• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Small Room Treatment

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Hello,

I am hoping that those of you with knowledge of room treatment could spare some advice.
I am setting up a system in the room shown in the diagrams below.

ROOM
The room is not large being 470cm L x 340cm W x 330cm H.
It has a brick wall on the left, front, and back, and a plaster wall on the right (backed by tiles on the other side - bathroom).
The ceiling is plaster and above there is the roof cavity.
The floor is suspended timber boards, probably about 50cm over the ground. It has a rug on it.
The front wall has a fireplace with a massive mirror mounted on top of it.
Behind the listening position I have a shelf full of records, and in the corner an Ikea kallax with my gear on it.

SYSTEM
Genelecs 8341 crossed at 70Hz to two KEF KC62 subs
MiniDSP SHD with Dirac live

As you can see I get a fair bit of comb effect. I attach the REW sweeps with the system equalized as best as I could, and smoothed at 1/24.

I have no major complaints subjectively, except that I am struggling to get a nicely stable center phantom image. Mind you this could be due to the fact that currently, the right wall is more reflective than the left (is covered in frames with glass), but this is going to change soon.

TREATMENT OPTIONS
This is a room that I share with family, so it has to remain fairly pretty.
Budget not much. I found these pretty cheap and decent-looking panels. I was thinking of perhaps taking care of the ER points on the side walls with absorbing/skyline panels. The ceiling is high at 330cm and the floor is covered with a rug.

Thoughts?

Thanks in advance
 

Attachments

  • room plan.jpg
    room plan.jpg
    95.4 KB · Views: 230
  • 3d room.jpg
    3d room.jpg
    187.5 KB · Views: 257
  • room response EQD.jpg
    room response EQD.jpg
    101.7 KB · Views: 258
  • room ETC.jpg
    room ETC.jpg
    98.8 KB · Views: 198
  • room RT60.jpg
    room RT60.jpg
    109.4 KB · Views: 182
  • room waterfall.jpg
    room waterfall.jpg
    152.3 KB · Views: 181
  • room spectro 10 Hz.jpg
    room spectro 10 Hz.jpg
    111 KB · Views: 181
  • room clarity.jpg
    room clarity.jpg
    96.4 KB · Views: 165
  • ETC dbfs.jpg
    ETC dbfs.jpg
    126.6 KB · Views: 151
  • impulse L and R overlaid.jpg
    impulse L and R overlaid.jpg
    66 KB · Views: 132
  • ETC L and R overlaid.jpg
    ETC L and R overlaid.jpg
    92.8 KB · Views: 134
  • LR FR.jpg
    LR FR.jpg
    106.9 KB · Views: 160
Last edited:

DJBonoBobo

Major Contributor
Joined
Jan 21, 2020
Messages
1,360
Likes
2,851
Location
any germ
Maybe some of this will help?

 

DJBonoBobo

Major Contributor
Joined
Jan 21, 2020
Messages
1,360
Likes
2,851
Location
any germ
BTW, i think your FR looks almost perfect already - not worse than mine after all the treatments, as far as i can see. Spectrogram also looks pretty smooth - better than mine actually, i think (less dry).
Maybe start by looking at the ETC graph?
 

chych7

Active Member
Joined
Aug 28, 2020
Messages
276
Likes
422
Overall your room actually looks pretty decent. Can you plot the spectrogram down to 10 Hz? That will reveal some more of the room modes; from RT60 it looks like there could be some benefit in putting some bass traps (at room corner primarily); the panels in the link you sent are not good enough for low frequency bass trapping. Can you also show the REW clarity plots? It will give some information about direct sound vs reflected sound, which can give insight into imaging/phantom image. Putting some of these panels strategically at reflection points can improve this for mid/high freqs, covering the vocal range.
 
D

Deleted member 46664

Guest
Overall your room actually looks pretty decent.

Yeah ... I was just sitting here thinking the graphs for my room should look that good. ;)

Suggestions... try moving the speakers back toward the fireplace a bit, maybe add a nice area rug in front of your couch, put a couple of bookshelves on the sides to scatter reflections ...
 

alex-z

Addicted to Fun and Learning
Joined
Feb 19, 2021
Messages
910
Likes
1,684
Location
Canada
Typical of most rooms, your 100-300Hz is the messiest due to room modes. I would fill the front 2 corners with bass traps, then put diffusion on the ceiling and both side walls to start.

