz and it could be better to cross at 80hz.
But the real reason is that this horn is pretty big and it should be worth it to spend the space for it
So x-over at 80Hz would be perfect. Here is a simulation with the 12PE32. It could go a little bit lower when I reduce the mouth size while keeping the length. I just played with the inputs of hornresp without paying attention to any ratios.
But the real reason is that this horn is pretty big and it should be worth it to spend the space for it
So x-over at 80Hz would be perfect. Here is a simulation with the 12PE32. It could go a little bit lower when I reduce the mouth size while keeping the length. I just played with the inputs of hornresp without paying attention to any ratios.Attachments
With that type of native response you would be crossing in the 80Hz region anyway.
What krivium said is correct, but you also have to look at this in terms of what the room will do. If you can space your main speakers far enough away from walls then you can avoid cancellation issues in the 80 to 160Hz range. That is easier said than done in most rooms with the distance to the front wall (behind the speakers) often being hard to get 2m plus away from.
You can easily see some of this in action by using REW's room simulation, place your speakers where you think they will go and what what havoc the room wreaks on the response, by moving them around you will be able to find the best balance of issues in practical placement. If your room is rectangular this will work, and as it is based on distances the accuracy is very high if all the information entered is correct.
Getting your speakers and subs positions and types to co-operate with the room will get a vastly improved result. In Toole's research bass attributes approximately 30% to the preferred result, getting it better makes the apparent sound improve significantly. I find the same with my own experiments, a smoother bass response with less nulls makes a huge difference.
The best sub crossover is completely impossible to decide before you have measured the speakers in practical positions. You may need to go high you may need to go low. This is one of the reasons I suggested a simple start with a direct radiator. The greatest weapon a diy designer has is the ability to match the room and speaker together, most commercial speaker designers make educated guesses hoping to get a fairly good result across the board.
Another interesting article series for you
Bass Integration Guide – Part 1
What krivium said is correct, but you also have to look at this in terms of what the room will do. If you can space your main speakers far enough away from walls then you can avoid cancellation issues in the 80 to 160Hz range. That is easier said than done in most rooms with the distance to the front wall (behind the speakers) often being hard to get 2m plus away from.
You can easily see some of this in action by using REW's room simulation, place your speakers where you think they will go and what what havoc the room wreaks on the response, by moving them around you will be able to find the best balance of issues in practical placement. If your room is rectangular this will work, and as it is based on distances the accuracy is very high if all the information entered is correct.
Getting your speakers and subs positions and types to co-operate with the room will get a vastly improved result. In Toole's research bass attributes approximately 30% to the preferred result, getting it better makes the apparent sound improve significantly. I find the same with my own experiments, a smoother bass response with less nulls makes a huge difference.
The best sub crossover is completely impossible to decide before you have measured the speakers in practical positions. You may need to go high you may need to go low. This is one of the reasons I suggested a simple start with a direct radiator. The greatest weapon a diy designer has is the ability to match the room and speaker together, most commercial speaker designers make educated guesses hoping to get a fairly good result across the board.
Another interesting article series for you
Bass Integration Guide – Part 1
I mentioned this is where the punch is located!
Fluid is right and why you heard from different source you are on a right track. Room and loudspeakers as one system.
If no horn here, the Faital 12" which has a low Le. The Xmax and the active system permitt a nice sealed front or lateral cabinet as we are still 4Pi radiating at 160 Hz. May be find such Faital at those frequencies, take subwoof cab. for the two first octaves then... The big 12" Seas from Linkwitz.
I planned to position the speakers to the corners to get the biggest possible listening distance. The simulations with REW were kind of eye opening and shocking at the same time. You can find the simulations attached. This was the best result I could get. IMO it doesn't make big difference if I use 3 or 4 subs. And moving the mains away from the walls made things different but not really better. Most significant is the listening position. It makes the biggest difference. I don't like the behavior in the upper bass frequencies. What do you think about this?
Is it possible to use a small bandwith eq for the room modes? Would this make things better? I read that's a great opportunity if you have FIR filters.
Anyway there will be plenty measurements needed. In the link you posted there is another way crossing the subs and mains with overlapping.
Is it possible to use a small bandwith eq for the room modes? Would this make things better? I read that's a great opportunity if you have FIR filters.
Anyway there will be plenty measurements needed. In the link you posted there is another way crossing the subs and mains with overlapping.
Attachments
Last edited:
Hi, this isn't this bad.
The 35hz will be difficult to treat but for the 100hz and up this is why basstrap are for.
Ok, you have no difference with the 3 or 4 subs because you locate them in corners.
Let's try with 4 but with 2 non located on corner and on more or less uncorelated location (with no axisymmetry: try more random location).
Then you ad another one not located on floor, something like 50cm from ground, here again uncorelated.
