davewantsmoore Posted November 9, 2022 Posted November 9, 2022 On 8/11/2022 at 8:20 AM, frednork said: Glad its making some sense. My practical experience is that you may be able to partly eq your way out of that 50-90 dip but unlikely to be smooth and not totally removable. Yes. What the excess phase chart is saying, is that (due to some sort of reflection/diffraction, ie. something with a delay) even if you EQ it flat ... then the amplitude and the phase won't match. Also... what it implies...... is that the thing making the delay, is not going to be the same vs position in the room..... so while you could EQ the response flat. If you moved to another location (say the seat next to you), then the EQ would no longer be correct, because the amount/delay of the reflection which is dominating the response, will be different. On 8/11/2022 at 8:20 AM, frednork said: So yes multiple subs or changing position of the current ones to make it less troublesome I had a dream that he dismantled the bass cannons, and put each driver in it's own small sealed box .... and the wired them all up around the room, driven by the same amp(s) that are used now (but with a mono/summed signal, of course). On 8/11/2022 at 8:20 AM, frednork said: free 1
acg Posted November 9, 2022 Author Posted November 9, 2022 7 minutes ago, davewantsmoore said: No, you can also fix it by adding additional sources ... or by moving the speakers or the listener. I tried an additional source in two locations within the room and one location did exactly zero to the droop and the other location gave 0.5dB improvement over no more than 10Hz from 55Hz. Moving the listener forwards or backwards will help a little I am sure, but I have not measured anywhere other than where I prefer to sit for optimal (for me) soundstage envelopment. 7 minutes ago, davewantsmoore said: It may well be, that you cannot solve the issue you have by adding more subs (you would have to simulate what suits at various locations would do the response where there's the problem) ..... the the excess phase chart doesn't tell you that you can't solve the issue with more subs. Ok. Trouble is that REW cannot properly simulate my room with its cathedral ceiling. I have walked around the room with the SPL meter and earplugs in my ears (and annoying the neighbours and the chickens no doubt) playing sines of specific frequencies to find the parts of the room that exhibit the high and low pressures in order to try to develop a plan for room treatment. My gut feel at this stage is that the ceiling directly above my listening chair is contributing to these issues. Moving the chair forwards or backwards is likely just to shift where the droop sits in the frequency range. To that end I've had some zincalume sheeting cut to size and if I get a chance this weekend I intend to build some VPR's...four of them if I have time...and go see what happens with them in various parts of the room. I've also ordered some Basotect foam to make some ceiling panels or VPR's (the other VPR's will use a softer polyester foam which is not suitable for hanging a steel sheet from the ceiling). 13 minutes ago, davewantsmoore said: This is kinda the sorta thing I was trying to explain to you previously ..... when suggesting more subs Yes, I've never ruled out more subs, but I wanted to give the midbass channel a go first to fix that droop. There are two amplifier channels spare to go towards more subs. Also, when I built the preamp, which is exclusively single ended, I added a solid state buffer (that I can switch on and off) to act as a second output but this time balanced just in case I need more than two additional channels/subs. So, there is scope for at least four more subs/midbass, some with separate amplification if necessary.