Normally I would say more absorption instead of diffusion, but your decay times above 200Hz are nicely consistent, and you don't want to mess with a good thing.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
Typical of most rooms, your 100-300Hz is the messiest due to room modes. I would fill the front 2 corners with bass traps, then put diffusion on the ceiling and both side walls to start.

Normally I would say more absorption instead of diffusion, but your decay times above 200Hz are nicely consistent, and you don't want to mess with a good thing.
Man do I love this advice! In small rooms, use only what you need when it comes to absorption. The corners are an excellent place to start.
 

chych7

Active Member
Joined
Aug 28, 2020
Messages
276
Likes
422
Be careful on the diffusion as well. Diffusion on sidewalls will randomize phase of the early reflected soundwave. When it hits your ears (after the primary wave), it can be interpreted as degradation in clarity. It may be better to either completely absorb the primary reflection (with a good broadband absorber), or not absorb it at all. For speakers with good dispersion (like Genelec), this would apply, but if the sidewalls themselves are different, it could skew the sound. Also keep in mind the floor/ceiling; those reflected waves can hit the ears at a different time than the sidewall reflection. It may be better to absorb the floor/ceiling but leave the sidewalls alone. Since OP has asymmetric sidewalls, I'd say it's best to absorb all of the early reflection points, which should improve the imaging (OP's stated goal).
 
Last edited:
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Really sorry, I realized that some charts were for the left channel only. I have now replaced these with the correct ones for both speakers. I also added ETC and extended Spectro to 10 Hz. I would be very grateful if you all could check if your previous comments still apply.
 
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Overall your room actually looks pretty decent. Can you plot the spectrogram down to 10 Hz? That will reveal some more of the room modes; from RT60 it looks like there could be some benefit in putting some bass traps (at room corner primarily); the panels in the link you sent are not good enough for low frequency bass trapping. Can you also show the REW clarity plots? It will give some information about direct sound vs reflected sound, which can give insight into imaging/phantom image. Putting some of these panels strategically at reflection points can improve this for mid/high freqs, covering the vocal range.
Thank you. I extended spectro and added clarity. Note that some of the previous graph were for the left channel only (now they are all correct for LR).
 

DJBonoBobo

Major Contributor
Joined
Jan 21, 2020
Messages
1,360
Likes
2,851
Location
any germ
What does ETC mean? Sorry...
Energy Time Curve - at 0 is the direct sound, peaks after it indicate reflections. On the x-axis you see the time in ms, so if there is a peak at 20m it means the sound has travelled 20ms longer than the direct sound. You can calculate how much length in cm the sound has travelled in 20ms (= 6.9 meter). With this information you can find what causes the reflection.
The graph gives an indicator if you have too much reflections. I read advice that one should aim for no peaks over -20dB in the first 20ms. Although there are different opinions about handling the reflections.

1651874380621.png


I had this example in my post i linked above:
1651874811285.png
 
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Typical of most rooms, your 100-300Hz is the messiest due to room modes. I would fill the front 2 corners with bass traps, then put diffusion on the ceiling and both side walls to start.

Normally I would say more absorption instead of diffusion, but your decay times above 200Hz are nicely consistent, and you don't want to mess with a good thing.
do you mean corner bass traps, or to fill the nooks next to the fireplace where the subs are with membrane bass traps? Note that there is a window in the left nook
 
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Energy Time Curve - at 0 is the direct sound, peaks after it indicate reflections. On the x-axis you see the time in ms, so if there is a peak at 20m it means the sound has travelled 20ms longer than the direct sound. You can calculate how much length in cm the sound has travelled in 20ms (= 6.9 meter). With this information you can find what causes the reflection.
The graph gives an indicator if you have too much reflections. I read advice that one should aim for no peaks over -20dB in the first 20ms. Although there are different opinions about handling the reflections.

View attachment 205000

I had this example in my post i linked above:
View attachment 205006
got it thanks. it is there now
 

DJBonoBobo

Major Contributor
Joined
Jan 21, 2020
Messages
1,360
Likes
2,851
Location
any germ
I don´t have any more advice. Every measurement i see from you looks pretty much perfect to me or at least "as good as it gets". Maybe someone else sees anything?
 

chych7

Active Member
Joined
Aug 28, 2020
Messages
276
Likes
422
Looking at your updated spectrogram, there are some typical room modes < 20 Hz. This is mostly inaudible, so we can ignore that. There are a few small peaks from 30-70 which may be some of the audible lower ordered room modes. Corner bass traps should help with them, but they don't look that severe.

