|
Post by Hexspa on May 12, 2016 2:05:10 GMT
|
|
|
Post by Hexspa on May 13, 2016 12:14:25 GMT
EDIT: The following is a condensation of my two main questions:
1. The front wells are equally wide and 16" deep. Should I make a 48" superchunk or fill them flat?
2. If I make a rear wall (and I do intend to do so) would it be sufficient to make 2'x4'x6" OSB/Plywood-faced panels filled with R-19 and stack them as a "modular" boundary?
I ask because building a wall that fails is going to make it harder to reuse and repurpose those building materials that could've made tried-and-true traps.
Or can I leave these modular frames "unfaced" and use them as a Stack of Panels/Absorptive rear wall and still yield good modes?
Thanks a bunch,
-m
|
|
|
Post by Ethan Winer on May 13, 2016 19:42:49 GMT
Can you post a drawing showing a top view of the room?
|
|
|
Post by Hexspa on May 14, 2016 0:17:59 GMT
Yes, of course. This is a layout of the living room and kitchen area. The door on the right is outside, the door on the left is to the bath and bed areas. The divider between the kitchen and living room is a 43" bar in front of a 36" cabinet/countertop. Above the bar is a 22" gap and then cabinets attached to the ceiling as per the pictures. The square there is a standing built-in cabinet. Presently my listening position is oriented 38% from the depth of the front wells to the flat rear. Thanks, -m
|
|
|
Post by Hexspa on May 14, 2016 4:43:07 GMT
After much stress I will now produce some fuzzy measurements. These were taken with the Radio Shack SPL (same as on Ethan's site) one round microphone base up from my present listening position. Basically if I lean my head forward to "listen closely" from where I am I'd be roundabouts here. The response looks promising to me. It is, of course, with my provisionally placed acoustic treatment affecting the sound. After 15 snapshots within a meter radius, the overall response remained consistent with it's characteristic null at 55-59Hz, the 70Hz peak, a smaller 76Hz null, the peak around 89Hz, another peak at 115Hz and subsequent "waviness" after 150Hz. At some other positions, nulls were emphasized at 180Hz, 210Hz, 248Hz, 275Hz ahd 291Hz. So the 55Hz null.. A quarter wavelength is just around 5' or so. About 25' is the length to my rear wall (23.83') and my front wall is stepped. Does that mean the 5/4th wavelength is causing that?
|
|
|
Post by Hexspa on May 14, 2016 4:49:14 GMT
Looking at the AMROC calculator here I can see 55Hz is unsupported with my rear wall. amroc.andymel.eu/?l=23.83&w=13.33&h=8&ft=true&r60=0.6Now I can understand that line graph superimposed on the keyboard. If I was to make my hypothetical rear wall I'd be looking at something more like nya: amroc.andymel.eu/?l=18.32&w=13.33&h=8&ft=true&r60=0.6So really my question is how to best proceed. I can build a rear wall made from 2x6 lumber with R-19 inside and faced with OSB for about $80. That would yield eight modular boxes which I would stack to create a barrier. Would this be reasonably effective on modal distribution? Given that I make some sort of rear-wall-shortener, how would you guys treat those front wells? It almost seems less important than getting my dimentions improved. Of course with some super chunks and more intelligently placed panels I'd probably get within the +- spec but, given the challenge and affordability of materials, I'm down to erect a surface. Sorry I couldn't give more graphs; I'm currently using a demo of FuzzMeasure and probably can't update due to my old OS. Thanks! -m
|
|
|
Post by Hexspa on May 14, 2016 5:53:46 GMT
Pretty much my plan at this point is as such:
1. Use what I got. Refabric my current panels, make them all 4" in thickness, reposition my desk so my listening position is at the above coordinate and install as many panels with as few drywall anchors as possible.
2. Superchunk. I really wanted to do something special for this new space so this is a pet project beyond improved response.
3. Wall. This is a last resort and dubious anyway. I can understand that shortening my rear boundary can potentially attribute to improved modal response. However, building a proper rear wall will be the most expensive of the three steps and the least transferrable to a new location. At best I will be building a barrier that isn't structure-grade (i.e. it will be modular, consisting of stacked 2'x4' frames stuffed with R19 and will not completely block the sound's view of the rear wall). Despite this I am considering it depending on how effective the first two steps have been.
Again, any input or feedback is appreciated especially regarding the response, listening position and, most of all, the rear and front walls.
Thanks,
-m
|
|
|
Post by Hexspa on May 14, 2016 7:10:42 GMT
This might sound crazy but questions like this happen when you learn new things:
Would it be helpful, in your experience, to place R19 in my kitchen cabinets along the tangential and oblique modes running from the front right to rear left corners? If so would it be better to keep the cabinet doors open or closed?
I'm not fully understanding which frequencies, and to what extent, pass through which materials and how much effect insulation contained within them will help.
Thanks,
-m
|
|
|
Post by Hexspa on May 14, 2016 12:15:43 GMT
Just realized my null at 55Hz is probably because my speakers are almost exactly 5' from the front and side walls. Double axial null action. I have a lot of moving stuff and buying cables to do.
|
|
|
Post by Rock on May 14, 2016 14:10:16 GMT
My suggestion of building a wall was conditional on "If you wanted a closet or iso booth". If you're building the wall (or modules as you describe) and it's only temporary and you have no practical use for the extra room you'll create, you'll be better off dealing with the existing room modes with more absorption, which is portable and reconfigurable in your next space. For new or permanent construction, it makes sense to consider dimensions for best room modes but even then, it's still only part of the equation. In your case, as you have discovered, it's the most expensive part of the project so you should probably skip it. Sorry, for leading you down this path but I learned something too.
Cheers, Rock
|
|
|
Post by Hexspa on May 14, 2016 14:32:51 GMT
Sorry, for leading you down this path but I learned something too. Cheers, Rock Rabbit Chaser is my middle name man. np Thanks, -m
|
|
|
Post by Hexspa on May 14, 2016 15:00:52 GMT
I moved my speakers (as per this article) into the corner to try and counteract that 55Hz null - and it worked! Unfortunately the overall response is now a bit worse; I'm probably now seeing my room's true colors. Thing is though, the soundstage is huge. Like Tim says in the article, and I'm guessing it's true, is that by placing my speakers in the corners I'm exciting all of the room modes. Subjectively the signal does sound much more like my room. I've never had my speakers so far apart and it's pretty cool. I assume that by fully treating my corners and reflection points that the response should even out. I'd love to make this work because I love how it sounds. It reveals like never before how bad my old mixes are lol/cry. The null before 400Hz is probably due to those wells as they're just over 3' away. Superchunks to the rescue. I should probably sleep but the unholy trifecta of internet, calculators and tape measures won't let me. -m
|
|
|
Post by Ethan Winer on May 14, 2016 16:13:35 GMT
I'll guess 4-foot wide chunks in the front will be best, but I guess you could measure / calculate. As I'm sure you know, bass trapping is all about surface coverage and depth.
--Ethan
|
|
|
Post by Hexspa on May 15, 2016 1:13:47 GMT
I'll guess 4-foot wide chunks in the front will be best, but I guess you could measure / calculate. As I'm sure you know, bass trapping is all about surface coverage and depth. --Ethan Definitely - can't wait. Apache! -m
|
|
|
Post by Hexspa on May 15, 2016 9:34:33 GMT
Five feet is that 55Hz null and the dimension suggested by the 1:1.67:2.33 ratio. Depending how many panels I have left I'll try to make my "wall" out of doubled up 4" mineral wool and test.
|
|