acg Posted November 9, 2022 Author Posted November 9, 2022 14 minutes ago, davewantsmoore said: What the excess phase chart is saying, is that (due to some sort of reflection/diffraction, ie. something with a delay) even if you EQ it flat ... then the amplitude and the phase won't match. That is my interpretation as well. My hope is that some Kong level room treatment will be enough to disrupt that reflection/diffraction. We will see...it is a lot of dB to make back! 19 minutes ago, davewantsmoore said: Also... what it implies...... is that the thing making the delay, is not going to be the same vs position in the room..... so while you could EQ the response flat. If you moved to another location (say the seat next to you), then the EQ would no longer be correct, because the amount/delay of the reflection which is dominating the response, will be different. Ha! The system is already wholly focused on just one seat... 20 minutes ago, davewantsmoore said: I had a dream that he dismantled the bass cannons, and put each driver in it's own small sealed box .... and the wired them all up around the room, driven by the same amp(s) that are used now (but with a mono/summed signal, of course). Glad I'm not the only one dreaming of my system Dave. This could be done with the Bass Cannons, they are all individually sealed and wired. But I would hate the look of it. 1
davewantsmoore Posted November 9, 2022 Posted November 9, 2022 1 minute ago, acg said: That is my interpretation as well. Yes... so it doesn't say that EQ won't work... or that more sources won't work. Just that with EQ, it's position dependant... and with the additional sources, well, they have to be in the right place (which applies, always, anyways). 1 minute ago, acg said: My hope is that some Kong level room treatment will be enough to disrupt that reflection/diffraction. We will see...it is a lot of dB to make back! What is the "bass response RHS" chart that you posted exactly? 1 minute ago, acg said: Ha! The system is already wholly focused on just one seat... Yea, I know .... was just to illustrate the explanation. 1 minute ago, acg said: Glad I'm not the only one dreaming of my system Dave. This could be done with the Bass Cannons, they are all individually sealed and wired. Yep. 1 minute ago, acg said: But I would hate the look of it. Imagine normies surprise when they see your system, and you tell them that it wasn't "function over form". hehehe. 1
acg Posted November 9, 2022 Author Posted November 9, 2022 3 minutes ago, davewantsmoore said: What is the "bass response RHS" chart that you posted exactly? It is the right speaker, all six channels including midbass box, but with the midbass attenuated 24dB so it barely contributes to the SPL response. To be honest though, that section of the graph looks the same with no attenuation to the midbass.
davewantsmoore Posted November 9, 2022 Posted November 9, 2022 11 hours ago, acg said: It is the right speaker, all six channels Ok. Have you measured a single source of known flatness? (eg. your ML1?)
davewantsmoore Posted November 9, 2022 Posted November 9, 2022 12 hours ago, acg said: I tried an additional source in two locations within the room and one location did exactly zero If it is caused by a mode, and depending on which mode, the source big need to be in a very specific location (perhaps even not on the floor). My point isn't that "you will/should fix this with more woofers" (although I think that is by far and away the best plan, i you cannot move the speakers/listener) ..... rather that: "the excess phase measurement doesn't tell you it won't work". 12 hours ago, acg said: I have not measured anywhere other than where I prefer to sit for optimal (for me) soundstage envelopment. I figured as much. Fair. Especially with the room dims the way they are, I expect there's not much wiggle room anyways.
acg Posted November 9, 2022 Author Posted November 9, 2022 @davewantsmoorethe ML1's went years ago. I measured the Midbass box in-room at the cone and the frequency response was what was expected.
acg Posted November 9, 2022 Author Posted November 9, 2022 2 minutes ago, davewantsmoore said: If it is caused by a mode, and depending on which mode, the source big need to be in a very specific location (perhaps even not on the floor). Interesting thing for me is that the two positions tested for the Midbass box varied a lot vertically, but not a lot horizontally, which indicates to me that not only is the sloped ceiling causing the issue, that it is likely also an in-conjunctive interaction with some walls, probably the front/back walls. 1
acg Posted November 11, 2022 Author Posted November 11, 2022 On a lighter note, here is the site supervisor inspected the freshly delivered sheet metal for the upcoming VPR weekend... You are correct, that is not a dog, it is the orphan goat we let the kids raise in town. He's a very thorough inspector and no doubt will sh%t all over it by morning... 4 2
BioBrian Posted November 11, 2022 Posted November 11, 2022 Fit of the VaPoRs? Really looking forward to how these work. They in themselves should alter the reflections causing the modes?
acg Posted November 11, 2022 Author Posted November 11, 2022 12 minutes ago, BioBrian said: Fit of the VaPoRs? Really looking forward to how these work. They in themselves should alter the reflections causing the modes? They are a panel resonator with quite broadband operation. It is effectively a flat sheet of steel glued to some foam and put against the wall. You can effectively tune the panel via its dimensions, panel thickness, foam thickness and foam density. The major premise of the design is to be able to tackle very low bass frequencies with minimal intrusion into the room. We will see how effective they are in my situation...
