I would try using a windowed (frequency dependent) minimum phase version of the listening position measurements as a basis for any analog eq that is to be applied:
-Generate a minimum phase version of each measurement (Controls - Generate Minimum Phase - File - Export - Impulse Response as WAV (check "Export Min Phase Version of IR"))
-Open the min phase versions with REW and apply a 5 cycle frequency dependent window (IR Windows - (check "Add frequency dependent window", and specify "Width in cycles" to be 5) Apply Windows To All )
At this point, you can use REW to calculate the settings for your 1/3 octave eq although I've no personal experience with that feature.
-Generate a minimum phase version of each measurement (Controls - Generate Minimum Phase - File - Export - Impulse Response as WAV (check "Export Min Phase Version of IR"))
-Open the min phase versions with REW and apply a 5 cycle frequency dependent window (IR Windows - (check "Add frequency dependent window", and specify "Width in cycles" to be 5) Apply Windows To All )
At this point, you can use REW to calculate the settings for your 1/3 octave eq although I've no personal experience with that feature.
Your probably a steps ahead of me... First I'd like to see the Impulse Responses as we are still looking for clue's here. Let's see what the speaker does in the early miliseconds, and the room, go from there...
But we'd probably need to see that without any EQ first. Just to get a bit of an idea of the speakers impulse and the room's contributions.
If we can fix some problems with placement and maybe even damping here and there etc. we don't need the EQ to do that much.
I agree though, with your proposal but I wouldn't start there. Start with the room first. See what's going on.
(I keep adding, lol) The Auto EQ in REW works really well. But you can always improve on it by softening some of the Q settings that are proposed. Broad strokes of EQ work better than high Q fills and boosts. REW does try to fill everything.
But we'd probably need to see that without any EQ first. Just to get a bit of an idea of the speakers impulse and the room's contributions.
If we can fix some problems with placement and maybe even damping here and there etc. we don't need the EQ to do that much.
I agree though, with your proposal but I wouldn't start there. Start with the room first. See what's going on.
(I keep adding, lol) The Auto EQ in REW works really well. But you can always improve on it by softening some of the Q settings that are proposed. Broad strokes of EQ work better than high Q fills and boosts. REW does try to fill everything.
Last edited:
Thank you for the responses. I will try to get more measurements this week, with the impulse response and also without any EQ.
One questions remains... how should I start assessing the room acoustics? I have built two bass traps that are basically light wooden frames, 6" deep, 17" wide, and 72" tall. I am about to finish these by putting these Ultratouch insulation batts into them, covering them with thick fabric, and mounting them about 3" off the wall right behind the speakers. That was just an idea that seemed most appropriate... I could place them elsewhere though, and I could also make four out of the two bass traps (they are essentially 2 x 2 frames mounted as one and could be split). Here is a picture:
(construction may look a bit awkward, but I wanted to use what I had at hand... and they are actually quite sturdy thanks to all those lap-joints!)
I am also considering to hang panels from the ceiling... there is a proper name for it, I just can't remember right now. It would not be a problem to cover most of the ceiling with hanging panels, but I am not sure how much that would help. It would be more acceptable by the wife than putting more bass traps into the corners though... I suppose...
One questions remains... how should I start assessing the room acoustics? I have built two bass traps that are basically light wooden frames, 6" deep, 17" wide, and 72" tall. I am about to finish these by putting these Ultratouch insulation batts into them, covering them with thick fabric, and mounting them about 3" off the wall right behind the speakers. That was just an idea that seemed most appropriate... I could place them elsewhere though, and I could also make four out of the two bass traps (they are essentially 2 x 2 frames mounted as one and could be split). Here is a picture:
(construction may look a bit awkward, but I wanted to use what I had at hand... and they are actually quite sturdy thanks to all those lap-joints!)
I am also considering to hang panels from the ceiling... there is a proper name for it, I just can't remember right now. It would not be a problem to cover most of the ceiling with hanging panels, but I am not sure how much that would help. It would be more acceptable by the wife than putting more bass traps into the corners though... I suppose...
Last edited:
OK, I have a few new graphs. I hope I am posting the things you asked for... I haven't really studied REW well enough to understand everything.
Here it goes. Yellow is before eq'ing, blue is after eq'ing:
Let me know if I should post some more/something else.
I still wonder, where should I put the bass traps? I though that behind the speakers would make the most sense in my room...
Here it goes. Yellow is before eq'ing, blue is after eq'ing:
Let me know if I should post some more/something else.
I still wonder, where should I put the bass traps? I though that behind the speakers would make the most sense in my room...
Are these from the "Filtered IR" tab? Could you post the plots on the regular "Impulse" tab? When you hover the graph window it should say "% FS" in the upper left corner of the graph. If this is the "Impulse" tab it probably states: "dB FS" there instead. Switch it to "% FS". Also see the time scale below the graph. Zoom in on the first ~ 20 ms. These graphs are showing a 1 second time span. You see 200 m, 400 m, zoom in until you get
Once we see the first 20 or 30 ms we can see what the room is doing. I see some large late peaks on this graph but it's not telling us much without zooming in.
