Room responses: What tolerances do you guys work to?

[Barish]

Hello there, first post here.

I have recently done a budget mixing room here in Istanbul. The bare room dimensions are 4.80m (w) 4.60m (L) and 2.60m (H).

I have installed a pair of Genelec 1038Bs in flush mount onto the wall at a 60 degree angle, and for the owner likes to work sitting pretty low -lower than usual desk height, I had to tilt the speakers forward by 15 degrees in order to make the acoustical axis of the 1038s fall within somewhere 1.10mt to 1.30mt from the floor at the seating point. I have also designed and custom-built a desk with racks so that he could manipulate the outboards without having to change listening position.


1038Bs were flush-mounted onto the front wall angled at 15 degrees to the listening point of 1.8 metres from the speakers on the acoustic axis. Room sizes after installation are 4.60mt width to 4.10mt depth and 2.50mt height at the listening point, which is also our measurement point, and is 1.75 metres from the front wall on the depth axis.

The red trace below is L speaker only, the blue is R speaker only, and the black is when both speakers are on. The target level is 69dB and all of the traces are smoothened to 1/6 octave.

The speakers have bass level trims down by -4dB, and all the other level and tilt trims are left flat.

I have achieved a pretty good mid and top end response with 2dB fluctuations on average beyond 700Hz , and a pretty flat response up to 100Hz at the bottom. But I have some 5dB to 9dB dips at 150Hz, 225Hz, 350Hz, 400Hz and 600Hz, which look like an effect of the room size according to the acoustical axes of the speakers. I try to avoid any further absorbtion for there is already a relatively high absorbtion affecting 100-700Hz region anyway, and also a double glazed window and entry door on either side which I am required to leave uncovered. For I was not able to change the positioning of the window within the budget I was given, I have placed the door 5 degrees off in order to avoid a flutter echo between the window and the door.

The budget does not allow me to treat that region with custom-built QRDs, and there's no room to place such deep installations at the back of the room either. To my experience, the depth of the room is simply too small to fix such variations without getting into complex diffusion solutions, for which we have no budget (man I have used that budget word so many times, I don't want to hear it anymore). I've tried to accommodate the user convenience and the acoustical benefits as much as I could in the placement, although they don't always agree with each other. Trade-offs.

The arrangement in the room was done exactly according to the client's requirements as he has a lot of outboard equipment that he wants to have an access to without having to move out of the listening position, so he is happy with the furnishing, and there's not much else I can do in that space anyway. He wants both space with minimal installation, and in the mean time almost theoretical acoustic perfection, which is pretty hard task to achieve.

The client was actually pretty chuffed when he listened to the room without seeing the chart, and was over the moon, but when he asked me to run the test and then saw the results -which I do as a standard procedure whether I am asked or not anyhow-, he started picking on these midrange dips now, although he admits that he has seen and worked in very prominent places in the UK and US with worse statistics than these. But he just won't give up and he wants to see that low mid range as flat as the top end.

So I would love to hear what tolerances you guys usually work to in terms of room responses.

Any feedback would be very much appreciated.

Thanks.

B.

[Soundman2020]

Hi Barish. Welcome.

There seem to be several issues that stick out from your post. You say:

"1038Bs were flush-mounted onto the front wall angled at 15 degrees to the listening point".

Did you angle the entire walls, or just the speakers? From the 3D model you posted, the walls are still vertical! If you only tilted the speaker in the wall without also tilting the panel around it, then you no longer have an infinite baffle. I have no idea what effect that will have, but I would guess it would create all kinds if strange interference patterns in the sound field coming off that wall, since different parts of each sound wave will be coming of different parts of the wall at different times. Phasing, interference, comb filtering, etc. Probably unpredictable, but definitely not nice!

Also, 15 degrees is a very large angle to tilt speakers. The EBU recommends no more than 10 degrees, max, and other experts say it should not go more than 7 degrees, due the way the human ear works. (Yes, you will find some places on the internet that say you can tilt your speakers at 20 or 30 degrees, but then again, you also find places on the internet that say egg crates and carpet in the walls make a great studio.....) The truth is, 15 degrees is a steep angle, and you are probably getting major comb filtering artifacts coming off the desk, racks and other furniture at that kind of angle.

The speakers have bass level trims down by -4dB, and all the other level and tilt trims are left flat.

It should be -6dB for flush mounted speakers, not -4. (Well, assuming that it is correctly mounted, that is!)

But I have some 5dB to 9dB dips at 150Hz, 225Hz, 350Hz, 400Hz and 600Hz, which look like an effect of the room size according to the acoustical axes of the speakers.

My guess is that what you are seeing there is comb filtering from the furniture, combined with the artifacts from the non-tilted flush mount panel, plus probably room modes from having the mix position at the 50% position (apparently, from looking at the diagrams).

I try to avoid any further absorbtion for there is already a relatively high absorbtion affecting 100-700Hz region anyway,

How did you manage to tune the absorption to only affect that frequency range?

Where is the absorption in the room, and how thick is it?

Do you have 6" thick absorption at all your first reflection points?

Do you have major bass trapping in all 12 corners?

What other treatment do you have in there?

and also a double glazed window and entry door on either side which I am required to leave uncovered.

You are probably also getting reflections and comb filtering from those. It sounds like you have an acoustic mess in there!

