Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Sept 22, 2020 14:33:15 GMT
Great, hope it's getting better after that! . So you can be actually pleased with your room, Its a fun state to be at.
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 6, 2020 13:55:00 GMT
Hi guys, just an update. i put in 6cm of rockwool with a 5cm air gap above the fiberglass ceiling tiles. Unfortunately this has made absolutely no difference to the 8dB peak at 120HZ. I also moved all the existing rockwool panels 10cm away from the walls to make a bigger air gap...
I'm really lost now. dont know what i should do. any advise would be great
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 6, 2020 19:17:03 GMT
Is the peak exact at 120hz? That 5cm gap is not enough, you can see by playing with that porous absorber calculator. I can't know what flow resistivity that Rockwool has you used. What was the density?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 6, 2020 19:20:47 GMT
Did you have new measurements, they would be nice. Need to see the waterfall, is it ringing much anymore or not.
|
|
|
Post by Hexspa on Oct 7, 2020 8:25:57 GMT
Ok, I did an analysis. Looking back, it lines up much with mine - good work. Without going through the entire thread in detail, I can offer three things:
1. Your increase in absorption probably helped your decay times a lot. Since I haven't seen you post any waterfall graphs, I can only wonder whether you've considered that. Decay is at least as important, if not moreso, as SPL.
2. SBIR. You have two main peaks as per your initial measurement: 320Hz and 120Hz. Those relate to ~106cm (42") and 280cm (110"/ceiling) respectively. Since 106cm is uncorrelated to your room's dimensions, it must be SBIR, a flaw in the measurement or speaker/mic response. Let's assume for now that it's SBIR. If you installed the ceiling panels correctly, some other boundary must be contributing to that peak. Half of 280cm is 140cm which is 55" and a plausible SBIR dimension - floor, rear wall, side walls (either side), desk, etc. It could also be related to 560cm but that's your ceiling height doubled. The length mode is 134Hz and, judging by your SPL response, it is having an influence as well.
3. EQ. It is not abnormal to finalize a room with some subtractive EQ after SPL is maximally flattened and decay mitigated acoustically. In the end, a 6dB peak in a small room is not a failure. FWIW, I used a -8.8dB cut at 68Hz for my final measurements. Just be sure to measure in a reasonable radius so you're not exacerbating a nearby point where you might listen.
Hopefully this helps. Check out 106 and 140cm from your speakers or microphone in each direction and see whether there is a large boundary there.
dimensions: Length: 16.4 feet (5 metres) 500 Width: 11.15 feet (3.4 metres) 340 Height: 9.18 feet (2.8 metres) 280
--- Dim in cm (Peak):
L: 1x; 500cm, 2x; 1000cm, 0.5x; 250cm
W: 1x; 340cm, 2x; 680cm, 0.5x; 170cm
H: 1x; 280cm, 2x; 560cm, 0.5x; 140cm
--- Dim in Hz (Peak):
L: 1x; 68Hz, 2x; 33Hz, 0.5x; 134Hz
W: 1x; 100Hz, 2x; 50Hz, 0.5x; 200Hz
H: 1x; 120Hz, 2x; 61Hz, 0.5x; 240Hz
--- Dim in cm (Null):
L: 1/4x; 125cm, 3/4x; 375cm, 5/4x; 625cm
W: 1/4x; 85cm, 3/4x; 255cm, 5/4x; 425cm
H: 1/4x; 70cm, 3/4x; 210cm, 5/4x; 350cm
--- Dim in Hz (Null):
L: 1/4x; 277Hz, 3/4x; 93Hz, 5/4x; 56Hz
W: 1/4x; 400Hz, 3/4x; 134Hz, 5/4x; 81Hz
H: 1/4x; 494Hz, 3/4x; 164Hz, 5/4x; 98Hz
--- Bunch Peak: L: 0.5x (134Hz), H: 1x (120Hz), W: 1x (100Hz)
Bunch Null: L: 3/4x (93Hz), H: 5/4x (98Hz)
Balance: L: 0.5x (134Hz), W: 3/4x (134Hz) W: 1x (100Hz), H: 5/4x (98Hz) --- Correlation: 494Hz null (H:1/4x) 81Hz null (W: 5/4x) 61Hz peak (H: 2x)
Surprise: 320Hz peak ~106cm (42") - SBIR from ceiling, side walls either direction, floor, desk, rear wall, etc. 120Hz persistent problem 280cm (110") - 55" is plausible SBIR distance
|
|
|
Post by Hexspa on Oct 7, 2020 8:36:47 GMT
One more thing, you used 6cm/2"? I don't know what to say other than 2" is not the right thickness for 120Hz. My understanding is that 4" is the minimum for that range and 8" is ideal for the critical audio band above 63Hz - air gap or no.
