REW results and possible acoustic treatment

gpt

Established Member
Joined
Nov 28, 2011
Messages
649
Reaction score
109
Points
158
Age
39
Hi All,

Obviously im here because I enjoy AV as much as the rest of us. Unfortunately im often guilty of trying to do too much in the limited space I have. My room is small (4m x 3.2m) and is used as a combination of a second sitting room and cinema room!

Hopefully these guys dont mind me tagging them too. Weve not spoke before but Ive read your material around the forums and hope you can help @MI55ION @Conrad. Apologies for bothering you if youre busy! :smashin:

Last December I got rid of my towers when the room was re decorated and purchased the XTZ cinema series for wall mounting (very wife friendly!). After missing the bass that came from my towers I took to DIY Subs, and made a pair of rather large beasts (not wife friendly!). In all honesty, size wise, they are too big for the room - but were getting used to them and they do dig deep – so hopefully they are staying.

Now on to the reason for the thread. Since doing the DIY subwoofers, Ive learnt a lot about REW. Whilst I can EQ my subwoofers significantly to get a nice curve – masking many problems – I am unable to now look away from the results of the speakers which Audyssey doesn’t seem to be able to fix. Eq’ing the subwoofers was also particularly difficult likely due to problems in the room, including the size. So this morning I have gone about modelling the room in sketchup, including some potential rooms treatment (corner bass traps and rear absorption panels)

I was hoping to get some thoughts and insight into what I can potentially do before I start building these things and if anyone can guide me on what improvements I might expect or If im simply chasing a white rabbit. I have attached the images below

One of my main concerns is my corner bass traps are not large enough but as you can see I do not have much room to go much bigger. they are 300mm out from the corner in both directions. From the corner out it is 212mm (8.5"). Absorption panels I was hoping to do 50mm deep with 50mm air gap behind. Also my rear corner traps would have to be cut around my side speakers.

So, here are my REW results with no EQ from my amp or audyssey, followed by images of my room (i have attached the REW file for anyone that wants to look deeper, admittedly I am still very much the beginner)

Also - Im very much aware this is not an optimal room, its part of the house too. Im just trying to do my best with what im working with :thumbsup: The room is rectangular. All speakers are mounted symetrically with tweeters at ear level. The separation between the fronts is quite wide due to having a motorised projector screen. Surrounds are at 90 degree to the listening position. Fronts are toed towards MLP

1599818766879.png

Red with EQ & Audyssey - Blue without


Key for sketchup images



Black – Speakers & Subs

Green – AV shelving

Yellow – Potential room treatment

Red – My curtains and curtain poles!



Windows and doors hopefully obvious. My room is plasterboard on top of concrete block for 3 walls (2 external, 1 old external). The wall with the double doors is a stud wall filled with insulation that goes to a kitchen. Were also in a bungalow, so concrete floor. Flooring is carpet. One big L shape sofa also dominates the room

1599817851906.png


1599817993531.png

1599818018498.png

1599818055484.png

1599818187828.png

1599818260426.png
 

Attachments

  • Forums Rew.zip
    3.5 MB · Views: 66
I'll have a look at this tonight in more detail and when I'm on a machine that I can use REW on.

First things:
- bass traps are unlikely to have any impact below 100Hz. You need significantly thicker traps to be able to affect very low frequencies. @Ultrasonic did some good tests, up to about 1m or so.
- Panels will help with frequency response, but they're mostly used to treat decay times. The two do go hand-in-hand though, improve one and you'll improve the other - it's all sound energy at the end of the day. In my room (rectangular, 2.4m x 3.7m, empty walls) I was getting slap echo and my decay times were up over 500ms. Adding panels got that down to <200ms and really helped clear up the clarity. They didn't do much for my response though. Again, when I can see REW I'll be able to comment more on this.
- A common place to put the panels is at first reflection points. It looks like your first reflections points are glass, on both sides. This is likely causing issues. Any chance you can get panels on the sides? Because of my room layout I actually went horizontally with my panels, that then covers the first reflection points of both main speakers from both seating positions, something to think about.
- Consider diffusion at the back, rather than absorption. Over absorption can create a dead and lifeless sound. To me that sounds clean and crisp, but lots of people like their surrounds to sound more diffuse. Especially in a 5.n system it might help give more of a sense of space at the back. That's what my calibrator told me at least. GIK is a good place to start for diffusion panels if you don't want to DIY.

