Measurement Fun - Basement Acoustic Tx underway

[mrhudson]

Okay, after some failed attempts at getting REW going on my iMac, I shifted over to Fuzzmeasure V2 and took some initial measurements of my space-in-development. Very exciting stuff to see after reading about it a bunch. This is the "untreated" starting point ("untreated" in that there was exposed fiberglass in the ceiling that was there when I moved in). From here, I'll start testing for improvement with trapping for low frequency modes.

Any impressions from these graphs - obviously will start with bass trapping, just curious if anything else jumps out. Is the lowend rolloff typical? - running Adam A7s, no sub. Thanks!

[dykesh]

I'm updating my mixing room in the basement and have just gotten some Adam A7s for christmas. Your graph looks good as far as I know. I think you should not have any smoothing on it though. That way you know exactly what's going on. I've got a thread going as well. Look at it for comparison if that helps. thanks, dykesh

[Ethan Winer]

Originally Posted by mrhudson

Is the lowend rolloff typical?

Without knowing the size of your room I won't even try to guess if it's normal.

--Ethan

[mrhudson]

Yes, of course....duh :-)

16'3" X 23'6" X 8'5"

[DanDan]

Nice big room. However you have an approx 8 and a 16 which is bound to cause trouble.
I expect that is what you are seeing around 100Hz.
I would try the ADAMs very close to your front wall, almost touching. See if that dip improves and if the roll off frequency shifts downwards.
Speaker height is worth experimenting with too.
If you have adjustable levels in the speaker I encourage you to try a HF roll off.
Old, and now seemingly ignored, wisdom from Bruel and Kjaer and many others suggest that a good mixing curve would be up 3dB or so at 100Hz and down 3dB or so at 10kHz.
A nice even slope. My experience totally confirms this in terms of mixes translating to the real world very well.
DD

[mrhudson]

Reflection free zone now in place: Six, 4" 703 panels - two to the left, two to the right, and a two panel cloud above . Impulse response was indicating something remaining at 4-5 msec, which was the floor - cement floor with thin carpet wasn't enough. Square of 703 removed that entirely.

Next, I had some fun this afternoon placing about twelve 6" 'panels', FRK faced, around the room - I say "panels" because they are not burlap wrapped yet, but I just wanted to test for their impact, get some hint of improvement from the 703. Saw some improvement, although improvements led to some new issues. Anyway, three questions to help guide further experimentation on placement:

(1) How critical is the vertical placement of the traps? It is obviously easier to sit things on the ground for a quick test of placements, but will that "miss" much of the sound energy (with such a panel being lower than the speaker for much of its area?). Not sure that bass is so directional that it would matter much......

(2) Is there any functional difference between a wall-floor corner or a wall-ceiling corner? Wall-floor corners are so much easier for me to deal with, and I have room for them.

(3) When placing a trap on a wall-floor corner, does it matter if you place it lengthways (i.e., 48" touching the ground) or widthways (24" touching the ground)? Lengthways would be analogous to the vertical corner placements.

I'll post some more pics tonight once I've found a configuration that seems to work better. Thanks again!

(Oh yeah - curious about the theory behind lowering the higher frequencies to improve translation - what's the idea behind that?)

[DanDan]

I am assuming you are using these 6 inch traps to tame the room modes, i.e. as Bass Traps?
When you say FRK, is that a single panel with one layer of FRK or three two inchers with several layers? One layer is right.
Many of us would start by placing eight traps vertically floor to ceiling in the vertical corners. If you want FRK bounce face it out towards you. If you don't want that HF bounce face it to the wall. Bass absorption seems to be better face out.
I suggest that you find the Bass hot spots by experiment. Try a Room Mode Calculator. There is one at www.bobgolds.com. This will indicate likely frequencies for your modes.
Sweep a sine wave around each of the say three axial frequencies and their octaves.
Your room should react quite loudly and obviously when you hit the frequency exactly.
Signalsuite is good for this but the one in your DAW may be good enough.
Move about the room, stepladder to hand and find the hotspots. That's where your Bass traps need to live. It's not always corners, watch out for whole hot walls, or ceiling. If you have a strong vertical mode, you might consider a bigger, thicker cloud.
This very physical, organic, approach is pretty sure footed, for obvious reasons.

