Primare Knob Posted January 8, 2019 Author Share Posted January 8, 2019 On 07/01/2019 at 5:10 PM, davewantsmoore said: I really think you need to forget everything you think you know, and go back to basics I am thinking about it, and I am also thinking about starting a new thread, .....................Part 2. I hope to create mote structure so that it might help other people better. Link to comment Share on other sites More sharing options...
Primare Knob Posted January 8, 2019 Author Share Posted January 8, 2019 On 07/01/2019 at 5:53 PM, davewantsmoore said: On 07/01/2019 at 8:44 AM, Primare Knob said: That would be the better solution when it comes to placement per type of trap, but I think you will be at risk, that the pressure traps end up to narrow, and not deep enough to be effective in the low frequency range, unless you can sacrifice a lot of space out of the corner. I don't think that the distance to the wall has effect on the pressure trap as I believe it is the amount of volume that dictates the efficiency. I could be wrong here, as always, its hard to find a clear answer. Distance to the boundary is always critical. For a velocity trap (ie. a fluffy thing) it will work zero against the wall .... and maximally 1/4 WL from the wall. They need to be thick to work at low frequencies. How much effect they have depends on the surface area. For a pressure trap (ie. a vibrating panel) it will work maximally against a wall .... and zero at 1/4 WL from the wall. How much effect they have depends on the surface area. Where they operate depends on the "tuning" of the vibration, which depends on the size, mass, stiffness, etc. of the vibrating panel. Understanding how these things work, is less important than quantifying your problem, and solving it efficiently. It's almost certain that the majority of the solution will be EQ. I was talking about a triangular shaped pressure trap. Since there is more pressure in a corner than just along the wall. And in the calculation for the efficiency of a trap, it doesn't matter if the distance start narrow, increases, and narrows down again, as it is the amount of volume that matters. I was also thinking along the lines, that if you build it into the corner that sound waves might have less change to go around it as it becomes part of the room boundary. The placement for the absorber or pressure type traps are well understood. The idea of Almikel is a good one for placement, but what I was aiming at is, the pressure traps need a certain width to be effective in the low frequency, and also need a certain volume (read depth). Place two of these in the corner, and they will already claim a lot of real estate. Straddle some absorber across the corner and you end up with a trap that is only suitable for people with very large rooms. Link to comment Share on other sites More sharing options...
Primare Knob Posted January 8, 2019 Author Share Posted January 8, 2019 On 07/01/2019 at 5:53 PM, davewantsmoore said: Quote If I build traps that effect the decay time but also the SPL response I either have to place the subwoofers at a different position, or I will need to crank them up by a lot to compensate. This is not a problem ... If you determine you have too little decay, then you should address that if possible, irrespective of the speakers It is the opposite that I am worried about. Affecting the decay time is what I want, but I don't want it to effect the SPL response in the region below 70Hz, as it is rather good. The question here is, if you build a trap to deal with decay time, will it automatically influence the SPL graph. And if it influences the SPL graph, is this a good thing, even if the response seemed good to begin with? Link to comment Share on other sites More sharing options...
Primare Knob Posted January 8, 2019 Author Share Posted January 8, 2019 (edited) On 07/01/2019 at 6:14 PM, davewantsmoore said: What are the room dimensions? The biggest change is at 30Hz. "No trap" is missing everything below ~30Hz. Something is wrong with your measurement(s). Quote I haven't sealed that off properly because I wanted to try different tuned versions of the trap by changing the front panel. I could seal the front panel up and have two different tuned traps and try again, but not sure if this is going to help or not. Is it supposed to be sealed in the design you built? Quote The 100Hz peak is a room mode (width and possibly length 101.5Hz), and the 70Hz peak could possibly be a room mode (height) Did you have any EQ applied to the woofers, for these measurements? We saw EQ you were applying earlier which was not good .... so it's difficult for me to look at your chart and say where the "baseline" is. (Eg. It looks about 88dB) You either need to make sure the basic EQ for the woofer is correct (you were PMing me about this a month or two back) .... or turn off the EQ, and just work with the raw driver response for now. Basically what we need to do is get some idea of what is the woofer and what is the room. If we work with the raw driver response to begin with, we can have a good go of drawing a response of what we should expect (12dB/octave drop below box tuning, plus some room gain). Room dimensions are 487cm deep, 337cm wide and 247cm high. Yes the design needs to be sealed, and has been sealed for the third attempt, which showed no visual changes in the data. Hence the comment that the traps are officially flawed, probably due to their size. No EQ is applied unless it is mentioned. (The measurement are from my main speakers, as they still show the room modes, and produce everything above 80Hz, as I should make that more clear) Edited January 8, 2019 by Primare Knob Link to comment Share on other sites More sharing options...