I don't know how much panels will cost you to make but I used Mafia Panels and they're excellent. I think I paid less than £400 for 7 panels - 5 x 100mm and 2 x 50mm thick.

For the sub response have a look at room sim, see if you can get a better native response with placement before you EQ. I'll have a look and see if you can do anything with delays, assuming the correct sweeps are in the MDAT. What I usually look for is:
- L with no sub but with the crossover
- R with no sub but with the crossover
- Sub1 on main channel, no main, but with the crossover
- Sub2 on main channel, no main, but with the crossover
All with a consistent timing reference.

If you have multiple positions the same measurement in each potential position would be great.
 
Thanks Conrad

RE the treatment, I agree the first and secondary reflection points are usually the places to treat first. Whilst I could potentially add treatment over the doors, I cant do anything over the window. Adding treatment on just one side im assuming would create a further imbalance. Whilst I don’t know how to completely understand and interpret REW, the XTZ I think are supposed to have a good off axis response? Therefore I assumed, some of the first reflections can sometimes be beneficial? It’s the reflections going across the room, e.g. Left speaker, to right wall to MLP that are likely to be the most problematic

I like watching this guy on youtube -



He does 4 episodes around room treatment, its all very useful information and probably prompted me to investigate things further.

I can EQ the stuff below 100hz on the behringer6000d but I rely on audyssey for the rest and I need to get closer to the target line first I think to give audyssey a chance of being able to do its job. Audyssey correction doesn’t solve the problem. I also wanted to attack the corner first since presumably the Bass frequencies can cause further problems in the upper frequencies, so it made sense to start there first. But as I previously have shown, ive got such limited space to do that. I can make bigger corner traps if I didn’t have to go from floor to ceiling, but I am not sure how beneficial that is? Or how wife friendly I can make them :rolleyes: The rear wall is obviously the easiest to treat for me.

I do have a REW file for all of my possible sub positions, Ive attached it to this post. I havent got readings for the position next to the couch (back right corner) although I tried to get the sub to work there a lot but it just didnt happen. The 8 positions are various positions from the left corner (where the AV shelving currently is) across to the right corner where the sub is currently. So I used positions 4 & 8, or graphs 4 & 8 as my basis to start. The sub on the left (pos 4) is set at -2db less than the other sub with a delay of 0.17 which audyssey had detected previously - I just put that into the behringer before i ran audyssey the next time. All sweeps should be just sub as I set the crossover high but ive noticed sometimes although I do the changes on the marantz, the change in crossover frequency doesnt always seem to happen. I have to go back or turn things off and on etc, its a pain! I also havent got banana plugs so unplugging things can be a further pain but naturally im happy to do any sweeps you need :thumbsup:

Really appreciate your input!

**EDIT - Looking at them rew readings it looks like the crossovers were active actually. I think they would have been at 120hz.
 

Attachments

  • Pos1-8.zip
    467.8 KB · Views: 65
I had a little look.

- You don't have a big ringing issue. Your decay times are hitting 50dB at ~300ms which isn't dreadful. Panels will help clear that up down to <200 I would guess. Mine were over 500ms and panels brought it down to 180ms across the board.
- Your response looks pretty good. I got slightly better results using positions 4 and 6 with a ~7ms delay, but it's negligible.

Here's a link to an MDAT showing the difference in decay times with and without the panels. Note that I don't think the response changes are as a result of the panels, that's other EQ and changes. The measurements were taken on different days. The decay is the panels though. Panel Comparison.mdat

I'm surprised that you're not getting more room gain, especially with the low shelf filter.

Red looks pretty good as an EQd response. I don't think anything you're planning will reduce the 40Hz ringing, but that's a lot less important than higher up.

Overall I don't see anything wrong with your response that basic EQ wouldn't manage. Small rooms can be more problematic than large. Some treatment will almost certainly tighten things up and give you some improved clarity. I don't know how much impact what you're planning will have on your response. You could try and model it in REW and increase the absorption figures against some of the walls?

For me, I just overdid it and went with as much as I could fit in thinking more was better. Apparently it's not but I think my room sounds great!
 
