|
Post by zyodmusic on Dec 21, 2019 20:09:15 GMT
Hi, first post here.. I'm in the process of building a home studio in my basement. (5,06m x 4,20m, 2,53m) I started treating around the first reflections with Ikea bookshelves filled up with rockwool on the side walls (28 cm thick), a cloud above the listening position (20 cm thick) and two absorbers in the corner of the front wall. (see attached) Today i ran the first measurement with Room EQ Wizard and got the following results. All i got left is treating the back wall and im not sure if i should only make absorbers in the corners, the whole back wall and some diffusers or nothing at all? Thank you!
|
|
|
Post by rock on Dec 22, 2019 14:54:28 GMT
Very good, It looks like you're on the right track. Looks like you've still got peaks and nulls, mostly probably from room modes so yes, bass traps in the back will help, covering the entire wall is also good. Diffusion AFAIK won't help with modes but can smooth mids to hi mids and help with localization. You can put diffusors in front of your entirely absorber treated rear wall but you don't need to go floor to ceiling with the diffusors. BTW, Ethan's company Real Traps makes a combination diffusor/absorber. Remember that you have 12 corners when you include wall/ceiling and wall/floor corners so those are areas for bass traps too. And then there's tweaking of speaker placement and listening position and you can experiment with.
|
|
|
Post by Hexspa on Dec 23, 2019 3:47:09 GMT
Ok, I did some numbers. If you're listening and measuring at 3/4 the height of your room then you need to move toward the middle. Also balance your front-back treatment with that of your ceiling because the two dimensions should even each other out around 72Hz but they're not. You have double nullage from your width and height just above 100Hz so try to shuffle your position to see if you can find the front-back peak at 114Hz.
In general, I think you can do better with your listening position. Try moving it more toward the 38% baseline because, as you've drawn it, it looks like you're in the middle of the room and it's messing up your response. Positioning matters a lot for SPL. Adding treatment progressively removes the dimension and mode's impact.
Here are my notes: 5.06m - 16.6' - 506cm - 57Hz, 114Hz, 28.5Hz peak (at, double, half wavelength) 4.2m - 13.8' - 420cm - 82Hz, 164Hz, 41Hz, 2.4m - 7.75' - 235cm - 147Hz, 294Hz, *73.5Hz 57(0.25) = 14.25, 47.75, *71.25Hz null (1/4, 3/4, 5/4 wavelength) 82(0.25) = 20.5, 61.5, ^102.5 147(0.25) = 36.75, ^*110.25, 183.75
*opportunity to balance ^reinforced null
big null 110Hz - possibly aggravated by vertical measurement height (3/4 wavelength) 71.25 null - length - treat rear wall null at 71.25Hz from 5/4 wavelength of room's length mode should be balanced by peak at ceiling's half-wavelength mode 73Hz. adjust listening height away from 3/4 wavelength balance ceiling treatment with front-back wall treatment
|
|
|
Post by zyodmusic on Dec 23, 2019 14:56:12 GMT
Very good, It looks like you're on the right track. Looks like you've still got peaks and nulls, mostly probably from room modes so yes, bass traps in the back will help, covering the entire wall is also good. Diffusion AFAIK won't help with modes but can smooth mids to hi mids and help with localization. You can put diffusors in front of your entirely absorber treated rear wall but you don't need to go floor to ceiling with the diffusors. BTW, Ethan's company Real Traps makes a combination diffusor/absorber. Remember that you have 12 corners when you include wall/ceiling and wall/floor corners so those are areas for bass traps too. And then there's tweaking of speaker placement and listening position and you can experiment with. Thanks for your help! Definitely will try experimenting with the listening position.
|
|
|
Post by zyodmusic on Dec 23, 2019 15:23:57 GMT
Ok, I did some numbers. If you're listening and measuring at 3/4 the height of your room then you need to move toward the middle. Also balance your front-back treatment with that of your ceiling because the two dimensions should even each other out around 72Hz but they're not. You have double nullage from your width and height just above 100Hz so try to shuffle your position to see if you can find the front-back peak at 114Hz.
In general, I think you can do better with your listening position. Try moving it more toward the 38% baseline because, as you've drawn it, it looks like you're in the middle of the room and it's messing up your response. Positioning matters a lot for SPL. Adding treatment progressively removes the dimension and mode's impact.