The overlap is a good idea and not different than E.Geddes proposition.
The 35hz will be difficult to treat but for the 100hz and up this is why basstrap are for.
Ok, you have no difference with the 3 or 4 subs because you locate them in corners.
Let's try with 4 but with 2 non located on corner and on more or less uncorelated location (with no axisymmetry: try more random location).
Then you ad another one not located on floor, something like 50cm from ground, here again uncorelated.
The overlap is a good idea and not different than E.Geddes proposition.
Last edited:
Placing the speakers in the corners will certainly reduce the impact of boundary interference and if that is your intended placement designing the speakers to work there would be a good idea. You can smooth out a lot of bass issues with multiple subwoofers, delayed, equalised and gain set independently.
Have a look at this paper from Welti and Devantier there are other sources of information and presentations with simpler illustrations as well.
https://pdfs.semanticscholar.org/00da/51387c572cfd27c0256cb15e44e976a1a72e.pdf
In a rectangular room they found that either 2 or 4 subs worked best and that mid wall or corner positions were best. Have a look at some of their research to see which sort of response you think is easiest to work with. Then have a look at Multi Sub Optimizer Software, which is kind of a free version of sound field management.
Yes minimum phase EQ will allow modes to be suppressed. No need for FIR to deal with this. With well placed multiple subs and a fairly fixed listening position the above software can work wonders.
In a home type of construction anything under 200Hz is going to be hard to deal with easily using absorbent due to the depth required. Helmholtz style targeted traps would be needed and are notoriously difficult to get right.
It is important that any treatment in a fairly normal room like shown is as broadband as possible. I have made some Fibreglass panel absorbers about 180mm thick spaced off the wall a distance and these are pretty good at absorbing evenly down into the transition region, below there EQ is the best thing to use and works really well.
Here's another interesting way of using subwoofers to combat room issues
JBL M2 Audiolense Digital Crossovers w/Subs | Audio Science Review (ASR) Forum
I agree.
From what i've seen this works fine in relatively large rooms ( cinema/theatre). In smaller room Geddes approach seems easier to apply and gave nice results when i tried. But it is worth trying both or other approach, there is sometimes surprise in 'real life' ( the build technique plays a role ime).
I don't agree with that. But it seems i'm the only one in there. Acoustic issues needs acoustic treatments in my view.
Limp mass absorbers are fearly effective and relatively broadband enough.
All tuned basstrap needs to be tuned in place and located where they are the most efficient.
This is not easy but not impossible ( time consuming i agree as you need to make a map of pressure in the room to find where to locate them).
Your panel must work great though. Seems inline of what i've seen used in pro situation.
Quote:
Yes minimum phase EQ will allow modes to be suppressed. No need for FIR to deal with this. With well placed multiple subs and a fairly fixed listening position the above software can work wonders.
I agree with you that acoustic issues are best solved acoustically, but low room modes are tricky to control that way and they really are a system problem at a point in space which can be controlled very well with EQ. An excess of energy that can be remedied by removing energy from the source.
my current implemnetation
Hi Drumberg,
I'm following this thread from the beginning, but was too lazy to comment earlier. Im running a horn system in a small basement room (6,4 m x 3,2 m x 2,13m) and tried a lot to get good sound over the last two decades. My setup is multi-channel dsp-driven with some acoutical treatment. Bass is delivered by four 15 inch drivers crossed over at 250 Hz. The upper horn is crossed over at 1000 Hz. The tweeter horns are added at 10000 Hz.
The punch between 80Hz and 160Hz is more than sufficient.
regards,
Bodo
krivium mentioned that there is a lot of punch in that octave 80-160hz and it could be better to cross at 80hz.
But the real reason is that this horn is pretty big and it should be worth it to spend the space for it
Hi Drumberg,
I'm following this thread from the beginning, but was too lazy to comment earlier. Im running a horn system in a small basement room (6,4 m x 3,2 m x 2,13m) and tried a lot to get good sound over the last two decades. My setup is multi-channel dsp-driven with some acoutical treatment. Bass is delivered by four 15 inch drivers crossed over at 250 Hz. The upper horn is crossed over at 1000 Hz. The tweeter horns are added at 10000 Hz.
The punch between 80Hz and 160Hz is more than sufficient.
regards,
Bodo
Attachments
Hi,
Bodo can you tell us a little bit more about your speakers? What horns and drivers are you using? How far is your listening distance?
@krivium: Of course I tried different random positions even with the mains. But I couldn't get a better result. The result is changing. By moving them around you improve one mode but another one gets worse. When using only 3 subs it was good at the middle of the back wall behind the listening position. Also the delays can change a little bit. But like fluid said hopefully it will be better with delay, eq and gain.