(Oh yeah - curious about the theory behind lowering the higher frequencies to improve translation - what's the idea behind that?)

Well, I just don't know. I can pick around the issue though. Bruel and Kjaer had a Testing LP with such a frequency response. They reckoned that 'modern' recordings are often done in the direct field, so this dulling of speakers would compensate. They steered a middle ground between this and a 'classical' approach, i.e. diffuse field. They reckoned their curve worked for both. I haven't been able to find the research etc. on this but I don't think they just made this up!
Take a look over at studiotips - tips on studio design, acoustics, and wiring understanding RTA. Same type of thing.
Lastly, I live in the world of making records, mixing, mastering, and I can assure you this works very well, while flat response does not. I am assuming smallish, heavily treated mix rooms here. Things would be very different in bigger rooms.
DD

[mrhudson]

Yes, I'm starting in on trying to get the bass okay. Yeah, the 6" panels consist of one panel of 2" FRK, backed up with two, 2" unfaced sheets. Of course at this point they are just 703 sheets standing against each other, just experimenting. I have a lot of the stuff, so I can afford 6" thickness across the board if it will help. I was definitely able to increase the levels at around 100 Hz from my tinkering, which was depressed originally. Will share graphs once I have tinkered some more.

You can see the room that I'm working in here (can't double-post a pic apparently):

Balancing asymmetries in possible basement studio room

I set up the listening position near the bottom of the picture, speakers firing up towards the upper wall.

The floor, left, and upper walls are cement (relative to the pic). The right wall kind of opens into an adjacent area (can one hang bass traps without a wall behind?). Ceiling is fiberglass stuffed between joists everywhere. The bottom wall is just studs with 4" of fiberglass, which was covered with burlap. The basement was "sort of" finished when we bought the house. So, not a perfect room for sure, but its what I have to work with. At least I have all my music making stuff actually in one room now, and that alone is happy times.

For bass traps, to start, I can do floor to ceiling and straddle the corners on that upper wall, even though they aren't 90 degree corners. Hopefully that's enough of a corner to get the "corner enhancement" effect. As you can see, I have one proper corner in the upper left hand side. I already put a 34" superchunk in that corner, floor to ceiling. Wham that one!

[DanDan]

Well that's quite a confusing space, that happens a lot though.
This will be a bit stream of consciousness but here goes.
Six inch is great.
Consider where your speakers are carefully. FM or REW are of great help here. A few quick tests should give strong indications of the best location. Go with the waterfalls, not frequency response.
I like the idea of a high ceiling any day. The stuffed joists you have are a great cloud, love it.
An open area, even a doorway, may function like a bass trap, e.g. if not much returns back through it. There is however absolutely no harm in hanging a bass trap there, perhaps for symmetry.
Bass is not fussy about exact angles. Those wavelengths are metres long. Playing sine waves at the modal frequencies will generate very hot spots which can be readily identified by ear. Good spots to place your Bass traps!
Best, DD

[mrhudson]

Can anybody read German? :-)

http://www.adam-audio.com/files/downloads/A7_Sound-and-Recording_german.pdf

I was trying to see what kind of frequency response I should realistically expect from my ADAM A7s below, say, 80hz. The first few graphs in the link seem to indicate some substantial rolloff (like what I am seeing), but without reading German hard to ensure that I have the context correct.

But, if that rolloff is real, and speaker related, certainly not worth pulling my hair out trying to chase it down.

[mrhudson]

Well, I've got an RFZ in place (two, 4" thick panels for a cloud, and two 4" panels on left and right - 6, 4" total). The back wall has a superchunk in my most well-defined corner, four floor to ceiling bass traps, even more bass traps on the back wall - so maybe 16, counting roughly. I have three more stacks of 6" 703 in other locations. Oh yeah, all traps 6" thick, most with FRK front faces. I probably have the equivalent of 22 bass traps, all but 6 of them being 6" thick. Impulse response looked pretty clear, with the various spikes being removed (w/in 30 msec). Some pesky spikes found along the lines of what SAC had suggested (use speed of sound, receipt delays, etc., to find probable distances for reflections)

So, I was getting some improvement, as tracked via my left speaker (see left speaker freq response - measurement 17 is after treatment)


And then I checked my right speaker (see Right Freq Chart - Measurement 7 is AFTER treatment). Ack! It is as if no traps were added at all!