davewantsmoore Posted January 9, 2019 Share Posted January 9, 2019 5 hours ago, Primare Knob said: And if it influences the SPL graph, is this a good thing Yes.... but you should not be evaluating the SPL ..... fix the decay issues by adding 'treatment' .... Then look at the SPL. Link to comment Share on other sites More sharing options...
davewantsmoore Posted January 9, 2019 Share Posted January 9, 2019 5 hours ago, Primare Knob said: No EQ is applied unless it is mentioned. (The measurement are from my main speakers, as they still show the room modes, and produce everything above 80Hz, as I should make that more clear) I'm confused. What making the sound below 80Hz? Link to comment Share on other sites More sharing options...
Primare Knob Posted January 9, 2019 Author Share Posted January 9, 2019 I'm confused. What making the sound below 80Hz?Front speakers with no crossover. I wasn't expecting any results below 80Hz, so used the front left speakers to test the traps. 1 Link to comment Share on other sites More sharing options...
almikel Posted January 10, 2019 Share Posted January 10, 2019 On 09/01/2019 at 8:38 AM, Primare Knob said: Recalculate the above using a 400mm distance, and 400mm fluffy vs 100mm XHD with a 300mm air gap models nearly the same. your definition of "nearly the same" is different to mine...which is somewhat funny as I've said the same thing numerous times and Dave has pointed out my "nearly the same" wasn't so near On 09/01/2019 at 8:38 AM, Primare Knob said: And then it comes down to cost, what is cheaper, 400mm of fluffy or 100mm of XHD agreed - it's a long time since I costed XHD - a quick googleweb search for fluffy shows $54 for a bag of 6 Greenstuff R2.5 1160mm x 580mm x 90mm batts On 09/01/2019 at 9:27 AM, Primare Knob said: Room dimensions are 487cm deep, 337cm wide and 247cm high. For a "superchunk" style trap of fluffy that would be 5 bags (with some crush) for a 1160mm wide x 580mm deep - floor to ceiling trap (batts simply stacked flat floor to ceiling) - $270 per trap... ...more expensive than I expected... Assuming stacked in a corner (no straddling), the length mode will "see" 1160mm deep and 580mm wide and vice versa for the width mode (opposite if you change the orientation of the trap) - "similar" performance in each direction...and as expected with absorption, effectiveness dropping with frequency. ...difficult to compare with XHD straddling a corner, as the modelling tools don't cater for straddling, but let's just run with a 1200mm wide sheet of 100mm XHD straddling the corner with 600mm deep air gap to the corner and assume an "effective" air gap depth of 300mm (50% of the 600mm depth to the corner) Even if we say the "effective" depth of the air gap of the straddled XHD is 500mm we get Large fluffy traps seem to "out perform" denser materials for bass trapping using absorption. Unfortunately the effectiveness of absorption reduces at lower freq. Truly massive absorption is required to get to 100Hz - pressure traps and/or EQ is required <100Hz Mike Link to comment Share on other sites More sharing options...
