Small control room design mk2

[endorka]

Source of the post That’s what I’d been wondering - are the dips around 60, 100 and 150 causing me to accentuate those frequencies? This is why I’m considering the proposal to build tuned absorbers at these frequencies.

These dips certainly seem the most likely culprit to my mind. But if they are caused by SBIR, I don't think a tuned absorber would directly address them.

In fact, I thinking about it there was one track in particular and I bet I have the version 1 mix saved. I should analyse it to see which frequencies are pronounced in case it correlates somehow.

Well worth checking out.

I suppose floor bounce is possible - I do have some more of those GIK absorbers in the loft so maybe I could drop them on the floor and see what happens.

Definitely try this. I have done similar and it does make a difference. Look at the impulse chart to see if it makes a difference to any of the peaks, particularly one at the time corresponding to the additional distance imposed by the floor bounce. Remember to do this for one speaker at a time; impulse charts don't really apply in this case with both speakers engaged.

Going back to the soffits, there is evidence that things change in the low mid, with a couple of dips between 400 and 600, however before soffits there was quite a peak there so one bad thing has been replaced with another. I do think that on balance they’ve had a positive effect, particularly on the waterfall plot.

I think it is the waterfall chart that matters most in this assessment as it shows the resonances through time, and that is what the acousticians you mentioned said a problem could potentially be. This is what a resonating structure would do. The SPL chart is only a snapshot and prone to interference from other factors.

As you are using full range speakers in the soffits there may be limits in how much you can address the floor bounce SBIR. I believe some folks high pass the soffit speakers above the nasty SBIR frequencies and use a sub woofer on the floor to replace them. Since it's on the floor there is no floor bounce.

Cheers,
Jennifer

[Starlight]

Source of the post... however before soffits there was quite a peak there so one bad thing has been replaced with another.

It might be that the previous peak that was a problem was strong enough to mask a lesser problem such that now that the previous peak has been dealt with the lesser problem is noticeable. Possibly but not definitely.

I am enjoying following the analysis you and Jennifer are doing. I am sorry that I am not yet knowledgeable enough to be of much help at this stage.

[garethmetcalf]

Interesting thoughts there from both of you.

I note your observations about floor bounce Jennifer, but the fact that these frequencies (62, 100ish and 150ish) are axial modes it does seem likely that they are not SBIR.

As I have some spare absorbers and a (currently unused) subwoofer, it seems the next thing to do is carry out some experiments with absorption on the floor behind the desk, and then high passing the speakers from the sub as well to see what the impact is/isn’t.

Assuming these nulls in response don’t change then I could try building tuned traps. My worry here is that they will have to replace some/all of the hangers and therefore what else might be compromised? Only the effort, expense and time will tell I suppose. Possibly worth it for science if nothing else!

Gareth

[endorka]

Source of the post I note your observations about floor bounce Jennifer, but the fact that these frequencies (62, 100ish and 150ish) are axial modes it does seem likely that they are not SBIR.

Good point. They could be both though, fortunately with your spare panels and the string trick it is easy, cheap and quick to rule out the SBIR possibilities

Assuming these nulls in response don’t change then I could try building tuned traps. My worry here is that they will have to replace some/all of the hangers and therefore what else might be compromised? Only the effort, expense and time will tell I suppose. Possibly worth it for science if nothing else!

Absolutely. In all seriousness, I find being actively involved in a field where new observations, understanding and discoveries are still being made concerning theory vs. practice in the real world utterly compelling. The acoustic hangers, although known to work, are not a well explained phenomena, for example.

Cheers!
Jennifer

[garethmetcalf]

I have completed a few experiments today, with the following results.

1. I took a new REW reading of the room, and then added my spare GIK traps to the floor between monitors and desk. Although these are only 50mm thick, I figured they would do something about the floor bounce if that is the issue.

2. I plugged in my unused Adam T10S sub in place of my monitors and ran an REW reading, to see if the sub would make any improvement with it being on the floor (ie no floor bounce), and whether it had any further low end extension than my monitors. I didn't take any care of where the sub was placed (basically behind the desk off centre) and made sure the crossover in the sub was bypassed.

3. I did a 'match EQ' in Logic between the two versions of the mix that I know I overcooked (over 8 bars only), which showed the difference in EQ curve between the two mixes and compared that with the most recent REW SPL reading.