And double glazed? you mean as in standard household / office double glazed panels? Then you are probably getting some ringing from the glass resonance, too. Those are made from thin plate glass, the resonant frequency of which is probably right in the range of your problems. So the glass is probably singing along with the music...

I have placed the door 5 degrees off in order to avoid a flutter echo between the window and the door.

5° MIGHT stop flutter echo, but it won't do a thing for comb filtering!

The budget does not allow me to treat that region with custom-built QRDs,

Good, because the room is too small for diffusers to be effective, and I doubt they would do anything to solve your problems anyway.

He wants both space with minimal installation, and in the mean time almost theoretical acoustic perfection, which is pretty hard task to achieve.

So let's get this clear: He wants acoustic perfection, but he won't let you do what needs to be done, and he doesn't have any budget anyway? Ummm......

But he just won't give up and he wants to see that low mid range as flat as the top end.

Well, that just is not going to happen unless he is prepared to let you change what needs to be changed and give you the budget to do it....

Do you think that this frequency response looks reasonable to you?

Frankly? No! You have some pretty big dips and peaks there, differences of around 16 dB by the looks of it, and you say that is already a smoother graph (1-6 octave), so there are most likely some REALLY big problems hidden in there as well, most likely in the lows and mid lows.

I have already run out of budget,

Then, frankly, there isn't anything you can do to fix it. You'd have to start by ripping out the entire front wall, tilting the infinite baffle panels to the same angle as the speakers, which should NOT be more than 10° anyway, the re-arranging the geometry of the room correctly, to get the mix position out of its current 50% location and closer to the 38% position, then treating the first reflection points with thick absorption, then measuring again to see how well that worked and what still needs fixing.

You might also have to re-design those racks, since I see an awful lot of angled surfaces there that seem to be facing the speakers. There's probably a reflection party going on in there, with all sound waves invited!

but the client is arguing that this should be improved either way as opposed to my argument which is that he may still not like what he hears after I start altering the placement.

Did he design and position the desk, or did you do it? If he did it, then you might be off the hook: It is his choice to sit so low that the speakers cannot be aimed correctly, and also his choice to have a desk that reflects stuff back all over the place, and causes major comb filtering. If he won't sit at the right height, then the speakers cannot be tilted to the correct angle, and the reflections and comb filtering will not go away.

Your escape might be that it's his desk and his listening position that is causing the problem.

Easy way to check: Place a very thick layer of absorption over the entire desk and racks, then measure again. My bet is that you'll see a large improvement just from that, without even fixing the infinite baffles.

we did a test and tried to lift that range by using a broad-Q transparent EQ

Well, I'm not a big fan of trying to solve an room acoustic problem with EQ anyway, since it will only ever work for one specific position in the room, but if he thinks it sounds better and the graph looks better, then who are yo to argue? "THe customer is always right" (even though he might be an .....)

So I would love to hear what tolerances you guys usually work to in terms of room responses.

Well, for a small room, you'll never get it flat, of course. Not even close. If you can get it reasonably flat and within 10dB, you are probably doing fine, but that means WITHOUT smoothing. I'm talking about the raw truth: 1Hz intervals from 20 up to 200 Hz at least, then you can relax it a bit for the high mids and even more for the highs. But smoothing can hide huge peaks and dips, so smoothing is out. And my guess is that you have some of those big peaks and dips hidden in there...

But anyway, 'm not an expert on this stuff by any means: I'm just offering my opinion of what I think the problems might be, and what I'd do to fix it if it were my room.

Maybe one of the real experts here can give you a totally different opinion!


- Stuart -

[Barish]

Hi Soundman hows it going? Thanks for your comments.

The infinite baffles are angled the same way as the speakers. And there is an absorbtion all along the ceiling and the back wall as deep as 2 ft at the back top corner. I couldn't get in any more than that due to the concerns of losing too much room space.

15 degrees was his choice as he did not want to sit far back. He was even wanting to use three sets of speakers to check his mixes apart from the flush mounted 1038Bs -I know, crazy idea- and all with the perfect placement -even crazier- and the speakers were Genelec 1032As, Adam P11s and a pair of NS10Ms

We finally came down to the 1038Bs, 7071 sub, and NS10Ms. The sub was switched off during above test.

The speakers are correctly mounted, so we do not get any resonance from the wall. The carrying structure and the front wall are two separate systems with correct neoprene coupling as is described in Genelec's installation manual. (Funny thing is, the drawing in there shows an even steeper angle, but I digress.)

I was also thinking of placing a thick overhead cloud over the listening point, and perhaps two panels on the sides to give the lows more taming, but I haven't tested how it would affect the overall mid level yet.

The doors had to be like that, for I had initially quoted a solid door but he wanted it glass so that he could see the drums room, and a sliding one as well, but when it came to the price difference there was a big argument so it had to stay there. I even paid for the glass difference out of my pocket.

The differences are not as wide as 16dB, but rather 13dB at most on per-channel test, and 9dB in L+R sum. The unsmoothened graph shows no other severe problems than above.

You can't get to 38% because it gets too close to the speakers and you can't fit the desk in that area anyway.

You just don't want to touch the desk. It was designed exactly according to his descriptions and it is probably the only thing he does not complain about. Oh wait, it was supposed to have a moving top on rails so that he could pull it in to get the stuff on top closer to him at will as well, but luckily we could live without it.

And as a final note, I would never make any comments about my client's personal character like that, and he IS always right, within reason, and I told him that I aim to please "within" reason.

B.