Thanks Conrad I appreciate you taking the time to look

My room doesnt sound bad by any stretch, I just want to know im getting the best out of what ive got

1599853629792.png

Thats my 100-500hz frequency range and you see the same sequence repeating itself all of the way up the frequency range so I was hoping I could knock kit on the head and tighten it up a bit. I figured Ive got to be getting reflections off of the rear wall as I sit so close to it. Obviously I only gave you the positions for the subs for now so I might try and get some readings for the LCR

Also Is it just co incidental having the dip at my crossover frequency or is that something I can improve?

1599854009254.png

This is the 500-1000 range

I know people talk about using a separate amp to really make the speakers shine. Im wondering how much of a difference that would make too. The speakers already get to a good volume, does the extra juice help in other ways?
 
A dip at the crossover isn't uncommon, at least in my experience. It's to do with the integration between subs and mains. If you want me to take a look at that I'll need the subs with crossover, and the mains with crossover, with the same timing reference on both.

You'll always get reflections from the back wall regardless of how far away from it you sit, the difference will be the affected frequencies.

Adding the panels might well improve your response but I don't know how to say for sure. It most probably won't make things any worse. You can always take them out if it measures worse or you like the sound less.
 
Hi gpt, I was notified of your post earlier today but was busy so didn't get an opportunity to reply.

It seems Conrad has pretty much covered most of it, not sure I can add much.

Nice to see a fellow forum member take it to the level that you are. You wouldn't believe the amount of people I see spend thousands on quality kit whilst paying no attention to the room itself.

The corner traps should help smooth out some of the peaks above 100hz whilst absorb some of that unspent energy. In fact I think that is your big challenge, to absorb, diffuse and dissipate all that energy.

Firstly thats a nice sub response for a small room. Seeing your diagrams I think you've got a good idea of the concept of it all. Due to space constraints you have an almost nearfield setup. There is good symmetry, curtains over both the door and windows will help as well as a rug on the floor if it isn't already carpeted. If it were possible I would say bring the sofa out a little but I see that may not be practical. Ass well as the rear wall, I would add some absorption to the front wall too. Some diffusion panels to the ceiling would be the cherry on top.

You may not see night and day difference in the frequency response in that 100 to 1khz range but it should significantly help with decay times.

As for separate amps, it isn't really about extra juice although I'm sure it helps with headroom. Its more about separating the pre and power sections for lower noise floor, dynamics, channel separation, etc so yes, if you can definitely look into dedicated processor and power amps (ideally mono blocks for LCR). The other advantage is as technology changes, you get to keep the power amps forever only updating the processor as and when required.
 
Curtains and rugs are a great call.

And yeah, power amps as MI55ION says, it's about separation. Also, even if you only get a stereo amp for the front LR you'll be asking less of your amp to drive the remaining channels which means more power to those channels, less distortion, etc. I'd try and do the front three though as the center is the most important channel.
 
I've not understood much of what has gone on here in this thread but that sub curve looks fantastic! I'd love to hear that going! Super low, sexy bass... 😁
 
Thanks Mi55ion, again appreciate you taking the time to repond! :thumbsup:

The corner traps should help smooth out some of the peaks above 100hz whilst absorb some of that unspent energy. In fact I think that is your big challenge, to absorb, diffuse and dissipate all that energy.

What did you think of the sizes of the corner traps I proposed, I noticed with GIK corner traps they traverrse a lot more of the corner than I proposed. Any idea if what Ive proposed would still be worthwhile?

1599892940481.png


This is the bit that really bothers me, Its a huge peak at 52hz (taken from an un EQ'd response), then it drops off down to 80hz (a near 15db dip), then another big peak again at 99hz. The 99hz part is particularly problematic. I could put my PEQ at -15 and it still wouldnt reduce any more than it has.

1599893089924.png

They are the filters I had to apply to knock it out of the frequency range up to 100hz. But the same pattern finds its way across the entire range.

1599893181619.png

1599893214771.png


I always assumed it all stemmed from the lower frequencies which is why I was keen to treat the corners but as Conrad has said the decay times are quite reasonable Im just wondering if im looking at this all wrong

Firstly thats a nice sub response for a small room. Seeing your diagrams I think you've got a good idea of the concept of it all. Due to space constraints you have an almost nearfield setup. There is good symmetry, curtains over both the door and windows will help as well as a rug on the floor if it isn't already carpeted. If it were possible I would say bring the sofa out a little but I see that may not be practical. Ass well as the rear wall, I would add some absorption to the front wall too. Some diffusion panels to the ceiling would be the cherry on top.