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 7, 2020 12:17:35 GMT
Thanks guys. I used 38,000 pa.s/m2 flow resistivity Rockwool. On the entire ceiling there is the following: 1. 2cm fibreglass celing tiles 2. 6cm of rockwool - 38,000 pa.s/m2 flow resistivity 3. 7cm airgap
I then tried this morning to put another 6cm of rockwool above the mix position with a 1m air gap to the celing (ceiling is already treated. This made no diference to the graph. I'm going to attach the new REW graphs.
Regarding SBIR, I have moved the speakers as close to the wall as possible. This made things worse at around 68HZ. The further I move the speakers from the wall and closer together, the better the 68HZ null (but this makes the stereo image very small).
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 7, 2020 12:20:19 GMT
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 7, 2020 12:24:05 GMT
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 7, 2020 15:29:00 GMT
@hexpa, really appreciate that detailed analysis! !! Regarding the 55" SBIR, what are you suggesting i do? I dont understand the next tests i should conduct? Should i move the speakers around? higher? lower? further from the wall? what distance?
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 7, 2020 16:32:58 GMT
|
|
|
Post by Hexspa on Oct 7, 2020 17:10:38 GMT
Had to delete my previous post because I accidently compared the same graph against itself.
The bass SPL is better on the left but decay and upper SPL is smoother on the right. As far as positioning goes, just put your speakers and mic where they were for the original measurement and find whether those two dimensions I mentioned line up with any large surfaces - desk, ceiling, floor, either wall, etc. Even if you halve the distance again and both speakers are that distance from a boundary then the buildup will add to the peak.
Regardless, your decay looks good. If repositioning doesn't help then just find the ergonomically best place and use some subtractive EQ to flatten your SPL some and you should be good.
Does anyone know how to average the response to see which is flatter overall or is smoothing the best option for that? In other words, if I gain 2dB in the bass range but lose 5dB at 2kHz over an octave, which is worse? Is bass more important to have flat than upper ranges?
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 8, 2020 7:47:12 GMT
Hmm. Is there something wrong with those measurements. If not, then you have serious problems with early reflections. And that "Peak" at 120db is not real problem. That null at 66hz is bigger.
Your bass spl seems to be almost 20db lower in volume compared to mid and high frequencies. That will definitely make those focal shapes sounds like shit. Add that really nasty comb filtering from really high in level early reflections, it just cant sound good.
Your SPL starts to roll of at 1khz, how's that even possible? Im suspecting you might have something wrong with the measurements. Your firstly posted FR graph didnt look that weird.
|
|
jared
Junior Member
Posts: 57
|
Post by jared on Oct 8, 2020 19:06:09 GMT
@hexpa, 100%. Thank you! @pasim, thank you. YES - that particular microphone was messed up. I need to repost a new graph which looks normal.
Listen to this for some news..... I removed my desk out the room and what do you know? The peak at 120hz is gone! Instead of an 8dB peak, there is a small and normal 2dB peak now. I thought desk reflections were operating at a much higher frequency from my reading on the net. How do you explain this??
The rest of my problems now I think is SBIR. I'm just going to play around with speaker position and try find the sweet spot and add some EQ via sonarworks to smoothe out the edges. thanks once again for all your help! you guys are awesome!
|
|
|
Post by Hexspa on Oct 8, 2020 20:04:40 GMT
Awesome. Glad you're making progress. I want to put insulation on the top of my desk or convert it to a vinyl covered mesh.
|
|