Btw I'm not sure if I mentioned it before. As subs I want to use the Beyma 12BR70. It's not a dedicated sub but it goes very low in a closed box and has a nice sensitivity. Only downside are the ~150L volume needed.
Broadband absorber are on my list of course. But I didn't plan in detail how thick they should be and how many my room can handle. It would be great if everything works without bass traps. But maybe I can add them if needed.
I didn't check Welti and Devantier paper and the software so far. Toole and Geddes on youtube took my attention...
But I'm happy that the simulation isn't that bad and there is hope with further tools. Now I need to decide which horn/driver combination I want to use.
Bodo can you tell us a little bit more about your speakers? What horns and drivers are you using? How far is your listening distance?
@krivium: Of course I tried different random positions even with the mains. But I couldn't get a better result. The result is changing. By moving them around you improve one mode but another one gets worse. When using only 3 subs it was good at the middle of the back wall behind the listening position. Also the delays can change a little bit. But like fluid said hopefully it will be better with delay, eq and gain.
Btw I'm not sure if I mentioned it before. As subs I want to use the Beyma 12BR70. It's not a dedicated sub but it goes very low in a closed box and has a nice sensitivity. Only downside are the ~150L volume needed.
Broadband absorber are on my list of course. But I didn't plan in detail how thick they should be and how many my room can handle. It would be great if everything works without bass traps. But maybe I can add them if needed.
I didn't check Welti and Devantier paper and the software so far. Toole and Geddes on youtube took my attention...
But I'm happy that the simulation isn't that bad and there is hope with further tools. Now I need to decide which horn/driver combination I want to use.
Start with the main speakers only in the simulation and try and find the best position with the least nulls. The lower frequencies will be taken care of with the multiple subs but big nulls in the wrong place will make the crossover difficult.
I simulated the Beyma drivers and they want a ~50 litre box with fairly heavy stuffing to get a 0.7 Q, with 120W input they do not exceed xmax. Seems pretty reasonable and no need for 150L in my sims.
To see the response of any absorber you can get a reasonable idea with the porous absorber calculator. You need to know the gas flow resistivity of the material you will be using to get a more accurate prediction. These vary wildly between materials that look very similar. In general something around 10,000 or less works pretty well in reasonable thicknesses. Air gaps really help to extend the effective frequencies downwards.
Hi Drumberg,
here comes a more detailed view on my system.
My listening room is located in the basement of our house. All walls including ceiling and floor ar either made out of
concrete or bricks. The measured frequency response of the untreated room showed peaks of more than 20 db caused by the (badly distributed) room modes and the hard surfaces of the room.
The room is an acoustical nightmare, but the combination of speaker placement, acoustic treatment and dsp based eq has solved some of the problems.
The lower bass speakers are Tad 1601 in 150 l onken style enclosures, but the reflex channels are completely blocked with damping material. The upper bass speakers are BD15s in 70l closed boxes.The bass enclosures are placed near the front wall to minimize boundary interferences. The horizontal and vertical placements of the bass enclosures are inspired by the dba concept (double bass array). There is no corresponding array on the back wall, but both back corners are stuffed with 50 cm deep acoustic absorbers.
Small remark: The in room frequency response of the 70l and the 150l boxes are not that different, as the room modes rule the bass response in small rooms. You have to eq the response anyway.
The horn satellites are driven by JBL 2490H compression drivers on 95 cm (tractrix?) horns and TAD TD4001 on 53cm (tractrix?) horns. The additional tweeter horns are B&C de10 on 18sound Xt120. The satellites are placed 1,5m in front of the back wall. The listening distance is 2,5 m.
All drivers are time aligned and controlled by a convolution based digital crossover located on an audio pc running jriver. The crossover filters and eq are done by acourate.
Early reflection are reduced by the directivity of the horn speakers and damping panels on both sidewalls, floor and ceiling. The result is a "dry" sound that requires some acoustical enhancement to make the speakers more listenable. This is done by two ambient speakers placed behind the listener. The ambient sound is generated by a combination of VST-plugins inside Jriver. I borrowed the idea from Wesayso, who presented this approach in his line array thread.
Feel free to ask further questions.
-Bodo
Hi,
I didn't have much time in the last days. I still need to do some room simulation especially with my 20m² just to see how it behaves. But I made some midbass horn simulation. I tried a ton of drivers until I had the idea to try the 15PR400 in a conical midbasshorn. The simulation doesn't look that bad either. It should be useable from 80hz to 500hz. Maybe I could pull 2db out at 150hz with eq. I even made it a little bit shorter (70cm vs. 77cm). What do you guys think? I needed to decide between 15PR400 in closed box without horn and 12PE32 (or other) in a midbasshorn. With the 15PR400 I can try both without buying different drivers.
@bodo: You have got a big system there
The tweeters look pretty far away from your tractrix horns. Does everything play as one closed system? Or can you hear that different drivers are playing?