Yes symmetry is perfect with tweeters dead on ear level. Room is fully carpeted with a thick underlay. Curtains I never thought really do much as they have little absorption but I was considering a wooden slatted blind over the window to help with dissipation


You may not see night and day difference in the frequency response in that 100 to 1khz range but it should significantly help with decay times.

But again, Conrad has found the decay times to be reasonable, so im not sure Im going to get the flatter line that I was looking for - or even if that flatter line would be detectable to the human ear anyway! - Could just be chasing a graph :facepalm:

And yeah, power amps as MI55ION says, it's about separation. Also, even if you only get a stereo amp for the front LR you'll be asking less of your amp to drive the remaining channels which means more power to those channels, less distortion, etc. I'd try and do the front three though as the center is the most important channel.

This is part of what I was wondering about. The Marantz is rated up to 200w per channel (I know its probably giving no where near that). My front 3 are rated up to 300w and are 4ohm. Since the Marantz can give up to 200w at 8ohm, it really shouldnt struggle to drive the XTZ speakers? The distortion is also measurable in REW isnt it?

I just dont want to spend a load and not really see the benefit. Have people done measurements before and after adding stereo amps?


Also since you mentioned the centre channel being important. I have the marantz app and looked into ratbudyssey for applying my own target curves. Its not ultra clear, but you can view the audyssey responses for each channel. I thought I might as well pop them in the thread for the front 3 speakers at least.

1599893976559.png

1599894044385.png

1599894103400.png

Also that red line is not my target curve ;)

I've not understood much of what has gone on here in this thread but that sub curve looks fantastic! I'd love to hear that going! Super low, sexy bass... 😁

Aquaman has been the best film on it so far. If youre ever near Crewe youre free to stop by :smashin:
 
Thanks Mi55ion, again appreciate you taking the time to repond! :thumbsup:



What did you think of the sizes of the corner traps I proposed, I noticed with GIK corner traps they traverrse a lot more of the corner than I proposed. Any idea if what Ive proposed would still be worthwhile?

View attachment 1364280

This is the bit that really bothers me, Its a huge peak at 52hz (taken from an un EQ'd response), then it drops off down to 80hz (a near 15db dip), then another big peak again at 99hz. The 99hz part is particularly problematic. I could put my PEQ at -15 and it still wouldnt reduce any more than it has.

View attachment 1364281
They are the filters I had to apply to knock it out of the frequency range up to 100hz. But the same pattern finds its way across the entire range.

View attachment 1364282
View attachment 1364283

I always assumed it all stemmed from the lower frequencies which is why I was keen to treat the corners but as Conrad has said the decay times are quite reasonable Im just wondering if im looking at this all wrong



Yes symmetry is perfect with tweeters dead on ear level. Room is fully carpeted with a thick underlay. Curtains I never thought really do much as they have little absorption but I was considering a wooden slatted blind over the window to help with dissipation




But again, Conrad has found the decay times to be reasonable, so im not sure Im going to get the flatter line that I was looking for - or even if that flatter line would be detectable to the human ear anyway! - Could just be chasing a graph :facepalm:



This is part of what I was wondering about. The Marantz is rated up to 200w per channel (I know its probably giving no where near that). My front 3 are rated up to 300w and are 4ohm. Since the Marantz can give up to 200w at 8ohm, it really shouldnt struggle to drive the XTZ speakers? The distortion is also measurable in REW isnt it?

I just dont want to spend a load and not really see the benefit. Have people done measurements before and after adding stereo amps?


Also since you mentioned the centre channel being important. I have the marantz app and looked into ratbudyssey for applying my own target curves. Its not ultra clear, but you can view the audyssey responses for each channel. I thought I might as well pop them in the thread for the front 3 speakers at least.

View attachment 1364285
View attachment 1364286
View attachment 1364287
Also that red line is not my target curve ;)



Aquaman has been the best film on it so far. If youre ever near Crewe youre free to stop by :smashin:
Cracking! Well you are only 200 miles away so if I'm in the neighbourhood.. 😂
Do you have a dedicated thread anywhere about your subs?
 
Cracking! Well you are only 200 miles away so if I'm in the neighbourhood.. 😂
Do you have a dedicated thread anywhere about your subs?


Only 200? 😋


I posted this thread not long after i got them. I designed them with a lot of help from Liam who i tagged in the thread.

It's been about perseverance since then, a bit like this thread! Keep trying to identify problems and keep digging.ive gotten a little more understanding since but I still no nothing about decay, distortion or how to interpret waterfall graphs, all things i would like to be able to do!

The subs have definitely been a big improvement though. I've spent about £1k on the pair which includes the amp and UMIK mic. I also had the subs built for me so had I done them myself i could have saved more. I am curious what I would achieved had i chosen sealed subs that were much smaller in foot print but that will be just a little fun project for the future.

Room treatment
DIY Atmos speakers
Stereo amp

Are probably my next improvements in that order 🙂
 
I'm very surprised that dual 18s in a room of that size are only getting you into the high teens. I have 15" subs in a room slightly smaller than yours and I'm getting well into single digits. Maybe sealed would be something to look at, if depth is something you feel needs improving.

I'm not convinced that the repeating pattern you're seeing is anything other than a regular measurement. Given that they're not octave spaced (doubling of frequency) but they are at different wavelengths, I don't know that they're really "repeating". The frequencies might be room related (multiples of your room dimensions, or multiples of distances between surfaces and the mic), but given the indirect nature of sound and reflections, it's very hard to model.

One thing you could do is build/buy one panel and try it. The nice thing about panels is that they'll have an effect on their own. More will have more of an effect, but they sum, they're not binary. You don't need 10 panels and then all of a sudden you see a difference where you didn't with nine, if that makes sense.

You could also try moving a speaker. I know that's not practical for long term, but just to test. Measure your left where it is now, then move it in a foot, measure again, then move it out a foot and measure again. At least that'll tell you whether the response changes. My guess it will, but your "problems" will move, not disappear.

That 100Hz spike on the centre is almost certainly affecting your speech clarity. I would tame that with EQ.

One other thing I like to do, just to see what's possible, is to give REW free reign to do what it wants. I choose a problem area, in your case it's about 300-1000Hz that we're looking at, set the level, set the EQ to generic so everything is available, set the max boosts to 18, the max individual boost to 12 and the flatness to 2 or 3 and see what it comes up with. I'd wouldn't recommend applying those filters, it'll probably sound terrible, but it gives you an idea of the kind of changes that need to be made to get to flat.

I've actually tested this by generating those filters in 500Hz batches and applying all of them using JRiver which has near infinite PEQ. It took a long time but I wanted to hear what ruler flat sounds like. It sounds dreadful. Metallic and highly digitally processed plus the speakers make some terrible noises. It measures great though!

I would worry less about peaks and dips further up the frequency range. They're less and less audible the further up you go. This is because they get narrower and is why variable smoothing exists.

If it helps at all (and you may not care at all!) this is my left channel:
1599899663052.png

As you can see, I have "repeating patterns" and a 15dB swing through that same 100-1000Hz range. I also have a hole around my crossover point that I just cannot fix. And that's in a heavily treated room. I have very limited placement options though. The conclusion I would draw is that treatment might not be your answer, you might have to just live with a slightly lumpy response. It's already better than 80/90% of people.
 
I'm very surprised that dual 18s in a room of that size are only getting you into the high teens. I have 15" subs in a room slightly smaller than yours and I'm getting well into single digits. Maybe sealed would be something to look at, if depth is something you feel needs improving.

When modelled in REW the sub would have gone louder had I increased the volume of the sub cabinet but I just picked the maximum size i was prepared to live with. I modelled 15" drivers in the same cabinet and the 18s still outputted more and I believed would have been better. It also wont help that i tuned it to 21hz thinking the extra SPL in the mid range would be useful :oops: live and learn

I'm not convinced that the repeating pattern you're seeing is anything other than a regular measurement. Given that they're not octave spaced (doubling of frequency) but they are at different wavelengths, I don't know that they're really "repeating". The frequencies might be room related (multiples of your room dimensions, or multiples of distances between surfaces and the mic), but given the indirect nature of sound and reflections, it's very hard to model.

Thats good to know! :) I didnt know if it was problem repeating itself along the line or just completely normal. If its normal thats great

One thing you could do is build/buy one panel and try it. The nice thing about panels is that they'll have an effect on their own. More will have more of an effect, but they sum, they're not binary. You don't need 10 panels and then all of a sudden you see a difference where you didn't with nine, if that makes sense.

Yeah im certainly thinking about giving the rears a go and putting a wooden slatted blind on the window

You could also try moving a speaker. I know that's not practical for long term, but just to test. Measure your left where it is now, then move it in a foot, measure again, then move it out a foot and measure again. At least that'll tell you whether the response changes. My guess it will, but your "problems" will move, not disappear.

This is actually harder for me to do since the cables in the wall and ive only got about a foot of cable to play with. Ive also got a motorised screen that drops down so long term I cant move the speakers in any further anyway. One thing that surprised me was I originally had the speakers flush wall mounted but they were aiming a fair way off the MLP. So I built a little 22 degree wedge (the xtz have their own 7.5 degree toe in on top of that) so that i was more on axis expecting maybe a little more clarity - I honestly cant really tell Ive done anything. Im going to keep them on axis because I know it has to be another improvement and step in the right direction, but in a blind hearing I couldnt have told which was which

That 100Hz spike on the centre is almost certainly affecting your speech clarity. I would tame that with EQ.

One other thing I like to do, just to see what's possible, is to give REW free reign to do what it wants. I choose a problem area, in your case it's about 300-1000Hz that we're looking at, set the level, set the EQ to generic so everything is available, set the max boosts to 18, the max individual boost to 12 and the flatness to 2 or 3 and see what it comes up with. I'd wouldn't recommend applying those filters, it'll probably sound terrible, but it gives you an idea of the kind of changes that need to be made to get to flat.

I've actually tested this by generating those filters in 500Hz batches and applying all of them using JRiver which has near infinite PEQ. It took a long time but I wanted to hear what ruler flat sounds like. It sounds dreadful. Metallic and highly digitally processed plus the speakers make some terrible noises. It measures great though!

Ive only got audyssey to EQ anything above the subs. Im using the dsp on the behringer powering the subs to EQ the lowest frequecies. The black and white ratbudyssey screen caps were what they audyssey mic recorded before it applied its EQ, so I dont know to what level it suceeded actually. I just wanted to tighten things up, giving audyssey less to do and more chance of success. I do think speech clarity is one area where my room suffers though. One of the big subs is very close to that centre channel (another reflection point) - but given I sit right in front of the centre channel and only 2.7 ish meters away, i figured the rear wall may be the most likely culprit.

Like you said, I may just make a few panels and try it. I have previously tried sticking a load of heavy feather cushions on the back wall but again, no change. But as we know different materials have different absorptions plus air gaps behind help. I think I just need to give it a go

If it helps at all (and you may not care at all!) this is my left channel:
1599899663052.png

As you can see, I have "repeating patterns" and a 15dB swing through that same 100-1000Hz range. I also have a hole around my crossover point that I just cannot fix. And that's in a heavily treated room. I have very limited placement options though. The conclusion I would draw is that treatment might not be your answer, you might have to just live with a slightly lumpy response. It's already better than 80/90% of people.

Absolutely does help! Its nice to see some sweeps from other people. Makes mine feel a bit more normal after all.

I think sometimes people spend a few grand on equipment and state there speakers are spectacular, justifying the amount of money spent. I can say going from my kef setup (which I sold for £800) to the XTZ setup (which cost me £2200) - the main difference ive seen is a more neutral sound (which has taken some getting used to - and i wouldnt necessarily say is better) and more separation and detail in each speaker (which I do like, the S5 surrounds are great). But I wouldnt say its £1500 difference to my ears. This is why I assume my room has more to do with it when I see the other great reviews out there of the XTZ cinema series, so im just trying to get the most out of my investment as possible.
 
I reckon that Center mode is due to the symmetry of the speaker/room/listening position. Your indirect sound is all positively summing at the listening position. I think panels at the sides would help but curtains and blinds might do too.

I’d be tempted to pick up some random speaker cable and measure the speaker in other positions. It’ll at least give you some insight into how moving the speaker changes the response which will give you a clue as to where your dips are being caused. If that’s a real pain though it might not be worth it.

Trying to get the most out of your system is a noble goal. Panels will help, more/better power and a separation of duties will likely lower distortion clearing things up another notch. It’s all the right thing to do, the decision comes on cost/reward and how liveable you need the room to be.
 
I reckon that Center mode is due to the symmetry of the speaker/room/listening position. Your indirect sound is all positively summing at the listening position. I think panels at the sides would help but curtains and blinds might do too

Don't forget the centre reading from audyseey will be just the centre speaker on its own though, and that shows the same peak
 
1599910072018.png


Just twigged ive got 6 measurement position with audyssey so I can compare my centre channel across those positions. For just the Centre - as you come forward into the room (and closer to the speaker) some of the peaks actually increase, which I wasnt expecting when I just opened it up. The green line is about 50cm forward (closer to centre and away from the rear wall) than the black line. Not 100% if its just an increase in SPL for being closer

1599910343820.png

Left graph is centre speaker left of MLP and then a bit forward, right graph is right of mlp and forward

I dont think this actually tells me anything now that ive wrote all of this :laugh:

Will have to check what Audyssey achieved just for centre
 
Thanks Mi55ion, again appreciate you taking the time to repond! :thumbsup:



What did you think of the sizes of the corner traps I proposed, I noticed with GIK corner traps they traverrse a lot more of the corner than I proposed. Any idea if what Ive proposed would still be worthwhile?

View attachment 1364280

This is the bit that really bothers me, Its a huge peak at 52hz (taken from an un EQ'd response), then it drops off down to 80hz (a near 15db dip), then another big peak again at 99hz. The 99hz part is particularly problematic. I could put my PEQ at -15 and it still wouldnt reduce any more than it has.

View attachment 1364281
They are the filters I had to apply to knock it out of the frequency range up to 100hz. But the same pattern finds its way across the entire range.

View attachment 1364282
View attachment 1364283

I always assumed it all stemmed from the lower frequencies which is why I was keen to treat the corners but as Conrad has said the decay times are quite reasonable Im just wondering if im looking at this all wrong



Yes symmetry is perfect with tweeters dead on ear level. Room is fully carpeted with a thick underlay. Curtains I never thought really do much as they have little absorption but I was considering a wooden slatted blind over the window to help with dissipation




But again, Conrad has found the decay times to be reasonable, so im not sure Im going to get the flatter line that I was looking for - or even if that flatter line would be detectable to the human ear anyway! - Could just be chasing a graph :facepalm:



This is part of what I was wondering about. The Marantz is rated up to 200w per channel (I know its probably giving no where near that). My front 3 are rated up to 300w and are 4ohm. Since the Marantz can give up to 200w at 8ohm, it really shouldnt struggle to drive the XTZ speakers? The distortion is also measurable in REW isnt it?

I just dont want to spend a load and not really see the benefit. Have people done measurements before and after adding stereo amps?


Also since you mentioned the centre channel being important. I have the marantz app and looked into ratbudyssey for applying my own target curves. Its not ultra clear, but you can view the audyssey responses for each channel. I thought I might as well pop them in the thread for the front 3 speakers at least.

View attachment 1364285
View attachment 1364286
View attachment 1364287
Also that red line is not my target curve ;)



Aquaman has been the best film on it so far. If youre ever near Crewe youre free to stop by :smashin:


I think the effect of the corner traps, as well as all the rest, will be cumulative. What you propose will be better than having none whatsoever, the deeper you could make them the lower down they’ll have an impact (within about 150hz) but that’s dependant on the constraints you have and only you can decide if it’s worth the effort. For reference, the first 1min of this video shows a clear audible difference with just corner traps and if you continue watching, the subsequent effect of further treatment.




I don’t think you should worry too much about that peak at 99hz, looking at the waterfall graph it doesn’t linger for long. You’ve addressed the main problem at 52hz however after EQ that 20hz region seems rather domineering, probably why the LFE effects of Aquaman seems impressive. Higher up the range, the repeated pattern that you see in your response is probably comb filtering. All of the panels as a whole, including the corner traps, should help in this regard. Depending on the material, thicker curtains can work a little to absorb and diffuse but whatever you decide on, I would ensure that both the window and door sides are covered with similar material to maintain symmetry.

The decay times are pretty good as is but there is room for improvement. I wouldn’t go chasing a ruler flat response as Conrad so eloquently put it, whilst it may look good on a graph, not only is it taxing of your amplifier, it usually sounds terrible as it comes at a loss of dynamic range whilst introducing other digital nasties. Furthermore, I would avoid eq’ing anything above 500hz if your version of Audyssey permits it as in my experience it starts to sound unnatural.
 
Thanks again Mi55ion

I don’t think you should worry too much about that peak at 99hz, looking at the waterfall graph it doesn’t linger for long. You’ve addressed the main problem at 52hz however after EQ that 20hz region seems rather domineering, probably why the LFE effects of Aquaman seems impressive.

