@docali I think the TMM 2.5 way is great solution as well. My configuration for that, if I chose that route I would put the 2 15's under the tweeter, the lower 15 and 2 18"s on the rear creating 0.5 section. I really like that idea too. All front woofers are allowed to carry full bass, and the tweeter mating woofer would still never cross 2mm excursion.
TM vs MTM
Who you guys think gives the better FR, generally speaking? Quality of FR has been a focus for this design as well. I could argue that FR should take slight priority over imaging. With that being said, a TM has smoother dispersion above the tweeter. The extra comb filtering brought about with a MTM, how detrimental is it to the final FR I wonder. In order to stay linear without speculation, @weltersys, I really only need to raise the crossover to about 300hz. Comb filtering isn't major at the point I do not think, off hand, for my CTC. Having the bass woofers at the floor baffle provides really good FR as well. The MTM would place 2 woofers in the air, to what consequence I am not sure yet.
TM vs MTM
Who you guys think gives the better FR, generally speaking? Quality of FR has been a focus for this design as well. I could argue that FR should take slight priority over imaging. With that being said, a TM has smoother dispersion above the tweeter. The extra comb filtering brought about with a MTM, how detrimental is it to the final FR I wonder. In order to stay linear without speculation, @weltersys, I really only need to raise the crossover to about 300hz. Comb filtering isn't major at the point I do not think, off hand, for my CTC. Having the bass woofers at the floor baffle provides really good FR as well. The MTM would place 2 woofers in the air, to what consequence I am not sure yet.
Last edited:
Part of the woofers in the air means less efficiency but also a more even excitation of room modes compared to all woofers on the floor.
I.e. you "waste" a bit of the SPL gained by the additional woofers in favour of a little more bass precision.
Regards
Charles.
I.e. you "waste" a bit of the SPL gained by the additional woofers in favour of a little more bass precision.
Regards
Charles.
What phase issues are you trying to fix? Between what driver sections exactly?
In my previous comment I took it to be linearizing the bottom end roll-off, as phase issues between driver sections are a matter of proper xover design, and not really "fixing" phase.
@phase_accurate Thank you for responding, Im interested in why you say having a woofer in the air, and exciting modes, is good for the response. In my experience it appears that the closer the source is to the walls, the better. If you continue to get closer to wall, you are eventually, in the wall, soffit or flush mounted as they say. This is great for FR as we already know. A woofer sitting at the floor for example, is as close as possible to being in the wall, and the frequency response has reflected that in my experience
@mark100
In the left chart, the black response line represents the summed response. The red line is the PPMSL response and the green line is a regular oriented woofer response. The dips in the response are from the areas that are out of phase, as you can see neither the red or the green line has any of these dips.
In the left chart, the black response line represents the summed response. The red line is the PPMSL response and the green line is a regular oriented woofer response. The dips in the response are from the areas that are out of phase, as you can see neither the red or the green line has any of these dips.
@camplo I never wanted to go with one midrange driver. Having 2 brings a lot of benefits.
The listening situation for this speaker is a well controlled room with close distance for mixing/mastering or bigger distance for cinema.
So I did my listening tests, made 2 presets and switched quickly between one or 2 midranges. Something I would recommend to do when standing at the decision how to use your TMTs - also compare in real live, not only theory.
I often described why I chose only one midrange - point source vs vertical "dimension" of the source. I'm sure you can't hear that difference in 3m distance in a living room, which doesn't mean it's not there ;-) I always liked (good) coax speakers and small satellites for their pin point behaviour - I can move around and the sound source stays stable. Maybe cause I have a live mixing/PA background - I like stable sound all over the room.
Driver distance is 83-85mm for my Bliesma construction, cross over frequency 1,8-2.0kHz.
It's "modern" to use wide CTC distances to steer the 0 away from the ceiling reflection point. It's stronger and more prominent there but can have less influence at your average listening position. So it's understandable that people chose to do so - but I don't. 🤓
Many of my personal listening needs are NOT the average listening position! I listen in 1m distance for mixing - with 1-3 more people in the room at different positions who also need consistent sound. I listen in well treated rooms with ceiling absorption (yes, also in my living room and kitchen 🤓). We as family listen all over the living room, not just in one chair. And there are 5 people on a big couch when watching a movie.
For me there are 2 ways to solve that issue. When you want wide radiation in a speaker (consistent sound with lot of HF energy also off axis) I go (KEF) coax or close distance of midrange and tweeter. As point source as possible.