50L would be a dream. But can you tell me a little more about the stuffing?
I already read a little bit about diy absorbers for studio use and the materials. But very useful to calculate myself before building something from the internet.
I didn't have much time in the last days. I still need to do some room simulation especially with my 20m² just to see how it behaves. But I made some midbass horn simulation. I tried a ton of drivers until I had the idea to try the 15PR400 in a conical midbasshorn. The simulation doesn't look that bad either. It should be useable from 80hz to 500hz. Maybe I could pull 2db out at 150hz with eq. I even made it a little bit shorter (70cm vs. 77cm). What do you guys think? I needed to decide between 15PR400 in closed box without horn and 12PE32 (or other) in a midbasshorn. With the 15PR400 I can try both without buying different drivers.
@bodo: You have got a big system there
The tweeters look pretty far away from your tractrix horns. Does everything play as one closed system? Or can you hear that different drivers are playing?Attachments
GHi,
Choice must be ruled by the off axis patern in the upper range for rhe mid crossover integration. The 12" should be better cause its excelent Le number and size, closed box possible, Qtc at 0,505 for time transcient and compression happy wedding...imho.
As youhace active amp and certainly no 3 W drivers you have to ask about horn utility in such frequency with in mind what is the Pi4 radiation patern at low frequency whatever the horn or not... If transcient is the targett more than electrical sensibility, you must focus on the best group delay imho in the drums fhz area you assume. The behavior of the filter at 80 hz has to be putt on the table as its slope towards the bass subs....bass strings vibration, piano....
Thought: Altec A5 is a good size floor spacing and sound behavior proof system in the 60hz-80hz to upper range you targett, while it has been created...for theater and movies venues. A sealed at a controlled 0.505 has the leading flag though... any way at the length waves below the room size length...the room will, once again, rules, whatever the EQ. In the transcient behavior and drums chest feeling you must think about upper harmonics till the trebles... think about indian tablas, indonesian small bells, the best the time behavior of the 80 to 200 hz drivers, the best the harmonics...just thinking horns and Sd is not a priority,
Imho, wmmv of course...
...
Choice must be ruled by the off axis patern in the upper range for rhe mid crossover integration. The 12" should be better cause its excelent Le number and size, closed box possible, Qtc at 0,505 for time transcient and compression happy wedding...imho.
As youhace active amp and certainly no 3 W drivers you have to ask about horn utility in such frequency with in mind what is the Pi4 radiation patern at low frequency whatever the horn or not... If transcient is the targett more than electrical sensibility, you must focus on the best group delay imho in the drums fhz area you assume. The behavior of the filter at 80 hz has to be putt on the table as its slope towards the bass subs....bass strings vibration, piano....
Thought: Altec A5 is a good size floor spacing and sound behavior proof system in the 60hz-80hz to upper range you targett, while it has been created...for theater and movies venues. A sealed at a controlled 0.505 has the leading flag though... any way at the length waves below the room size length...the room will, once again, rules, whatever the EQ. In the transcient behavior and drums chest feeling you must think about upper harmonics till the trebles... think about indian tablas, indonesian small bells, the best the time behavior of the 80 to 200 hz drivers, the best the harmonics...just thinking horns and Sd is not a priority,
Imho, wmmv of course...
...
Last edited:
Hi,
50L would be a dream. But can you tell me a little more about the stuffing?
I already read a little bit about diy absorbers for studio use and the materials. But very useful to calculate myself before building something from the internet.
I didn't have much time in the last days. I still need to do some room simulation especially with my 20m² just to see how it behaves. But I made some midbass horn simulation. I tried a ton of drivers until I had the idea to try the 15PR400 in a conical midbasshorn. The simulation doesn't look that bad either. It should be useable from 80hz to 500hz. Maybe I could pull 2db out at 150hz with eq. I even made it a little bit shorter (70cm vs. 77cm). What do you guys think? I needed to decide between 15PR400 in closed box without horn and 12PE32 (or other) in a midbasshorn. With the 15PR400 I can try both without buying different drivers.
@bodo: You have got a big system there
The acoustic path length of the upper tractrix horn is as long as the path length of the tweeter horn. I can't detect single drivers. Even the bass section is integrating nicely due to the digital delay on the horn speakers.
Don't waste your time on absorber construction, if your (new) listening room should have hard walls. I've tried different kind of absorbers. The best I could get was about 3-4 dB on a single room mode. Finally the combination of eq + broad band absorbers (acoustic hemp, melamin foam, Caruso iso-bond) solved the problem with room modes. Don't fear to use eq for solving problems with room modes. I had to use an attenuation of 16 dB on single bass frequencies. So 2 dB at 150 Hz is nothing to worry about.
-Bodo
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.