John Sayers' Recording Studio Design Forum Archive

Hi
I don’t know if Stuart or anyone else is reading all my posts, it is rather quiet here. Anyway I will ramble on in the hope that somebody can help out…

So I finally screwed up the soffits and attached the heavy duty side wings and ran yet another rew test. Comparing the previous test (with just the first layer of baffle attached) with this most recent one (with full baffles and side wings) I was disturbed to see the results seem to have got worse. Maybe I am misunderstanding the results, I am very far from being an expert, but at this point I am tempted to just take the side wings off and chuck them away.

I thought I would share the results here in case anyone had any opinion about this.

As a comparison here is the previous mdat (with just the first layer of baffle attached)…
https://www.dropbox.com/s/7a2hik8ltk7nz ... .mdat?dl=0

and here is the new mdat (with full baffles and side wings)…
https://www.dropbox.com/s/ltv21wrhlmd3x ... .mdat?dl=0

My areas of concern are …

400hz - a worrying dip around 6db worse when I add the wings. Not that much of a difference at other frequencies except …

…between 4k and 7k there is a huge 20dB dip which is completely absent in the rew taken before I added the side wings. Also this is not apparent on the individual L and R speaker readings which look ok in this range even wth the side wings. I don’t even know if I should be worried by this such is my level of incompetence.

Waterfalls - I was hoping that the addition of the side wings would result in the room being a little less dry but the addition of several m2 of hard reflective surfaces does not seem to have made an appreciable difference. I suppose it would probably be better with some more life in the room but I can’t see how to improve this. In a larger room maybe slats would help but already my side walls are almost entirely reflective being mainly made up with the side wings, windows and doors. I am not sure putting slats on the rear wall or ceiling is a good idea.

Anyway there it is. Quick reference graphs are below. Any advice greatfully received.
Cheers David

I don’t know if Stuart or anyone else is reading all my posts, it is rather quiet here.

Still here! Still following! Still interested! Just a bit overwhelmed with workload right now...

I thought I would share the results here in case anyone had any opinion about this.

Great! I just downloaded the MDAT files, and I have to say that your room response is looking pretty good!

However, there's one thing that worries me: one set of measurements is called "withbaff and eq". I'm not sure what that means, but if you had some type of equalization applied when you took those measurements, then they are no use. If that's the case, then you'll have to go back and do those readings again. EQ will have totally screwed up any possibility of seeing the REAL response of the room, so you can throw out any measurements you did with EQ, go back to the room situation at that time, and repeat the measurements with totally flat response across the entire signal chain.

The reason is quite simple: using EQ disguises the true response, so you don not know what the remaining problems in the room are! It's like going to see the doctor for your fever, and giving him all your temperature readings made AFTER you already took analgesics and anti-inflammation medication! Those readings are totally useless and invalid, since the medication is disguising the fever... So the doc won't be able to diagnose anything about your illness, since you covered up and hid the symptoms with the medication.

Yes, EQ, can be used to smooth over some final remaining issues in the room, but only under certain circumstances, and only AFTER the room has been fully completed, and fully treated, and fully tuned acoustically, until the necessary conditions are met. Using EQ before you even have your treatment in is pointless, since it grossly distorts the true situation.

400hz - a worrying dip around 6db worse when I add the wings

Are you certain that you designed the wings to be at the correct angle? Did you ray-trace that, to make sure? Are you SURE you did the ray-tracing correctly? From my point if view, the angle of your wings does not seem large enough.... It's hard to tell just from a photo, but that's what it looks like to me. There are new reflections appearing in the L and R IR measurements, that seem to be at about the right delay (location) for a bounce from the wings...

…between 4k and 7k there is a huge 20dB dip which is completely absent in the rew taken before I added the side wings. Also this is not apparent on the individual L and R speaker readings which look ok in this range even wth the side wings. I don’t even know if I should be worried by this such is my level of incompetence.

Nothing to worry about: that dip is not actually there in reality. It's likely due to just having the mic perfectly located on the center line of the room. You would never actually hear that, since you do not have just one single ear located in the exact middle of your head! Rather, you have two separate ears, located a few inches off to the side of where the mic was for that reading, and the response that each ear will hear is far closer to the individual L and R readings, than it is to the combined LR reading. The LR reading is mostly to check the low end with both speakers on, not the highs. The low end can often suffer from interference between the two speakers, which is why it needs to be checked, as you really will hear that (low frequency = long wavelengths = valid at the mic position).

So don't be concerned at all about what you see about about 500 Hz or so, in the combined LR readings.

I was hoping that the addition of the side wings would result in the room being a little less dry

It is too dry, certainly. The overall decay times are all around 110 ms, which is way too low for that room. It should be more like 230 ms. for that sized room.

the addition of several m2 of hard reflective surfaces does not seem to have made an appreciable difference.