davewantsmoore Posted January 11, 2019 Share Posted January 11, 2019 12 hours ago, almikel said: your definition of "nearly the same" is different to mine...which is somewhat funny as I've said the same thing numerous times and Dave has pointed out my "nearly the same" wasn't so near Aside from the ripples, these are essentially the same. The reason is that the most important parameter, how far the absorption is from the wall (ie it starts at 400mm) is the same. The simulator over represents the ripples, which will not be as sharp or as big in practise. http://www.acousticmodelling.com/mlink.php?im=1&s13=2&d13=100&v13=12000&s14=1&d14=300&s24=2&d24=400&v24=12000 Don't make the mistake of splitting hairs, or reading to much into things... it will cause you to miss what matters, and to try and optimise/chase things which are mirages..... or worse, cause you to misdiagnose why you see a certain result in your testing. Link to comment Share on other sites More sharing options...
davewantsmoore Posted January 11, 2019 Share Posted January 11, 2019 On 09/01/2019 at 10:11 AM, Primare Knob said: I was talking about a triangular shaped pressure trap. Since there is more pressure in a corner than just along the wall. And in the calculation for the efficiency of a trap, it doesn't matter if the distance start narrow, increases, and narrows down again, as it is the amount of volume that matters. The amount of surface area is important (for how much sound it will "catch") and the distance from the wall is important .... but at low frequencies it's hard for it not to be "close to the wall". In practical terms, I would build either type of trap "straddling the corners" ..... the reason being to work at low frequencies they need to be large, and the corner is really the only place they can go, unless you have a wall you can dedicate to a pressure trap .... or maybe you can hang a velocity trap from the ceiling, sometimes called "cloud". The unfortunate reality is that unless you are prepared to give up a lot of space, the traps will be relatively ineffective ..... although theory and calculators will probably understate this (they will be more effective than expected). Honestly, I would just put some "superchunks" in the corners.... and be done. Optimise multiple subwoofers independently. On 09/01/2019 at 10:11 AM, Primare Knob said: I was also thinking along the lines, that if you build it into the corner that sound waves might have less change to go around it as it becomes part of the room boundary. If the trap is tuned right, then sound of the appropriate frequencies will make the trap move, as opposed to "going around it". Link to comment Share on other sites More sharing options...
Primare Knob Posted January 11, 2019 Author Share Posted January 11, 2019 What I have now is a form of superchunck style traps in the front corners of the room.At 550mm out from the front wall, and 550mm out from the side wall are two sheets of 50mm 48kg/m2 forming a square tower. The inside is filled up with fluffy R1.7 stuff, all floor to ceiling.I have room at the front wall to build traps up to 550-600mm out from the front wall. I still need space for a center speaker behind an AT screen, and room for equipment with some air for ventilation.At the back of the room I have some space 1100mm deep behind the sofa, up to 900mm high that can be used for trapping. Not along the full width of the room as there is a door in the Left rear corner.At the moment I have 2 layers of 50mm 48Kg/m2 absorbers 900mm high along the full width of the sofa, at 1000-1100mm from the back wall.At the moment I have a sheet of 1200x2400x50mm 48kg/m2 suspended slightly lower than the exposed roof joists roughly in the middle of the room. I am planning to cover the entire ceiling with absorbers inbetween the exposed roof joists, and hold them in place with a binary pattern of slats. I am still looking for better decay times below a 100Hz if possible. What kind and size trapping should I look into?The biggest culprits seem to be 100Hz, (second width and or third length mode) 70Hz (second length and or height mode), and 35Hz (length mode), but actually everything below 100Hz has to much decay time.The bigger question I might probably need to ask is: How can I be sure that what I measure and read is actually real and related to my room? How can tell reality from phantoms under a 100 Hz, as I understand that it is not that simple. Link to comment Share on other sites More sharing options...