When you want/need a controlled radiation in a speaker (reverberant room but you still need detail/direct sound) you go with wavegudies/horns and digital corrections + crossovers to control speaker delays and make the crossover steep enough without to many downsides.
Both work, both have different use cases.
My big studio/home cinema speaker is a wide radiation design to get even sound in a controlled room. And works VERY well for that.
In my living room I have controlled speakers at the front and coax/wide radiating speakers at the rear cause the ROOM and the positions demand for that. Tried KH120ii for rear speakers (which are positioned closer as they should as in many living room cinemas) - wide + coax radiation works better for uniform sound.
There is no BEST solution - there is always just the best solution for your use case.
The listening situation for this speaker is a well controlled room with close distance for mixing/mastering or bigger distance for cinema.
So I did my listening tests, made 2 presets and switched quickly between one or 2 midranges. Something I would recommend to do when standing at the decision how to use your TMTs - also compare in real live, not only theory.
I often described why I chose only one midrange - point source vs vertical "dimension" of the source. I'm sure you can't hear that difference in 3m distance in a living room, which doesn't mean it's not there ;-) I always liked (good) coax speakers and small satellites for their pin point behaviour - I can move around and the sound source stays stable. Maybe cause I have a live mixing/PA background - I like stable sound all over the room.
Why claiming about CTCs? Just do the calculation?
Driver distance is 83-85mm for my Bliesma construction, cross over frequency 1,8-2.0kHz.
It's "modern" to use wide CTC distances to steer the 0 away from the ceiling reflection point. It's stronger and more prominent there but can have less influence at your average listening position. So it's understandable that people chose to do so - but I don't. 🤓
Many of my personal listening needs are NOT the average listening position! I listen in 1m distance for mixing - with 1-3 more people in the room at different positions who also need consistent sound. I listen in well treated rooms with ceiling absorption (yes, also in my living room and kitchen 🤓). We as family listen all over the living room, not just in one chair. And there are 5 people on a big couch when watching a movie.
For me there are 2 ways to solve that issue. When you want wide radiation in a speaker (consistent sound with lot of HF energy also off axis) I go (KEF) coax or close distance of midrange and tweeter. As point source as possible.
When you want/need a controlled radiation in a speaker (reverberant room but you still need detail/direct sound) you go with wavegudies/horns and digital corrections + crossovers to control speaker delays and make the crossover steep enough without to many downsides.
Both work, both have different use cases.
My big studio/home cinema speaker is a wide radiation design to get even sound in a controlled room. And works VERY well for that.
In my living room I have controlled speakers at the front and coax/wide radiating speakers at the rear cause the ROOM and the positions demand for that. Tried KH120ii for rear speakers (which are positioned closer as they should as in many living room cinemas) - wide + coax radiation works better for uniform sound.
There is no BEST solution - there is always just the best solution for your use case.
Ok, just get the phase traces to overlay in whatever frequency range they will be asked to sum. (down to -20dB is decent, further down better)
Ripple will go away. Play around in Vcad, using their low-passes in place.
What is your acoustic goal for each section? Why are you using both?
I look at the green and red trace above, and I wonder, how exactly how to get that to overlay on top of each other lol.....
If I had had two PPSL to mate together, it would be two phase traces like the red trace but unidentical, and still I will wonder how to get them to overlay on top of each other if they are not identical. @mark100.... the question is, can I make the Red trace look like the green trace, or can I make the red and green trace, flat, with DSP. Is that a challenge or is that normal business.
if I were to build my idea, which is MTM made of PPMSL. The individual FR is fine for each section. But the phase will be unidentical thus, cannot be overlayed over top of each other with out the help of DSP. If by "just get the phase traces to overlay" you are speaking about using DSP... sure, as long as you say its possible. I believe you.
The acoustical goal is a MTM made from PPMSL in which I can use various Xover points. I am still learning the off axis nature of slots.
TM, especially with the center to center distance forced by your huge "tweeter" horn.
MTM is detrimental for both imaging and (reflected) frequency response.
A single front loaded sealed 15" will be under your 2mm excursion goal at 115dB if you cross it at 120Hz.
Amps are cheap, go tri-amp as you planned and stop effing around with all the compromises.
Put your subs in the corner if you want to get some additional room gain from them, and place the mid-high in an equilateral listening triangle at your desired distance.