Results:

1. I couldn't find a good colour combination to show the differences here, but in essence introducing the absorbers on to the floor did nothing for the low frequencies, and instead changed the mid band so that some frequencies ended up as nulls and others as peaks. The very sharp null at 650Hz is odd for sure.

Conclusion - no improvement.


2. This is interesting, I think. The sub performs worse than my soffit mounted Neumann monitors below 25Hz, performance between 25Hz and 75Hz is similar, and then things change. The sub produced a null around 92Hz, however from 100Hz-150Hz the response is much flatter with the sub rather than the monitors.

Conclusion - I don't know what to make of this. I would love to get the response the sub provides between 100Hz and 150Hz from the monitors.


3. Even more interesting. I don't like the sound of the mix I ended up with - even on headphones I find it a bit lacking in the bottom end, but look how well the EQ curve correlates to the dip in SPL response of my room between 50Hz and 100Hz. Basically what this shows is that in version 1 of my mix I had turned up the 50-100 space too much, which sounded too loud on other systems, but in my room sounded right because the response of the room is quieter in that area. However, I'm not sure how scientific this is, and may be related to the sound sources in this track as much as anything else. And, more importantly, other tracks have worked out fine. That said, with an album I was mixing in July we noted we were overboosting around 160-170Hz, when playing the tracks in my lounge.

Conclusion - I do need to take care around those all important 50Hz to 100Hz frequencies - I don't know what would be the right move for trying to even out that part of the SPL response. Maybe I will give the tuned traps at 62Hz and 100Hz a go.

[gullfo]

in this regard, placement of the sub matters. one way is to put the sub on your chair in the listening position, then run short sweeps between 25-100hz and get down on the floor and listen to where it generally sounds best for as much of the sweep as possible. put your sub on that spot. next, the sub should augment. if your main speakers roll off around 80hz, then set the sub to augment 25hz-90hz - you may have to adjust timing (phase) and polarity (180 phase flip) to get the right levels and balance. just run short sweeps for the speakers from say 40hz-150hz and monitor the levels, tweak your crossover box for timing and polarity until it evens out as much as possible. you do have a crossover unit? right?

[garethmetcalf]

Thanks Glenn. At this stage the sub isn’t part of my system, and as such I don’t have a crossover other than the built in one.

I tried the sub only to see if it made the room perform any differently, to help understand the next steps for treatment. I thought it interesting that the sun doesn’t actually perform any better in the very low frequencies than the monitors, suggesting that if I do use a sub I would possibly need a different one.

Should I integrate a sub I will definitely take your advice and play with crossover and phase settings etc.

Cheers
Gareth

[endorka]

Fascinating! Would you be able to post the REW file? Perhaps I could find some other areas for you to rule out

I've been at this point too by the way, where I just can't quite figure out what is causing what. Frustrating.

Cheers!
Jennifer

[garethmetcalf]

Hi Jennifer

Please do! The sharp nulls in the mid band are concerning (phase related?) and of course this dip between 60 and 100.

As I’m on my phone just now it’s easiest to share via Dropbox: https://www.dropbox.com/s/46v8twtihj5oa ... .mdat?dl=0

Any thoughts welcome.
Gareth

[endorka]

I think you are on to something with the phase. This is all conjecture and at the limits of my knowledge, but anyway...

Have a look at the blue rectangle in this amroc plot of room modes. The lowest in that area is length axial 125Hz. The higher three are 152, 157, 158 and are all very tightly bunched together. These are all axial modes corresponding to height, length and width respectively.

Note how these correspond to a big shift in phase in the spl & phase plot below. Is it possible that the bunching of those is causing a perfect storm of bad effects?

I think phase changes are caused by reflections. If room modes are contributing to this, moving the measurement mic (say) six inches forward, then six inches back should show a difference in this area. Same with width and height. Of course the latter two are fixed, but moving the mic for the sake of pinpointing the problem might gain useful knowledge.

Have a look at this post from Stuart on his walking mic measurement technique, there may be some useful info there.
https://www.digistar.cl/Forum/viewtopic.php?f=9&t=24

Would it also be possible that you might be getting deleterious effects from similar bunchings at multiples of those frequencies? E.g. 500Hz to 632Hz ?