Is this result a big tipoff for something - the non-effect, and the left vs right speaker discrepancies? Any suggestions would be a huge help - I tried Ethan's pink noise (felt like the ghost hunters, searching for hot/cold spots!)

I think that the ADAM A7s might have some roll off in the low end, so I'm not totally surprised to see the low end 10dB down at 50Hz I suppose, but the right speaker has this 80Hz/120Hz swing going on......Is that a clear clue?

[PaulP]

Your room modes aren't great (your ratios are ~1:2:3) :



You've got three of them coming together somewhere around 70hz and multiples thereof.

I think you'll have to either do something very specific to address this or else move a wall
to get better ratios.

Paul P

[mrhudson]

I experimented with speaker positioning, trying to improve the frequency response of the right speaker. Left speaker was best where it is, but the right speaker was better pushed back some. However, doing so means that I'm no longer firing sound straight down the room. Can this work, if, say, you position panels to create frontal symmetry?

Any tips for how to find the best trap placement to kill the 80/120Hz stuff? I have enough 703 left to make twelve, 6" thick traps - surely that ought to be able to help if placed correctly? I'll just play some 80 Hz tonight and try to find where it lives :-)

[DanDan]

I'll just play some 80 Hz tonight and try to find where it lives :-)

I reckon that is the best shot. I have gotten much very clear positions and many surprises by triggering the actual modes. The hotspots get very loud and somethimes they are not where you might expect !
DD

[mrhudson]

Will do, thanks again. I figure that there must be something to the fact that the right speaker is having the bigger issue with that 80 Hz stuff. Maybe something is happening on the right side of the room, will start there. Back to bass chasing.......

[DanDan]

Swap the speakers to be sure there is no fault.
Is one side of the room concrete the other sheetrock? This would cause a bass response inbalance. Concrete reflects very firmly and thus supports modes. Sheetrock absorbs bass. Thump it and you will hear the frequency area of resonance/absorption.
DD

[mrhudson]

For the right speaker, I was able to knock off about 3 dB from the 80Hz peak, and bring up the 100-150 dip, with a TON of 703 in my front left "corner" (bottom right corner in referenced pic/thread). So, some reasonable improvement.

Of course, this treatment then opened up on the left speaker a -10dB gash at 175 Hz or so!! A new carnival game - "Whack-A-Bass" - hammer down 80Hz, another one pops up! ha.....

Locating the 175 Hz dip was interesting. Moving from left to right in the room, facing monitors: Left side was a dip, then high as I moved right towards the left speaker, then back to null-ish in between speakers, high again in front of and to the right of the right speaker, then nullish at the right wall. I got some improvement by putting about 6" of 703 on the floor in a hot spot, but that's hardly a great place for a permanent trap! Is this when a ceiling trap might be called for - when the hot spots aren't near a wall? Gapped floor traps? :-) Harder to test ceiling traps.

Just for my own sanity, what's a reasonable end point, in terms of +/- variation? If I average the right and left speaker signals (as presumably one's brain does while sitting and listening?), I have a bump between 60 and 100 Hz, peaking about 4dB up at 80Hz. Maybe 4dB variation from 100 to 200. High end has bumps around 6-8 Hz, and a surge heading up to 20k (might be character of the adams). But, I can roll that down a bit via the speaker dials.

The 10 dB rolloff down to 50Hz that I'm seeing might be speaker related. Levels at 60 dB are more or less in line with the average low end level, but there is about 10dB of rolloff down to 50Hz. One test: If I push up the <150Hz dial on the back of the ADAMs by the full amount (+6dB), there is a dB for dB increase at 50Hz. I'm assuming that this means that nulls aren't causing the rolloff down to 50Hz - if they were, wouldn't it be tougher to get the lowend dial to increase the bass? That, plus the measured frequency chart referenced above for the ADAMs, which shows the same rolloff........

[mrhudson]

Did a bunch more work on repositioning the speakers, trying places very far flung from the 38% location. I found a position where a 37" listening triangle seems to have the most even bass. I've attached four pics showing 40Hz to 250 Hz, -30 dB to 0dB.

The first two files show the left and right speaker separately, unsmoothed and then smoothed.