Here are my notes: 5.06m - 16.6' - 506cm - 57Hz, 114Hz, 28.5Hz peak (at, double, half wavelength) 4.2m - 13.8' - 420cm - 82Hz, 164Hz, 41Hz, 2.4m - 7.75' - 235cm - 147Hz, 294Hz, *73.5Hz 57(0.25) = 14.25, 47.75, *71.25Hz null (1/4, 3/4, 5/4 wavelength) 82(0.25) = 20.5, 61.5, ^102.5 147(0.25) = 36.75, ^*110.25, 183.75
*opportunity to balance ^reinforced null
big null 110Hz - possibly aggravated by vertical measurement height (3/4 wavelength) 71.25 null - length - treat rear wall null at 71.25Hz from 5/4 wavelength of room's length mode should be balanced by peak at ceiling's half-wavelength mode 73Hz. adjust listening height away from 3/4 wavelength balance ceiling treatment with front-back wall treatment
Wow, appreciate your help! These measurements are actually made with the listening position being 38% of the baseline. (unfortunate drawing sorry :/) I'm quite new to the the topic, so I´m trying interpreting these deductions: 1. experimenting with listening position 2. moving the speakers up in height 3. treat rear wall - behind the speakers i guess is best? 4. treating the back wall - would you suggest filling out the whole or just the corners? Thanks again!
|
|
|
Post by Hexspa on Dec 23, 2019 18:26:36 GMT
The thing is to experiment. You have a massive cloud which is mainly negating your floor-ceiling mode. However, I like to look at the room's natural tendencies - found with a baseline measurement in an empty room. This way you can find the ideal location, add essential treatment like RFZ then progressively treat the bass in harmony with the room.
So, barring starting over, just take that into consideration and maybe it'll help. I can't say 'do X and get Y' because there are always surprises - again, it's a balancing act.
|
|
|
Post by rock on Dec 23, 2019 23:30:44 GMT
Hexspa, nice analysis. You have given Zyod has some choices but I'll just simply say that if the front to back mode seems to need some attention, for that, rear corners are good but full rear wall is better. Hexspa's got way more hands-on experience than I do but I believe Ethan has mentioned that absorption behind the speakers is not as directly effective as the rear wall. This is not to say the front wall can't have absorption it's just that the rear wall might be a better choice if it's one or the other. This is were experimenting comes in. My 2 cents.
|
|
|
Post by Hexspa on Dec 24, 2019 4:22:49 GMT
Rock just reminded me to look closer at the waterfall. Zydoc, I aim for 20dB decay within 150ms above 63Hz so try seeing how more treatment affects that. I'm having a little trouble reading the waterfall so if you want to put a decay plot showing around 150ms, that'd be great.
|
|
|
Post by zyodmusic on Dec 24, 2019 12:47:58 GMT
The thing is to experiment. You have a massive cloud which is mainly negating your floor-ceiling mode. However, I like to look at the room's natural tendencies - found with a baseline measurement in an empty room. This way you can find the ideal location, add essential treatment like RFZ then progressively treat the bass in harmony with the room. So, barring starting over, just take that into consideration and maybe it'll help. I can't say 'do X and get Y' because there are always surprises - again, it's a balancing act. Yeah, not sure about the cloud either... what do mean exactly by massive? to thick or to big in size?
|
|
|
Post by zyodmusic on Dec 24, 2019 12:49:32 GMT
Rock just reminded me to look closer at the waterfall. Zydoc, I aim for 20dB decay within 150ms above 63Hz so try seeing how more treatment affects that. I'm having a little trouble reading the waterfall so if you want to put a decay plot showing around 150ms, that'd be great. Sure. You mean this one? Attachments:
|
|
|
Post by Hexspa on Dec 24, 2019 19:10:33 GMT
Yes but please only show those closest to 150ms. As you have it, I don't know what any of those slices mean. What I mean by the cloud is that it seems to be too much for the space even if you were to treat the rear wall. Not sure how you built it but you may benefit from breaking a piece off and sticking that on your rear wall. What's easiest for you: adding treatment or rearranging what you have?