You can zoom in by typing in the limits manually, click on limits in the upper right corner:
Set it to something like this:
That should fit the first 20 ms of the impulse on the screen...
Be sure to set it to "% FS" first...
Attachments
1) Why? you didn't answer my question. Putting a speaker in a room null severely over damps it, the opposite of what is generally needed.
2) ??? Please explain.
3) The one the speaker is in of course!
GM
OK, but that's not what you wrote unfortunately.
No contradiction, just apparently too large a language barrier.
GM
OK, adjusted it as suggested. I just changed the top and bottom limits to 120 and -120 though, that did fit better. The red one is without EQ, the green one with:
Assuming your layout is still like in post #1, your whole bay window ideally needs to be damped and you ideally need to isolate the piano and damp one of each parallel wall around the speakers, i.e. one side and back. When the LP is up against the back wall, the bass traps normally go to the corners, but neither is viable, so [re]tuning the speakers or using room EQ is about all you can do. That, or roll off the mains and use a separate sub system as a coffee table positioned at an odd harmonic between the speaker and LP.
As already noted, some serious damping is required around the LP. Wedges in the corners, both vertical and horizontal will help, though ideally all the exposed wall around the LP needs damping.
GM
That's better, for a more clear picture you might want to un-tick Envelope (ETC) and Step Response. That will show only the impulse. Are these speakers still in the position as outlined on the first page?
Looking at this picture you can see the large peak, that's your speaker and some smaller ones at 3-4 ms and 7 ms, 11 ms 12 ms. All those peaks are most probably reflections of nearby surfaces. See the discussion between the two gentlemen posted before your reply 😉.
I'm going to sleep now, much to late already but knowing the speed of sound, and the milliseconds at which you see the spikes you can translate that into distances. If you measured at or in front of the couch I kind of expect the wiggles at 7 ms to be from the wall behind it. The ones at 3-4 ms are somewhere closer to the speakers. So try to find these spots by looking (and imagining) where the sound could bounce back to the listening position. (and arrive 3 ms later or 7 ms, get the drift?)
Tackle one at the time and confirm with a new measurement. Get the big and early ones first! You could also try and move the speaker to see what changes and figure out why.
It's a learning process but you have to be prepared to do some work. Are you 🙂
I'll tell you, it is rewarding.
Looking at this picture you can see the large peak, that's your speaker and some smaller ones at 3-4 ms and 7 ms, 11 ms 12 ms. All those peaks are most probably reflections of nearby surfaces. See the discussion between the two gentlemen posted before your reply 😉.
I'm going to sleep now, much to late already but knowing the speed of sound, and the milliseconds at which you see the spikes you can translate that into distances. If you measured at or in front of the couch I kind of expect the wiggles at 7 ms to be from the wall behind it. The ones at 3-4 ms are somewhere closer to the speakers. So try to find these spots by looking (and imagining) where the sound could bounce back to the listening position. (and arrive 3 ms later or 7 ms, get the drift?)
Tackle one at the time and confirm with a new measurement. Get the big and early ones first! You could also try and move the speaker to see what changes and figure out why.
It's a learning process but you have to be prepared to do some work. Are you 🙂
I'll tell you, it is rewarding.
Thanks GM and wesayso.
The room setup is still as in the first post, only that the bay windows have been "dressed" with curtains, and I hung a thick blanket behind the couch. Also, the coffee table is covered with a doubled-up table cloth.
I measure (and sit) about 3 feet, or a meter, in front of the back wall. The couch is really deep, and I usually sit on the edge rather than "kicking back".
Damping the piano is not really an option...
I was going to put the bass traps behind the speakers, so that should help then. They will be 7 feet tall, 6 inches deep frames filled with "jeans wool" (R19 insulation). I was planning on mounting them off the wall to leave a 2" air gap. They are still not finished (waiting for a staple gun), but once they are, I will experiment with placing them behind the speakers, and possibly in some different locations, and measure.
A separate subwoofer is something I am thinking about. I have that idea to get a pair of AE TD12S and try these as subwoofers, and maybe pair them with a mid/high range speaker at a later point as a complete speaker... but I will try room acoustic treatments first.
I am prepared to do the work... but it's always a compromise between great acoustics and visually and practically acceptable solutions.
Any thoughts about ceiling clouds? I could probably pull that off with WAF points 🙂
The room setup is still as in the first post, only that the bay windows have been "dressed" with curtains, and I hung a thick blanket behind the couch. Also, the coffee table is covered with a doubled-up table cloth.
I measure (and sit) about 3 feet, or a meter, in front of the back wall. The couch is really deep, and I usually sit on the edge rather than "kicking back".
Damping the piano is not really an option...