BioBrian Posted November 11, 2022 Posted November 11, 2022 So, the wall becomes part of the structure? I was wondering, if they had independent backing sheets, whether you could try them in other areas, eg across the V of the ceiling (hence the ref to altering modes). But any news will be good. We've seen the frames a few times, if my long-term memory serves. OT: I have a soft spot for goats - had up to 120 Angoras for a while. Trouble is, the Devils also liked them, or should I say their newly-born children. Their intelligence and communication with humans was however greatly eclipsed by a very darkly funny "pet" sheep. 1
acg Posted November 11, 2022 Author Posted November 11, 2022 2 hours ago, BioBrian said: So, the wall becomes part of the structure? Yes. 2 hours ago, BioBrian said: I was wondering, if they had independent backing sheets, whether you could try them in other areas, eg across the V of the ceiling (hence the ref to altering modes). Also yes. 2 hours ago, BioBrian said: OT: I have a soft spot for goats - had up to 120 Angoras for a while. Trouble is, the Devils also liked them, or should I say their newly-born children. Their intelligence and communication with humans was however greatly eclipsed by a very darkly funny "pet" sheep. Yep, goats are pretty cool and are really useful in rougher, lighter country where they control the woody weeds and let the grass grow as a result. In conjuction with cattle they are hard to beat. 1
davewantsmoore Posted November 11, 2022 Posted November 11, 2022 19 hours ago, BioBrian said: Really looking forward to how these work. They in themselves should alter the reflections causing the modes? If they're designed correctly (they will work efficiently over a narrow range) , and cover a large area.... in theory you can completely nuke a "room mode".
acg Posted November 12, 2022 Author Posted November 12, 2022 4 hours ago, davewantsmoore said: in theory you can completely nuke a "room mode". I sure hope so. First one is done... 1.2m x 1.8m. I've set it up so that there is a 1mm steel sheet one side of the foam and a 2mm sheet on the other side. Have a look at this... ... the whole idea is that the resonant steel panel touches nothing but foam... see the gaps around the front... but 1mm will resonant differently to 2mm so with these test VPRs I can screw either the 1mm or 2mm sheet to the frame to change their performance. The frame is there for safety as I really do not want a steel sheet becoming unglued and damaging my hornsor decapitating my children. The frames also let me add diffusion to the front or adornment or mount the TV or whatever. There is plenty of room between the sides of the panel and the frame to perhaps stick a little more foam. The frame gaps can also be sealed to reduce the bandwidth of absorption if need be. I know what I'm doing now so off to build the second. 1
davewantsmoore Posted November 12, 2022 Posted November 12, 2022 11 minutes ago, acg said: 1.2m x 1.8m. I've set it up so that there is a 1mm steel sheet one side of the foam and a 2mm sheet on the other side. Have a look at this... ... the whole idea is that the resonant steel panel touches nothing but foam... see the gaps around the front... but 1mm will resonant differently to 2mm so with these test VPRs I can screw either the 1mm or 2mm sheet to the frame to change their performance. Aaaahhhh <lightbulb> Nice. 11 minutes ago, acg said: The frame gaps can also be sealed to reduce the bandwidth of absorption if need be That's going to make bgrall difference (right on the wall)
acg Posted November 12, 2022 Author Posted November 12, 2022 2 hours ago, davewantsmoore said: That's going to make bgrall difference (right on the wall) No difference at the bottom end, but certainly a difference at the top end. See this graphic for some smaller and thinner VPR's that use different foam... Placed on the floor (green and red) of a reverberant room with VPR sides open or closed the difference starts at about 200Hz. The blue line with the same VPR's placed in wall and floor verticies of the same reverberant room are pretty impressive. If I want this panel to be even more broadband all I have to due is place some more foam on the front (room side) of the resonating panel and I could likely extend effective absorption past 5kHz. So flexible...assuming I can get them to work properly in the first place...