The left speaker has a range of +/- 4dB from 42 through 250, with most falling within 2 or 3 dB. The right speaker has a similar range but with a bit more rolloff around 60.
With smoothing, its within +/- 3dB from about 45/55Hz through 250 dB.

I figured that the left and right speaker combine at the listening position, so the next two files average the response from the left and right speakers. This average looks to be within +/- 2 or 3 dB down to 45 Hz. Some "offsetting errors" from the L and R sides. The smoothed version of the L/R average is within +/-1 or 2 dB from 50 through 250.

Of course, my desk won't fit in between the speakers - but I can always work around that. This placement was pretty precise, unfortunately. Didn't have a ton of wiggle room outside of this sweet spot.

So, is this analysis valid? Looking much better I hope!

[PaulP]

Originally Posted by mrhudson

Did a bunch more work on repositioning the speakers, trying places very far flung from the 38% location. I found a position where a 37" listening triangle seems to have the most even bass. I've attached four pics showing 40Hz to 250 Hz, -30 dB to 0dB.

Your monitors are pretty close together, what's the soundstage like ? It
seems to me that what you've done is taken a lot of the room out of the
equation, like headphones but to a lesser extent.

Just out of curiosity, what does the room sound like ? (Before you moved
your monitors so close together that is).

Paul P

[mrhudson]

I thought I read somewhere that 3-5 feet was a reasonable spread for monitors in a home studio, so I am definitely on the low side. I haven't done music listening in that spot yet, save for my seemingly one millionth sine wave sequence :-) Didn't have a chance yet: I'll listen today and see. (What would I be listening for as goes soundstage, depth/breadth?) Other positions, with that 3 foot spread, didn't have the same advantage, so I'm not sure that the narrower positioning alone caused it, though it could have - I certainly didn't test all combinations, though it feels like it :-)

Based on days of positioning and testing, I seem to have two basic options, without more drastic interventions:

(1) New position - 3 foot triangle with pretty even bass. That's a big plus (?), as I thought even bass was the hardest thing to attain. Cons - slightly smaller triangle (I was using 4 foot usually) (need to rethink my little desk), and there is a slight front/right asymmetry that I'd have to address (?) via symmetric placement of hanging RFZ panels.

(2) Various other positions that I tried yesterday allowed for a slightly wider triangle and somewhat more symmetric room position (although the left side still opens to another room, using hanging 703 panels), but they had worse bass response. The bass rolled off quite a bit from 60 down to 40 Hz. From 60 Hz up, I saw +/- 6-8 dB fluctuations - level at 60Hz, swells up to +8 dB to 80, down to even @ 100, 6-8 dB down @ 120 - stuff like that. Maybe there was a +/-5 or 6 dB type spot. As for how it sounds, more bass shy in the measured bass shy positions - but tight and clean imaging. I could enjoy listening in either position, sounds way better than where I was before (totally untreated bedroom, so anything is better!), but I suppose I'm trusting the measurement for the low end.

What should one aim for in terms of +/- dB fluctuations? Which of these two options would you base things around? Compromises :-)


(This is where I found the 3 foot recommendation for the ADAMs - http://www.adam-audio.com/files/down...06_english.pdf )

[Ethan Winer]

Originally Posted by mrhudson

What should one aim for in terms of +/- dB fluctuations?

If you can get the response within a 10 dB total window you're doing really well. But this means when viewing the response at high resolution, not with 1/3 octave averaging.

--Ethan

[mrhudson]

Great, thanks - 10 dB total window, meaning up 5 and down 5. Got it.
I think I have that in my current position (at least down to 45Hz, Repositioned Part Two above has no smoothing), unless the 3 foot triangle on the monitors will mess up soundstage/panning......

[DanDan]

That response looks good to me. The software was really useful in searching out the best positions eh? If the 3 foot thing sounds good to you I wouldn't sweat it. Just remember to listen to reference great mixes (mine ) frequently to calibrate yourself.
I use headphones frequently when mixing. No room acoustic. A very useful tool, I have used the same headphones for decades.
I would suggest that you take a look at your Waterfall plots before and after treatment.
You may get a very pleasant and reassuring surprise. The decay spectrum is at least as important as static frequency response.
Well done, DD

[mrhudson]

Is this what you might expect for a waterfall plot? This is left speaker only for now, I still need to get a y cable to send the FM2 output to both speakers. Should it look more even with respect to time?