almikel Posted January 11, 2019 Share Posted January 11, 2019 3 hours ago, Primare Knob said: What I have now is a form of superchunck style traps in the front corners of the room. At 550mm out from the front wall, and 550mm out from the side wall are two sheets of 50mm 48kg/m2 forming a square tower. The inside is filled up with fluffy R1.7 stuff, all floor to ceiling These are decent sized traps...you've probably told me already, but what is the construction of the room? With those sorts of issues <100Hz, I'm guessing multiple brick/block walls, slab floor - ie rigid boundaries? 3 hours ago, Primare Knob said: I am still looking for better decay times below a 100Hz if possible. What kind and size trapping should I look into? An option that requires room construction and $ would be to install furring channel and ordinary gyprock with fluffy behind over the rigid walls. That would make the room "lossier" - the gyprock operates as a large membrane trap - no need for multiple layers of gyprock and Greenglue as you're not chasing isolation (the brick does that), and you want the gyprock to flex. 3 hours ago, Primare Knob said: I am planning to cover the entire ceiling with absorbers inbetween the exposed roof joists, and hold them in place with a binary pattern of slats Unfortunately the modelling tools don't cater for 1D BAD patterns, but I would do the same thing (use a 1D BAD pattern) - you can use the Porous Absorber Calculator to model a simple regular slat pattern - see how it models with your joist depth etc - the ceiling is a big surface, so I would expect a good result. 3 hours ago, Primare Knob said: The bigger question I might probably need to ask is: How can I be sure that what I measure and read is actually real and related to my room? How can tell reality from phantoms under a 100 Hz, as I understand that it is not that simple Good process and lots of measurements, documenting the measurements along the way, from multiple mike positions with and without treatment. Leave the mike in exactly the same place for a measurement before and after treatment is placed. Turn off extraneous noise sources (pool pumps, washing machines, air conditioners) Re-do a measurement if a plane flies over Have the volume high enough to produce a good signal to noise ratio post graphs here for input focus on the RT60, Decay, Waterfall and the Spectrogram tabs in REW - and the Impulse graphs to identify specific reflections that could be causing SBIR beg for forgiveness from the family for the endless "Woop Woops" It's laborious, but over time you'll learn how to interpret the measurements. cheers Mike Link to comment Share on other sites More sharing options...
Primare Knob Posted January 13, 2019 Author Share Posted January 13, 2019 (edited) Ok, I am going to post a bunch of pictures from measurements taken yesterday. First batch will be the Left Right and Center listen postion, before and after adding absorbtion around the room for Left, Right and Sub speakers Second batch will be of the Center listen position Subwoofer graph with absorbtion in place. I am posting all graphs that REW can produce, but not sure how to use them. Third batch is a picture of my left speaker and the influence on the 40-80Hz range with positioning. I came across this post on gearslutz (year 2010) from the author of REW. " Waterfall plots cannot be used to read decay times, unfortunately. The plot is produced by moving a window through the response and plotting the frequency content within that window. Using longer windows makes decays appear to take longer, which is at its heart a manifestation of the time/frequency trade-off of these kinds of analysis. Longer windows give increased frequency resolution, but correspondingly reduced time resolution. On the plus side, there are ways of measuring both the frequency and decay time of modal resonances with high precision, I'll include that capability in the next REW V5 beta in a few days time. " I am not sure what method he is referring to, regarding the proper interpretation of decay time. Legenda explanation: First letter L_ R_ Subs_ refers to the speaker Second letter Lpos Rpos Cpos refers to the listen postion Base refers to the untreated room Absorb refers to the absorption treatment put in place Edited January 13, 2019 by Primare Knob Link to comment Share on other sites More sharing options...
Primare Knob Posted January 13, 2019 Author Share Posted January 13, 2019 (edited) Left Speaker, before and after. Var smoothing has been applied to the no Absorbtion graph to make it more readable. Edited January 13, 2019 by Primare Knob Link to comment Share on other sites More sharing options...
Primare Knob Posted January 13, 2019 Author Share Posted January 13, 2019 (edited) Right speaker, before and after. Var smoothing has been applied to the no Absorbtion graph to make it more readable. Edited January 13, 2019 by Primare Knob Link to comment Share on other sites More sharing options...
Primare Knob Posted January 13, 2019 Author Share Posted January 13, 2019 Subs, before and after Link to comment Share on other sites More sharing options...