almikel Posted November 12, 2022 Posted November 12, 2022 @davewantsmoore, based on @acg's excess phase plot On 08/11/2022 at 6:26 AM, acg said: The excess phase is "not zero" anywhere, but it's flat but gently falling between low 20's to 50Hz, then mostly flat again from 90Hz up with a big peak between 50-80Hz. My understanding of excess phase is that it needs to be zero for the response to be "minimum phase". Clearly the peak between 50Hz - 90Hz is not minimum phase, but could you help me understand why the "mostly flat but not zero" parts of the excess phase curve aren't minimum phase? My understanding of "minimum phase" is also limited, other than that room effects that "are minimum phase" can be rectified by appropriate EQ - the EQ (which is also minimum phase) will fix both the time domain and the frequency domain responses. I don't want to take @acg's thread too far off topic - I'm happy to start another thread. cheers Mike
acg Posted November 12, 2022 Author Posted November 12, 2022 @almikelthis is in topic so discuss away. My interpretation is that the non- zero thing is due to delay via the electronics and then the sound waves leaving the speaker and finding the microphone. 1
davewantsmoore Posted November 13, 2022 Posted November 13, 2022 18 hours ago, acg said: Placed on the floor (green and red) of a reverberant room with VPR sides open or closed the difference starts at about 200Hz. Exactly. I was only considering ~ < 200Hz.... (Because) with a narrow DI speaker, sitting quite close .... do you want to be adding absorption > 200 Hz ?!!? (I would suggest, no).
davewantsmoore Posted November 13, 2022 Posted November 13, 2022 13 hours ago, almikel said: @davewantsmoore, based on @acg's excess phase plot The excess phase is "not zero" anywhere, but it's flat but gently falling between low 20's to 50Hz, then mostly flat again from 90Hz up with a big peak between 50-80Hz. My understanding of excess phase is that it needs to be zero for the response to be "minimum phase". Something sort of measurement issue (time delay embedded in the data?) I'm happy to only consider where it peaks, and write off the rest as "flat"..... but like with most of this sort of stuff, it is very hard to say, without knowing the precise measurement / data setup, and playing around with the setup to confirm that what is being recorded is actually representative of reality. 13 hours ago, almikel said: Clearly the peak between 50Hz - 90Hz is not minimum phase, but could you help me understand why the "mostly flat but not zero" parts of the excess phase curve aren't minimum phase? As above, it could be some sort of issue with the data ..... it could also be that multiple out of phase sources are playing over the entire range, but I'd expect it to be a little bit more wonky if that were the case. 13 hours ago, almikel said: My understanding of "minimum phase" is also limited A very simple way to picture it is to say: We recorded a certain amplitude (SPL) vs frequency In theory, the phase response should be X (eg. if the frequency response was flat, the phase reponse should be flat) What is the difference between the actual measured phase, and the phase we should expect for this amplitude response? (this is the "Excess phase") What could cause the phase to be "wrong" for a certain measured amplitude? ..... something that arrives "late". ie. a reflection/diffraction.... or two speakers were respell sable for the measured SPL but they are not in time.
davewantsmoore Posted November 13, 2022 Posted November 13, 2022 13 hours ago, acg said: My interpretation is that the non- zero thing is due to delay via the electronics and then the sound waves leaving the speaker and finding the microphone. You should get a measurement setup where this is corrected for... you can use either a timing channel, or an acoustic reference. You can also ask REW to "guess" the correction, which works fine for a single measurement.... but it will make (the timing of) one measurement not comparable to another, as the "guess correction" may not be the same.
acg Posted November 13, 2022 Author Posted November 13, 2022 17 minutes ago, davewantsmoore said: You should get a measurement setup where this is corrected for... you can use either a timing channel, or an acoustic reference. You can also ask REW to "guess" the correction, which works fine for a single measurement.... but it will make (the timing of) one measurement not comparable to another, as the "guess correction" may not be the same. I do. Left channel is looped back as timing reference and right channel feeds the preamp, or the other way round, I can't remember.
acg Posted November 13, 2022 Author Posted November 13, 2022 37 minutes ago, davewantsmoore said: (Because) with a narrow DI speaker, sitting quite close .... do you want to be adding absorption > 200 Hz ?!!? (I would suggest, no). Well, I'm not sure. At the moment the room sounds ok, but I think it could do better at really loud volumes which I sometimes/ often do partake. At the end of the room tuning process I figure I'll experiment with a little bit of higher frequency absorption to nut out the issue of not incidentally sorted by then. It's not major... think 100dB perhaps.
Recommended Posts