Two of the same PPSL should have the same phase traces, measured at the same distance. (or rather close enough to be considered identical)
1. With IIR and delay, and perhaps all-pass, you can get the green trace to match up to the existing red trace.
2. With FIR on the red trace, you can get it to match the existing green trace.
3. With FIR on both, you can make them match up to, about whatever acoustic target works for you
FIR will add latency for either 2 or 3.
All are normal business once one gains enough experience with REW, VCad, favorite simulator, etc.
Thanks for joining the convo. My tweeter horn is elliptical, and on the vertical, is 19.5" tall which isn't small is it. The NS15 wave guide is 18", for reference.
You are not wrong lol...
I can do both, in the corner, and equilateral listening triangle. The beauty of these rectangular basement spaces.
Thats being very conservative as that's the anechoic response. All the noise I was making in the other thread about PPMSL was to fuel my project. With all 18"s and 15"s in use I have 4096 sd per channel. The goal was to get all woofers under 2mm. 120db definitely not but lower, more likely
Not going sum as well as this sum suggest but also not listening in 2.0XPi. All 8192 cm2 gets me 2mm at 30hz at 110db/1m. Thats pretty good for music.
I know I am not wrong about the Central presentation of MTM, but I also know about comb filtering and its effects on the Final FR. I know about it, not mastered it, btw. So if you are saying what it is that I already was suspicious of, then I am correct in my suspicion's.
So anyway, if I you are saying all that you've said, which you are, saying it. That would be a vote for the TMM Rear-MM 2.5way. The top M would mate to the tweeter and the rest of the M's would cover the lows in combination with the top M. No high pass needed on the top M because efficiency is high enough to keep excursion <2mm. Or if a higher pass is needed, it is readily available.
The two PPSL would not be identical. I'm not certain option one works in this case. I was under the impression that 2 and 3 work, but I have not the experience to say their limitations. I also want to be able to use the system, without additional latency, when desired, without the detriment to FR, so I think that means this PPSL MTM idea isn't floating.
I don't think Vcad has a phase EQ. I am beginning to see that an allpass has potential, though. Without seeing how much latency I've caused... whats the point? The sims phase is nothing like the real life phase, and I've never measured PPMSL in real life, just the PPSL. Sounds like work.
Last edited:
Room modes spoil the low-end.
PPSL is one solution favoured by quite few respected people from the DIY community (SLOB is another).
I lack personal experience with the concept, but have listened to a few systems with PPSL subs(ystems).
Corner placement/loading can work very well, but - again - depending on the room.
The marketing blurb for the sub bass solution below that Aussie MTM:
"Each pillar sits atop an extremely large sub bass system which displace well over a cubic metre each. Dual 21” opposed drivers fire into a cavity, whereby their mechanical forces cancel, resulting in purely acoustic energy that vents both fore and aft - with rearward reflex ports tuned well inside the ordinary ‘infra’ bass region. This segment has 2200w RMS on tap, for bass that is controlled, punchy, deep, thrilling. And there’s plenty of it, and the rest."
The complete system (retail price: 120k) is powered by Pascal class D amplifiers and equipped with DSP.
I'd say it basically consists of a large waveguide flanked by two direct-radiating 15s + a compact horn subsystem.
IOW, short path lengths.
Earlier I referred to some systems from the Norwegian DIY community.
Since your Tratrix horn is quite deep, it makes sense to 'hornload' the mids and subs as well.
In this way you increase efficiency and improve alignment of the 3 channels, but such systems tend to eat up a lot of space.
Someone shared his experience at the GT-Sound factory with me.
He listened extensively to the 'Godzilla' (as they call it) and was absolutely floored, a.o. because it didn't sound like a horn system at all.
A customer's system:
Contrary to what you might expect, these are not (infra)bass monsters.
Last edited:
Hopefully the Norwegians won't blame me for stealing some pictures 😉
Beginning:
Concept that might work, but will be expensive:
To end up with 21" horn subs:
Beginning:
Concept that might work, but will be expensive:
To end up with 21" horn subs:
Another option is a horn loaded OB in WMTMW-fashion, similar to docali's system and Bert Doppenberg's original Orelo prototype.
Though I'm not sure the 18s and 15s like to share the same horn mouth and slot.
Combined with your elliptical horn this should (in theory) approximate a point source.
In any case, the 'wings' can be constructed modularly.
Though I'm not sure the 18s and 15s like to share the same horn mouth and slot.
Combined with your elliptical horn this should (in theory) approximate a point source.
In any case, the 'wings' can be constructed modularly.
Last edited:
That was 8 years ago, but a smaller variant was used at ETF this year:
