View Single Post
Old 14th August 2019
Lives for gear

Originally Posted by nms View Post
1. The frequencies which define the stereo image are primarily above 150hz.

2. Low frequencies radiate from the speaker in an increasingly omnidirectional pattern the lower down you go.

So with that in mind, we know that the first reflection off the front wall is going to be a low frequency bounce.

And what does physics tell us about delaying that powerful LF reflection? By increasing the distance from the front wall you've now carved a hole in your LF response courtesy of phase cancellation from the delayed low frequencies. And it's not just the reflection right in front of you, but also the front wall to ceiling edge, floor edge, and front room corners. You've increased the delay time in the LF reflections from the entire front end of the room. This usually doesn't go well.

Maybe you think placing speakers close to the wall compromises stereo image because you tried it and the LF boost washed out the upper frequencies? If you haven't done a good job of treating the front end of the room, and you don't have any LF EQ adjustment on your speakers to compensate, then it may not work out well. Just don't make the mistake of thinking that's going to apply to someone else's room where circumstances are different.
Exactly! Well explained. If someone does experience degraded stereo imaging from moving the speakers close to the front wall, that's a sure indicator that the room has serious acoustic issues. It needs proper treatment on the rear wall, and likely the other walls and the ceiling too. There's a problem with the treatment in that room.

Originally Posted by Synthpark View Post
Not really. You can measure the influence of placing an absorber behind the monitors at least up to 500 Hz,
That's not coming from the back of the speaker: you are measuring an improvement because you are treating the reflections and decaying field from the rest of the room. At 500 Hz, it's going to be -10dB down at least at the rear of the speaker, and probably more like -15 or -20, depending on the speaker.

Besides, there's not much stereo information at 500 Hz: the HRTF only really gets useful from about 1 kHz upwards. Below 1 kHz, your brain uses mostly ITD to determine direction, and for a pair of speakers equidistant from your ears, there's not much ITD. For higher frequencies, ILD comes into play as well. Your ear and brain do best for directionality in the range 1 kHz to 5 kHz, with some additional stuff up to maybe 15 k or so. But below about 500 Hz, direction detection isn't that great, and certainly below 250 there's very little at all.

In other words, if your stereo imaging improves from front-wall absorption acting around 500 Hz, then that's a sure indicator that there's a problem with treatment elsewhere in the room.

Originally Posted by nms View Post
This is why comprehensive rear wall treatment is so important, and where that's lacking you use absorption between the speakers to deal with what remains. The sound radiating out the rear of the speaker is primarily low frequency content though.

By moving the speakers closer to the wall, the problems become higher in frequency and easier to treat.
Exactly. +1

Originally Posted by Synthpark View Post
I have comprehensive rear wall treatment in the mids realized with RPG skyline diffusors (just light enough for the door in the middle). They work very well at 500 Hz. If rear wall was such a problem, I would measure it in the impulse response.
That confirms what we suspected: your rear wall treatment is insufficient, and incorrect. It's not surprising that you are having imaging problems, if you don't have a fully treated rear wall. Diffusion at 500 Hz is irrelevant to the REAL problem with your room. The real problems will be below 200 Hz. Since those are not being dealt with, once again it's not uprising that you don't get good acoustic response in the room. The back wall is the single most important surface in the room, in pretty much any control room. The other surfaces are important to, to a lesser extent, but the rear wall is the key to getting good response in the room. If all you have there is a diffuser, then that explains the problems you are describing.

It would help if you could do some proper testing with REW, and post your MDAT file here, so we can see what is actually happening there, and help you fix it.

So my conclusion, once again, is that the front wall directly, via the first reflection, is responsible for this deviation.
Only because the rest of the room is not treated properly. You are placing the blame in the wrong place: You hear a difference when you put the speakers against the front wall, not because there is a problem with speakers against the front wall, but because the rear wall of the room is not properly treated, and the resulting poor acoustics response of the room is overwhelming your perception of the speakers when they are against the front wall. You are seeing a symptom, and attributing it to the wrong cause. Sort of like assuming that if your car does not steer straight, it must be the steering wheel that is bad, because that's where you feel the problem. But the REAL issue is that one of your front tyres is flat....

And that reflected wave would hit me first. Otherwise it would loose even more energy travelling to the front wall and back.
Not really. Attenuation of low frequencies over short distances is just a fraction of a decibel. Not even noticeable.
You have only some diffusors on the rear wall? Of course you're going to have lots of reflected sound returning to the front wall. We have a very different definition of comprehensive.
+1 So true.

500 Hz, even 300-350 Hz, doesn't it follow more or less geometrical acoustics?
Not necessarily. It depends on the speaker itself, and also on the room. Don't forget that what you see in the acoustic data published by the manufacturer is take under perfection conditions in a large anechoic chamber. But once you put the speaker in a real room, the actual loading changes dramatically, so the performance is rather different. Read Floyd Toole's excellent book, "Sound Reproduction", and you'll start to see just how big that change can be. Or Kliener and Tichy's book "Acoustics of Small Rooms", to see other examples of how a speaker's response can be so different in a real room, as compared to the ideal measurement in an anechoic chamber.

Even if I put some thick absorbers below the diffusors it doesnt essentially change the frequency reponse, I tried it.
You are looking in the wrong place for the wrong reasons! Frequency response is not what you should be looking at for rear wall treatment: time domain response is.

I can post an impulse response later.
That would be good, yes. But don't post just the IR: post the actual MDAT file.

Most problems result from the arrangement directly in the front of the room, around the speakers (windows, ground, table), and the ceiling.
Likely because the rest of the room isn't treated correctly, especially the rear wall. In a room that has all of the surface treated except the front wall, there won't be major changes from doing things to that front wall. If there are major changes from doing something on the front wall, that indicates that the rest of the room is not treated properly.

There are side wall absorbers with a length of 1,80 meters from the front corners into the room, and as heigh as the ceiling, so the sidewalls are treated more comprehensively,
Not necessarily! It's not just the location of the treatment, but also the type and thickness, and the speakers themselves also have an effect... but even if the side walls are well treated without the rear wall being treated heavily, the overall acoustic response won't be good.

50 cm thick rockwool,
Your treatment on your first reflection points is 50 CENTIMETERS thick? Nearly 20 INCHES? In a room that is 460 cm wide? So you are taking up one entire meter of room width for side-wall treatment (50cm on each side wall)?

I was never able to measure any significant change in the frequency response using additional treatment (by placing 50cm thick rockwool absorbers) at not covered areas at the rear wall. That surprised me
Once again: the changes you want to see will not be very evident in the frequency response. The rear wall is all about time-domain response. There should be some changes in FR, yes, but the changes you are really looking for will be in the time domain. It's a common mistake to concentrate too much on the frequency response graphs for a room: the other graphs are far more informative. Especially ETC.
And reverberation is mostly covered with the side wall absorbers,
Not really. That will only help with flutter echo, and width-wise axial modes. It won't do anything for length-wise axials (which are always the biggest issue), nor vertical axials, nor SBIR from the front or rear walls, nor floor bounce, nor ceiling bounce, or the ITDG.

mostly covered with the side wall absorbers, down to 50-60 Hz.
With one meter thickness of absorption on your side walls (50 cm on each side) you should be getting a lot lower than that, for your width-related axials, as well as tangentials that involve the side walls, and all obliques too.

Not that I wouldn't like to be able to improve room response using rear wall treatment, this would be more convenient than other methods.
Then post your MDAT, and we can help you do that! But probably better do that on your own thread, so we don't hijack the OP's thread any more... Your room is big enough that it can be rather good, if set up and treated properly.

- Stuart -