Could you elaborate on the 20hz region a little? Other than seeing the line on the chart - my understanding on REW is limited so id love to know what it is I should be looking at.

All of the panels as a whole, including the corner traps, should help in this regard. Depending on the material, thicker curtains can work a little to absorb and diffuse but whatever you decide on, I would ensure that both the window and door sides are covered with similar material to maintain symmetry.

I will have to see what I can do here. I could go for slatted wooden blinds over the window and some form of dissipation panelling on the door. I was worried corner traps only spanning 425mm of the corner and being 212mm deep wouldnt have been effective enough but im pleased to hear theres some merit in it!

The video clearly demonstrates the difference room treatment can make. I noticed the corner traps in the video arent full floor to ceiling, is this something you feels matter? Should i for example pull a subwoofer out and away from the corner to make room for a corner trap, or leave the sub in the corner and start the corner trap from the height of the subwoofer?

I came across this thread the other day

There some good info on there about different materials and absorption rates so I will give that a go (also useful for anyone else who reads this thread

The decay times are pretty good as is but there is room for improvement. I wouldn’t go chasing a ruler flat response as Conrad so eloquently put it, whilst it may look good on a graph, not only is it taxing of your amplifier, it usually sounds terrible as it comes at a loss of dynamic range whilst introducing other digital nasties. Furthermore, I would avoid eq’ing anything above 500hz if your version of Audyssey permits it as in my experience it starts to sound unnatural.

Yes fortunately this is one thing the marantz/denon audyssey app is really useful for and I have played around with this a little already :) - when I get the room sorted I will make a point of not letting audyssey tackle anything above that range! :thumbsup:
 
This is your waterfall graph before EQ:
wf2.jpg


This is your waterfall graph after EQ:
wf1.jpg


It shows the time it takes in milliseconds for the sound to dissipate. As you can see whilst you've dealt with the room mode at 52hz, the energy at 20hz is still considerable and has increased. You could probably massage that at a later time, for now though in relation to acoustic treatment you primary focus is on everything above 100hz. You could possibly reduce that decay time closer to 100ms. It's what gives you that instant "click" or attack, speed, definition or whatever other word audiophiles like to use.

I wouldn't sacrifice positional EQ of the sub to fit a couple of corner traps, just isn't worth it. However if moving the subs give a better response then by all means stick the floor to ceiling corner trap in there, you can have your cake and eat it too.
 
@MI55ION @Conrad. Just one final Thank you from me guys - Appreciate you both chiming in with as much help as you have, invaluable info.

I will get to making some panels over the next few months (things never happen fast around here 😂) and return with results!

:beer:
 
I had understood under 400ms was pretty good, so 300 looks more than acceptable? I don’t think many manage to absorb the low frequency room modes, they just dial them out by EQ and the decay and RT60s are improved but getting below 400/300 here I think is hard ?
 
anything under 100/150Hz is very tough to target with traps so we're largely ignoring those.
300ms is pretty good, but lower is better. 300ms without treatment is likely to come down to <200ms with treatments and will make a decent difference. It's quite strange just how much improvement it makes to the overall clarity of the sound, specifically speech.

The test I use is, when there's a scene and there's a band or music playing in the background (behind the main actors who have dialog), can I make out the words of the background song. Usually I couldn't before I treated, although I was up in the 500ms range. Now I almost always can. Sometimes it depends on the content but it's usually pretty clear.
 
I have spent quite some time EQing with REW and setting PEQ's manually and with 3 subs, learnt quite a lot as I have an awful room acoustically (around 5m square) and I sit a fair bit into the room I have some big bass nulls as well as peaks. I have ordered some panels from GIK to helps a little around ear level first relection points, my EQ has improved a little around the 20Hz mark and I have cut a little further at the 34 Hz main bass peak since these average curves, here is my messy EQ's averaging the response across the seating area (2m wide). My RT60's are well into the 300s and double that at the low end, so hoping for some improvements. But sonically its sounding superb and I have listened to a lot of purpose built set ups !?

Screenshot 2020-09-13 at 19.06.46.png
 
I'd love to get my hands on your MDAT. That sounds wrong.
 

The latest video from AVForums

Is 4K Blu-ray Worth It?
Subscribe to our YouTube channel
Back
Top Bottom