Looking at your Clarity plots, it looks like the ratio of early to late sound is quite low, resulting in low numbers. This may have a negative impact on imaging. It is fairly even over the frequency range, which is good.

For comparison, this is what my clarity plot looks like in my treated room (with room correction disabled). It's considerably higher than your response, although it's not as even (my room treatments are not uniformly absorbing frequencies, unfortunately). The imaging in my system is excellent, IMO. For your system, adding more absorption should increase the clarity values by reducing late reflections. However, adding too much absorption will reduce decay times further, but I think your decay times are fine. One solution would be to apply broadband absorption to early reflection points and some diffusion elsewhere, to improve clarity without impacting decay times too much.

1651875920614.png



Also, can you plot L/R sweeps separately? L/R error is an indication of imaging performance. Look at both frequency response and impulse.
 
Last edited:
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Looking at your updated spectrogram, there are some typical room modes < 20 Hz. This is mostly inaudible, so we can ignore that. There are a few small peaks from 30-70 which may be some of the audible lower ordered room modes. Corner bass traps should help with them, but they don't look that severe.

Looking at your Clarity plots, it looks like the ratio of early to late sound is quite low, resulting in low numbers. This may have a negative impact on imaging. It is fairly even over the frequency range, which is good.

For comparison, this is what my clarity plot looks like in my treated room (with room correction disabled). It's considerably higher than your response, although it's not as even (my room treatments are not uniformly absorbing frequencies, unfortunately). The imaging in my system is excellent, IMO. For your system, adding more absorption should increase the clarity values by reducing late reflections. However, adding too much absorption will reduce decay times further, but I think your decay times are fine. One solution would be to apply broadband absorption to early reflection points and some diffusion elsewhere, to improve clarity without impacting decay times too much.

View attachment 205010


Also, can you plot L/R sweeps separately? L/R error is an indication of imaging performance. Look at both frequency response and impulse.
Thank you. My room correction is ON in all measurements by the way
 
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
Looking at your updated spectrogram, there are some typical room modes < 20 Hz. This is mostly inaudible, so we can ignore that. There are a few small peaks from 30-70 which may be some of the audible lower ordered room modes. Corner bass traps should help with them, but they don't look that severe.

Looking at your Clarity plots, it looks like the ratio of early to late sound is quite low, resulting in low numbers. This may have a negative impact on imaging. It is fairly even over the frequency range, which is good.

For comparison, this is what my clarity plot looks like in my treated room (with room correction disabled). It's considerably higher than your response, although it's not as even (my room treatments are not uniformly absorbing frequencies, unfortunately). The imaging in my system is excellent, IMO. For your system, adding more absorption should increase the clarity values by reducing late reflections. However, adding too much absorption will reduce decay times further, but I think your decay times are fine. One solution would be to apply broadband absorption to early reflection points and some diffusion elsewhere, to improve clarity without impacting decay times too much.

View attachment 205010


Also, can you plot L/R sweeps separately? L/R error is an indication of imaging performance. Look at both frequency response and impulse.
Added L/R screenshots
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
Be careful on the diffusion as well. Diffusion on sidewalls will randomize phase of the early reflected soundwave.
I agree with this. A balance of absorption and diffusion in any room is better than a reliance (or overreliance) on either.
 
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
I think I found the reason of the soundstage being constantly shifted to the right.
Cheap hearing test shows a sizeable gap between my right ear (red line) and left one (blue line).
After some trial and error I compensated by increasing the right channel affected frequencies by 1\4 of the gap. Image is now perfectly stable in the middle.
This is the price I am paying for 15 years of professional DJing during which I sistematically used my left ear to monitor the next song. Oh well. It was fun though.
 

Attachments

  • Screenshot_20220509-093445_Hearing Test.jpg
    Screenshot_20220509-093445_Hearing Test.jpg
    182.6 KB · Views: 77
OP
o2so

o2so

Active Member
Forum Donor
Joined
Sep 9, 2019
Messages
250
Likes
143
Location
Sydney, Australia
So, reading back the suggestions I get quite confused. Should I use broadband absorption at first reflection points, or diffusion, or both, or nothing? I would not mind increasing clarity a little now that the centre image is fixed.
 
Top Bottom