I disagree! There's a clear increase of about 10 to 15 ms, across the board. It's a bit more in the highs than the lows, but it's there all over.

I suppose it would probably be better with some more life in the room

Definitely!

I am not sure putting slats on the rear wall or ceiling is a good idea.

Not on the ceiling, probably, but there's nothing wrong with doing it on the rear walls ... provided that it is done intelligently and correctly: correct sized slats, correct placement, and correct coverage area. It's an easy way of increasing decay rates that are too low. You wont get a huge increase from making the front o the sides of the room more reflective, if the rear is dead! The front and sides of the room should be designed to send all the reflections to the rear of the room, and if the rear is pure absorption, that energy will never get back to your ears: hence, the room remains "dead". You do need "stuff" on the rear to send some of that energy back towards you, but in a carefully controlled manner...

If you want to test this to see what I'm taking about, just put a couple of full sheers of plywood standing up vertically, leaning against the back wall, and do a REW test like that. I think you'll see the difference!

But before you do that, remove all EQ from your signal chain and repeat the above tests. And also double-check the angles of your wings: I strongly suspect that they are not angled correctly.



- Stuart -

Soundman2020 wrote: Great! I just downloaded the MDAT files, and I have to say that your room response is looking pretty good!

That is very good to hear Stuart.

Soundman2020 wrote: However, there's one thing that worries me: one set of measurements is called "withbaff and eq". I'm not sure what that means, but if you had some type of equalization applied when you took those measurements, then they are no use. If that's the case, then you'll have to go back and do those readings again. EQ will have totally screwed up any possibility of seeing the REAL response of the room, so you can throw out any measurements you did with EQ, go back to the room situation at that time, and repeat the measurements with totally flat response across the entire signal chain.

My purpose in doing a ‘with eq’ test was to try to work out what were the optumum speaker settings before burying the speakers in the baffles. Neumann recommend settings for various placements and so I thought I should try them out before the controls became unavailable - there is not much point in discovering that there is too much bass at the point where it then becomes impossible to turn it down. The graph should really then be labelled ‘with speaker compensation’ but it is just eq really isn’t it. Now the speakers are buried in the baffles and it would be hours of effort to get to them so I hope I set it for the optimum performance.

Soundman2020 wrote: But before you do that, remove all EQ from your signal chain and repeat the above tests.

Would you advise to do all the acoustic testing without using any speaker compensation?

Soundman2020 wrote: Are you certain that you designed the wings to be at the correct angle? Did you ray-trace that, to make sure? Are you SURE you did the ray-tracing correctly? From my point if view, the angle of your wings does not seem large enough....

Well I am not quite certain. I did it to the best of my ability and without any instructions which is not quite the same thing! The constraints of the room made it hard to get enough angle but I thought it looked ok on the ray tracing. (diagram with wing angle and ray tracing at the end of this post)

It is possible I messed up slightly the actual building of them, I am far from being a good craftsman, but it should be accurate to about a cm.
Anyway my options at this point are —
- leave it how it is and live with it
- replace side wings with fabric
- make adjustments to side wings

Having looked at the ray tracing and design one more time I don’t know if there is much adjustment i can do. I may be able to bring the inside edge of the wings forward an inch (maybe 2) which would result in a steeper angle but that would leave a gap between the side wing and the front baffle, don’t know if that is bad or not?

Soundman2020 wrote: It is too dry, certainly. The overall decay times are all around 110 ms, which is way too low for that room. It should be more like 230 ms. for that sized room.

Yes I thought so, I just can't see how I can begin to extend the decay to 230ms.

Soundman2020 wrote: You do need "stuff" on the rear to send some of that energy back towards you, but in a carefully controlled manner...

That is one adjustment I was half expecting. But before I add “stuff” I would like to work out what “stuff” that should be and where to add it effectively. I can see that higher frequency reflections that are directed to somewhere other than the listening position might be a good idea to increase the general liveliness so I suppose I should aim for that.
I haven’t worked in the room for that much time yet but so far I don’t find the dryness too bad, maybe because there are massive big baffles right infront of me that reflect all the sound I make directly back at me. If can improve things by adding a some slats I will definitely give that a try.

I am guessing that I should use slats of differing widths across the sides of the back wall with the spacing between them varied. Is that approximately correct?
I have looked around for info on slat size and spacing but there seems to be lots of different recommendations out there. As a rough guess I am thinking 3 sizes of slats between about 3cm and 10cm wide (whatever the wood shop has at the same thickness) spaced apart with gaps of around 1-3cm, placed in a pseudo-random order? Any further tips on this Stuart before I do it wrong as usual?

diagram of front baffle/wing angle with faint ray traces of limiting angles below …
also a rough sketch of dimensions of rear wall area - there are hangers to the L and R of the sofa and the dotted line indicates the fabric covering. Behind the sofa is 20cm of rockwool and some smaller hangers above …