|
|
|
Post by zyodmusic on Dec 25, 2019 9:49:35 GMT
The cloud is just a wooden frame (box) 2m x 2,40m filled up with rockwool (20cm thick) and mounted with a 10 cm air gap. I could make it smaller in size or thinner if thats causing the null. I´ve still have plenty of rockwool left so adding treatment or rearranging is an option (or both). Here´s the decay graph. Its 140ms, cause thats the closest REW shows. Attachments:
|
|
|
Post by Hexspa on Dec 25, 2019 19:08:53 GMT
Ok. Well, I know REW doesn't show 150ms which is why I said "those closest." That would be 140ms and 160ms IN ADDITION to the original SPL level - makes it easier to compare when it's on one graph, you know? Anyway, you're getting about 10dB of decay at your peaks which means you don't have enough absorption. If you didn't take an 'empty room' baseline measurement then, if you're serious, I'd empty the room out and start over. Try to understand how the the dimensions of your room support certain frequencies and not others. Position your listening spot and speakers so you get the flattest SPL, treat RFZ to the minimum then measure again to see where your modal treatment needs to go. You can always do the 'shotgun' approach and just stick panels everywhere but that's not the most accurate way.
After LP and RFZ and baseline are done, you can see which dimensions balance one other out, i.e. 'width has a peak at X frequency and length has a null there so maybe I won't treat those as much, etc., and start experimenting with placement. Again, it's about balance so don't drill any holes until you've measured each hypothetical array. Also try to appreciate that absorption is mainly about decay although it will help with evening out SPL as well.
|
|
|
Post by zyodmusic on Dec 25, 2019 19:59:47 GMT
Sorry, completely read over it... Here´s the right graph. I actually did measurements when the room was empty. I´ll attach them aswell. Do you have any good sources where to the whole room mode topic is explained well? Or how do I know exactly in which area to put e.g. the cloud?
|
|
|
Post by Hexspa on Dec 26, 2019 23:56:06 GMT
I looked but can't easily find any succinct resources on modes. Let me simplify what I know.
Shaped materials and space within hard boundaries have resonant frequencies. The AMROC explaination was that of a pendulum in that there's a rate at which it moves most easily. Think of a kid on a swing.
Rooms fit these criteria hence modes. The dimensions of your room correlate to frequency wavelengths. Those frequencies are then either supported, reinforced, unsupported or cancelled out. This is why you have 'golden ratio' rooms where the boundaries evenly support the bass range.
Dimensions which are 1x, 1/2x and 2x a given frequency's wavelength will reinforce it. Ringing and an SPL peak may, and often do, result. The converse is that dimensions which correlate to a frequency's 1/4, 3/4 and 5/4 wavelength will create a null. I believe that the dimensions closest to the actual frequency will have the most impact, i.e. 1x peak and 3/4 and 5/4 nulls rather than 1/2x and 1/4x. In addition, some ranges are neither supported nor nulled but they aren't generally a problem so long as all your support (anti-nodes) and nulls (nodes) aren't bunched up due to identical or 2x multiple dimensions like an 8' ceiling with 16' length.
Therefore, all you need to do to predict modal behavior is to use a frequency chart which gives wavelengths then line it all up like I did in my previous post. Naturally, sound moves in three dimensions so these simplified modes between two boundaries, called axial modes, are just the beginning. Even so, axial modes are the strongest among tangential and oblique modes and easiest to understand so we just focus on those.
I don't see a lot of people in forums going this deep. Indeed, targeting modes and balancing treatment based on a room's dimensions is only something I've done myself out of being perplexed by the response of my current oddly-shaped room. In other words, other people might do this - particularly paid professionals - but I just have never seen anyone talk about it. This is the only room I've done it in plus I'm a math sweater rather than a tournament winner so I'm not 100% convinced it's valid.
Even so, it's worked once and makes sense so I tell people about it with the caveat that, like I mentioned, room acoustics are complex beyond A+B=C - my mathematical pay grade - so the principles of experiment and measure still apply like always.
No matter how you slice it, I think you need more absorption in your room. If you aren't ready to tackle a possibly experimental method of modal balancing then just do the shotgun approach and stuff absorption everywhere. Eventually something will work.
|
|