Just conjecture, like I say I am at my limits here, this is something I'd personally test though, at least to rule out it.

Cheers!
Jennifer

[garethmetcalf]

That's really interesting, thanks Jennifer. The room dimension with the f0 at 150ish is the height, so I thought I'd have a look at before/after the ceiling treatment (100mm rockwool).

I'm not very good with how REW represents phase other than knowing the abrupt vertical line is wrapping the reading rather than showing a massive change in phase. Here are the graphs of the SPL and phase before adding ceiling treatment and after - the phase change around 150Hz changes, as does the SPL response around that area (for the worse I'd say).

Before ceiling treatment

After ceiling treatment

One of the proposals from the first acoustic consultant was to add some reflection to the ceiling, so I wonder whether this phase/150Hz area issue could be changed in part by putting a reflective surface/cloud above the desk and angled to the back of the room, to break up that axial mode, thus changing the phase relationship with the others?

What I might do is run a live REW reading and move the mic around to see if I can see the frequency response changing in that area, rather than trying the walking mic test just now.

[endorka]

The abrupt vertical line is indeed wrapping. In the graph I posted I'd selected the option to "unwrap phase", so the sudden change there is real. I reckon when looking at smaller portions of the graph it allows me to see sudden changes more clearly.

Sometimes adding treatment can seem to make things worse. It's worth bearing in mind the thing Starlight mentioned about solving problems that allow previously hidden ones to reveal themselves.

Good plan to do the live frequency response thing, I look forward to hearing what it reveals.

At the back of my mind I have something Rod Gervais mentioned in his book. From memory, about not focusing too much on getting one precise listening point because it is not how we hear things - we have two ears and we don't keep our head precisely located. I think he recommended taking an average of readings around a volume of space of the listening spot. I worry about saying this because not all the acoustics folks recommend this, and the potential to lead to a path of confusion. But there it is.

EDIT: I just looked it up. Home Recording Studio, build it like the pros. Page 183 onwards, "Operation of Analyzers". The method he describes has some things in common with the walking mic technique, albeit with a slightly different philosophy. That whole chapter is excellent by the way, I could certainly do with reading it again for a refresher.

Cheers!
Jennifer

[gullfo]

very true - you'll see in some acoustic testing specs the listening position is defined as a circle or oval, and there are at least a dozen or so spots within the listening position for the mic to be placed. this may be a bit much but certainly testing a minimum of 4 spots - one for each ear and nose and back of head would be a good choice as well as maybe another 4 out 12" (30cm) to check on the extent of the listening position.

[garethmetcalf]

Hi all
I can't believe it's been three months since I last posted about this (sounds like a confessional!). I've been focussing on building my business processes and website etc but I did a mix last week that I had to go to and from the car and hifi in house to get the bass right.

I therefore thought it was worth investigating the impact of listening position/measurement mic position on the low end frequency response.

I think in this thread we'd already determined that although my treatment has absorbed the general wash of low end reverberation I'm left with the axial modes and some phase interactions and it was suggested to try a moving mic type technique.

I used the REW RTA which allows for snapshots to be captured and played pink noise from 10Hz to 1000Hz. I tried a few mic positions - making relatively big movements, the most interesting/useful of which are posted below.

The position 60cm forward of the designed mix position seems to have the flattest response, but feels a bit odd as I'm right up close to the speakers and soffit wall. My ears are a bit shot so I can't make any more sense of this just now, but I wondered if anyone had any observations?

I fixated on a few of these frequencies using a tone generator and wandered the room. The 55Hz ish null is worst at around 1/4 and 3/4 room dimensions, but according to the bob golds calculator is associated with the ceiling dimension!?

Would you move your mix position to the 60cm forward?

Cheers
Gareth

[garethmetcalf]

Update - listening again with fresh ears I've put the listening position back where it was. I didn't like how it sounded being so close to the monitors, so I think it's back to treatment options. I've messaged Stuart for some paid advice, but if anyone has any thoughts I'd appreciate it.

Considering:
1. hard backed cloud to break up floor/ceiling mode
2. slatted wings that allow for thicker insulation to the left and right, and maybe tune them to the 100Hz (ish) mode
3. tuned absorbers at the back in place of some of the hangers (maybe in the middle?)
4. EQ to turn down the modal peaks (esp 33Hz)