I was going to put the bass traps behind the speakers, so that should help then. They will be 7 feet tall, 6 inches deep frames filled with "jeans wool" (R19 insulation). I was planning on mounting them off the wall to leave a 2" air gap. They are still not finished (waiting for a staple gun), but once they are, I will experiment with placing them behind the speakers, and possibly in some different locations, and measure.
A separate subwoofer is something I am thinking about. I have that idea to get a pair of AE TD12S and try these as subwoofers, and maybe pair them with a mid/high range speaker at a later point as a complete speaker... but I will try room acoustic treatments first.
I am prepared to do the work... but it's always a compromise between great acoustics and visually and practically acceptable solutions.
Any thoughts about ceiling clouds? I could probably pull that off with WAF points 🙂
You're welcome!
By 'isolating the piano', I meant making a [decorative] false wall absorber between it and the speaker and in retrospect, unless the speakers are toe'd in at > 12 deg/ea., then damping all parallel surfaces just enough to absorb any 'slap echo' reflections between them and the speaker might be beneficial.
Anytime one can damp one of two parallel walls is a good plan, especially if it can be enough to quell its fundamental eigenmodes [~SoS/2/distance], though the latter normally requires room walls to be tapered at >12 deg included same as a cinema, etc..
Not sure what you mean by ceiling clouds, but I assume you're talking about ceiling 'drapes' or suspended built up panel absorbers, neither of which I've done. The only 'full tilt boogie' acoustic room I've done was an LEDE HT where the entire tapered side walls were bass traps of a sort with sloping ceiling and the speakers built in flush, so no additional acoustical treatment was deemed necessary, though was carpeted with a plush pile over 3/4" jute padding.
GM
By 'isolating the piano', I meant making a [decorative] false wall absorber between it and the speaker and in retrospect, unless the speakers are toe'd in at > 12 deg/ea., then damping all parallel surfaces just enough to absorb any 'slap echo' reflections between them and the speaker might be beneficial.
Anytime one can damp one of two parallel walls is a good plan, especially if it can be enough to quell its fundamental eigenmodes [~SoS/2/distance], though the latter normally requires room walls to be tapered at >12 deg included same as a cinema, etc..
Not sure what you mean by ceiling clouds, but I assume you're talking about ceiling 'drapes' or suspended built up panel absorbers, neither of which I've done. The only 'full tilt boogie' acoustic room I've done was an LEDE HT where the entire tapered side walls were bass traps of a sort with sloping ceiling and the speakers built in flush, so no additional acoustical treatment was deemed necessary, though was carpeted with a plush pile over 3/4" jute padding.
GM
I have to ask, are we looking at a stereo pair sweep or was it a single speaker measurement. If it is a stereo pair I'd like to see the separate left and right response. If it was one speaker I'd like to see the other as well 🙂.
If you turn of Envelope (ETC), is the main peak downwards? Can't see it clearly but I think your response might be inverted.
So this measurement was with damping in those several places? To educate yourself, take measurements, one by one removing one element of damping at the time. Look for differences in this plot compared to the original. That way you'd learn what panel or blanket does what peak in the impulse. Then you can decide what a good way of damping would be. For instance, if removing the blanket behind the couch makes one or more peaks bigger you could try 2" or more insulation material instead of the blanket. And maybe even one with the blanket and isolation material. That's how I learned what my room was doing.
If you turn of Envelope (ETC), is the main peak downwards? Can't see it clearly but I think your response might be inverted.
So this measurement was with damping in those several places? To educate yourself, take measurements, one by one removing one element of damping at the time. Look for differences in this plot compared to the original. That way you'd learn what panel or blanket does what peak in the impulse. Then you can decide what a good way of damping would be. For instance, if removing the blanket behind the couch makes one or more peaks bigger you could try 2" or more insulation material instead of the blanket. And maybe even one with the blanket and isolation material. That's how I learned what my room was doing.
Will do. This weekend I should have some time to play around. The bass traps are not done yet, so the measurements were done without them, but with the blanket behind the couch, the curtains etc.
And it's a stereo measurement. I will post some left/right measurements later too, and with the ETC off. I think I am getting a pretty good idea of what to look for now, so I can experiment. Since the bass traps are not done, I was thinking of just bringing the big roll of wool insulation up and just try to see what happens when I place it in different locations. It's inside a plastic back, so not so good for HF, but it should have an impact on LF.
@GM:
I see now what you mean with the piano. I will experiment with damping around the piano. The speakers are toe'd in a bit, maybe 10°.
By ceiling cloud, I meant something like this:
And it's a stereo measurement. I will post some left/right measurements later too, and with the ETC off. I think I am getting a pretty good idea of what to look for now, so I can experiment. Since the bass traps are not done, I was thinking of just bringing the big roll of wool insulation up and just try to see what happens when I place it in different locations. It's inside a plastic back, so not so good for HF, but it should have an impact on LF.
@GM:
I see now what you mean with the piano. I will experiment with damping around the piano. The speakers are toe'd in a bit, maybe 10°.
By ceiling cloud, I meant something like this:
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
- Status
- Not open for further replies.