Primare Knob Posted January 13, 2019 Author Share Posted January 13, 2019 Sub graphs generated from the Center position with Absorbtion in place. Not sure how to interpret the distortion graph, but might this explain why there is an uneven decay rate/time at 20Hz? Looking at the waterfall and decay graph, I would like to do more work on the room, but the RT60 graph doesn't seem to bad. Link to comment Share on other sites More sharing options...
Primare Knob Posted January 13, 2019 Author Share Posted January 13, 2019 I have been trying to get rid of the nasty dip between 50Hz - 70Hz, and only moving the mic closer to the side wall seem to make an impact. I would guess that the problem here is SBIR. The same happens in the other graphs and sub and right speaker in a similar region. My biggest question is, how can I locate this SBIR? I have walked around the room along the side wall, at various distances with 200mm thick absorbtion, but never been able to make a difference. What forces are in play, and how do they work? Link to comment Share on other sites More sharing options...
Primare Knob Posted January 19, 2019 Author Share Posted January 19, 2019 @davewantsmoore @almikel Any thoughts or ideas you would like to share on these measurements. Are they any useful, and what does it tell me? Link to comment Share on other sites More sharing options...
Primare Knob Posted January 19, 2019 Author Share Posted January 19, 2019 (edited) I have mapped some pressure point in my room using specific frequencies from the decay graph of my subwoofers. 35, 49, 69, 98Hz Numbering is starting in the front left corner of the room, and next point is 1/4 wall length away from the next one. The highlighted sections are the pressure point between the sub and the wall/window. Position 3 is where my Left sub is placed at the 1/2 way length, and position 11 is where my Right sub is placed at the 1/2 way length. It is rather clear how leaky a windows is compared to a double brick wall. Looking at this myself I would draw the following conclusions: 35Hz trap would be best suited in the front of the room as that area shows up as the biggest pressure zone. 49Hz is all around quite equal, but shows a weaker area in the back of the room. 69Hz is a clear winner for the ceiling area. 98Hz is a bit more difficult, as it seems more random. Edited January 19, 2019 by Primare Knob Link to comment Share on other sites More sharing options...
almikel Posted January 19, 2019 Share Posted January 19, 2019 On 13/01/2019 at 2:26 PM, Primare Knob said: I have been trying to get rid of the nasty dip between 50Hz - 70Hz, and only moving the mic closer to the side wall seem to make an impact. I would guess that the problem here is SBIR. The same happens in the other graphs and sub and right speaker in a similar region. My biggest question is, how can I locate this SBIR? I have walked around the room along the side wall, at various distances with 200mm thick absorbtion, but never been able to make a difference. What forces are in play, and how do they work? Hi PK, can you explain what's happening in this graph a bit more? I'm not sure what you've measured On 13/01/2019 at 2:26 PM, Primare Knob said: I would guess that the problem here is SBIR not sure, but other than the green curve, the freq of the dip stays very consistent when I think you've moved the left speaker left and right? You would expect the frequency to change if the cause was SBIR On 13/01/2019 at 2:26 PM, Primare Knob said: What forces are in play, and how do they work? are you asking what SBIR is? sorry if this is sucking eggs, but it's the constructive and destructive interference from boundary reflections between the speaker and the listener - change the distances and the peaks and nulls will move in frequency, whereas with modal behaviour the frequency stays the same and only the magnitude changes with position. A member here on SNA I haven't seen for a while has an SBIR calculator on his website http://tripp.com.au/sbir.htm cheers 1 Link to comment Share on other sites More sharing options...
davewantsmoore Posted January 19, 2019 Share Posted January 19, 2019 On 13/01/2019 at 3:20 PM, Primare Knob said: might this explain why there is an uneven decay rate/time at 20Hz? No. They're totally unrelated. On 13/01/2019 at 3:20 PM, Primare Knob said: the RT60 graph doesn't seem to bad. The RT60 (is not super meaningful in a small room, but) is quite low ... and then jumps up. Most people IME would generalise that this is too low. On 13/01/2019 at 3:26 PM, Primare Knob said: I have walked around the room along the side wall, at various distances with 200mm thick absorbtion, but never been able to make a difference. Absorbtion on a wall doesn't work .... 200mm is only just sufficient to significantly affect such low frequencies, when it is placed 1/4WL for the wall (ie. where it work max). On 13/01/2019 at 3:26 PM, Primare Knob said: What forces are in play, and how do they work? The distance between the speaker and you ... and the distance between the speaker to wall to you. 4 hours ago, Primare Knob said: Any thoughts or ideas you would like to share on these measurements. Are they any useful, and what does it tell me? Why does the frequency response of the previous charts, look completely different to these??! Example: VS On 13/01/2019 at 3:26 PM, Primare Knob said: Are you moving the microphone? In short, no.... they are useless, and impossible to interpret. Put the mic in the listening position. Put the subwoofer in different places, takes sweeps, choose the best place. Apply EQ to solve problems. Finish. Make sure you either have the EQ for your speaker, ie. the EQ which flattens the driver response, etc.... either turned off, or setup correctly ... before you do any of this. If you do all of this with the subwoofer driver EQ and crossover in place, but they are wrong - then this will give you bad results. Link to comment Share on other sites More sharing options...
davewantsmoore Posted January 19, 2019 Share Posted January 19, 2019 On 13/01/2019 at 3:12 PM, Primare Knob said: Right speaker, before and after. Var smoothing has been applied to the no Absorbtion graph to make it more readable. Impossible to see anything in these as one of them is smoothed. Link to comment Share on other sites More sharing options...
almikel Posted January 19, 2019 Share Posted January 19, 2019 2 hours ago, Primare Knob said: I have mapped some pressure point in my room using specific frequencies from the decay graph of my subwoofers. 35, 49, 69, 98Hz Hi PK, can you explain how you generated that data? sounds like you chose those specific frequencies based on "problem frequencies" in your decay graph, then generated test tones at those frequencies, and used a Sound Level Meter or your mike at floor/mid/ceiling for each position? huge effort if that's what you did ~220 measurements with test tones - wow 3 hours ago, Primare Knob said: Looking at this myself I would draw the following conclusions trying to get my head around what it means...without reviewing in detail I would make the following suggestions: modal behaviour would see pressure peaks at the opposing side of the room (ie compare 1/13, 2/12, 3/11, 4/10 for the width mode) floor/ceiling modal behavior should see peaks at both floor and ceiling for the same location 3 hours ago, Primare Knob said: It is rather clear how leaky a windows is compared to a double brick wall Double brick (or any rigid construction) is a challenge for the "in room sound". It may sound an extreme solution, but installing fluffy/clip/channel/Gyprock on all rigid walls could be your best option. The fluffy/clip/channel/Gyprock will operate as a big "membrane" trap, providing boundary compliance to soak up bass. You lose some room space overall, but it will reduce your <100Hz issues substantially. Your only alternative once you've put as much absorption in as possible (and managed not to kill the top end) is large tuned pressure traps. I don't have first hand experience in building pressure traps to manage issues <100Hz in a rigid room. My leaky room doesn't have the issues you have...loads of absorption and a few bands of EQ cut <100Hz is sufficient to manage the bass in my room. ...I've considered building limp mass traps for ages as I like the concept, but then my lazy DIY gene kicks in... Rigid rooms are challenging...I'm glad I don't have one Mike Link to comment Share on other sites More sharing options...
almikel Posted January 19, 2019 Share Posted January 19, 2019 56 minutes ago, davewantsmoore said: Put the mic in the listening position. Put the subwoofer in different places, takes sweeps, choose the best place. Apply EQ to solve problems. Finish. certainly try that approach, but I can't see EQ managing the low bass issues with rigid walls - there's too much energy bouncing around 1 Link to comment Share on other sites More sharing options...
Recommended Posts