Second opinion on my measurements

[pollock0424]

@neRok
Thanks, got it! I pressed every button except the "Actions". When I was exploring, I thought that I pressed that button but apparently I didn't because I don't remember seeing the subsequent pop up window!

Look at this peak at 37Hz, FIRs align very well! Interesting to verify...

 

[neRok]

@neRok
Thanks, got it! I pressed every button except the "Actions". When I was exploring, I thought that I pressed that button but apparently I didn't because I don't remember seeing the subsequent pop up window!

Look at this peak at 37Hz, FIRs align very well! Interesting to verify...

View attachment 312096

You're a whole cycle off there (nearly 28ms delay). That's why there is a new cancellation at ~120Hz. Put the cursor at 120Hz and "Filter at cursor" to see the impact of using that delay.

As a general comment - you've got 2 identical speakers arranged symmetrically about the listening position and room, so the responses should be close to aligned with 0 time offset. If you had accidentally wired one "backwards", you could tell that here (they would be aligned, but opposite).

Your response aren't the same because your right speaker has way more reflections than your left. You can see the problem at other random frequencies;


I would move the right speaker in to alignment with your left, measure both again with acoustic timing ref, and see what its like then. If that doesn't improve the right much (which I doubt it will), then there must be something slightly different about the room that is affecting the right speaker more than the left. It could be the door or a window opening, or a specific piece of furniture? Looking at your photos, I wonder if it is the door. Perhaps the right speakers sound is going straight through the door, reflecting on next wall on the other side of the door, and then coming back through the door?

Edit: Maybe it's the couch, but I would think that would cause more absorption on the right speaker, rather than more reflections. If you can manage it, perhaps try remove the couch, or temporarily move it to the back wall so that it is symmetrical. If things stay the same and the door is still looking like a culprit, perhaps stand the couch on its end and block the door with it! See if it absorbs then, lol.

 

[pollock0424]

It could be the door or a window opening, or a specific piece of furniture? Looking at your photos, I wonder if it is the door. Perhaps the right speakers sound is going straight through the door, reflecting on next wall on the other side of the door, and then coming back through the door?

Edit: Maybe it's the couch, but I would think that would cause more absorption on the right speaker, rather than more reflections. If you can manage it, perhaps try remove the couch, or temporarily move it to the back wall so that it is symmetrical. If things stay the same and the door is still looking like a culprit, perhaps stand the couch on its end and block the door with it! See if it absorbs then, lol

This office is in the backyard and there is a gap of about 20 feet between the door and the main house. The footon is here temporarily and will go back to my daughter's room which is getting painted (with honeybee theme ) . I have plans to bring a 2 person bean bag in footon's place though!

I am busy again for this upcoming whole week and I will take some measurements hopefully next weekend. I will try tweaking the position of the speakers to minimize the mutual cancelations but that seems like a non-convex optimization problem given the chaotic behavior of acoustics in this small space.

 

[ozzy9832001]

Changing the delay on the speakers could cause more harm then good, especially if the problem is limited to only a select frequency range. You could use filters to change phase...but it will take awhile to design them properly.

Honestly, I think you're getting caught up in the micro details and losing sight of the larger picture. The room needs to be reworked and positioning needs to be tweaked.

I'd try to keep things simpler and try to change the flow of the room first. There is a lot of stuff in the room and most of that is going to cause comb filtering or other problems.

You should probably put the speakers near the front wall to minimize SBIR and push it to a higher frequency that will be easier to treat later on. If the MLP is the back wall, then I'd have the rack and your equipment on the front wall, between the 2 speakers or if you have to have it on a SW, then it should be away from a position that is a primary reflection point for the MLP.

The foam wedges you have all over the room do nothing for absorption...while they may help slightly with flutter echo, they are probably causing reflections off of them. To maximize their minimum effect, they should be placed near a corner or where 2 surfaces meet, i.e. wall and ceiling. This gets them out of the way.

The couch is probably acting as both an absorber and a reflector and is probably the source of the added reflections from the right side...the desk chair and desk another source.

If all else fails, get a sub or 2. Place them in the front 2 corners.

It can become very difficult to get quality sound when there are a lot of stuff in a room. This is why a lot of listening rooms are sad, lonely places because we have to make compromises to achieve the best sound we can. Sometimes a couch needs to be there (duh), but maybe it's location can be moved to achieve a more acoustically symmetrical look.

 

[pollock0424]

@ozzy9832001
Yes, I'm slowly coming to the realization. Unfortunately, I have to have most of the stuff in the office if not all. Without the foam that I have it was very echoey and I didn't like the sound at all. High frequencies were emphasized and even some out of phase sound also was present at some mid/treble frequencies.

What would be a reasonable (no nonsense) broadband absorption solution for me? I will try to push the speakers back another 8 inches and try to get 2 to 3 sets of measurements in between and also vary the toe in.

I will wait until the footon is moved out and the bean bag is brought in.

But I agree, my best way forward is to move the speakers to where they were originally and get 2x subs and cross them at 80hz. The original arrangement had the best imaging in my opinion.

 

[ozzy9832001]

@ozzy9832001
Yes, I'm slowly coming to the realization. Unfortunately, I have to have most of the stuff in the office if not all. Without the foam that I have it was very echoey and I didn't like the sound at all. High frequencies were emphasized and even some out of phase sound also was present at some mid/treble frequencies.

What would be a reasonable (no nonsense) broadband absorption solution for me? I will try to push the speakers back another 8 inches and try to get 2 to 3 sets of measurements in between and also vary the toe in.

I will wait until the footon is moved out and the bean bag is brought in.

But I agree, my best way forward is to move the speakers to where they were originally and get 2x subs and cross them at 80hz. The original arrangement had the best imaging in my opinion.

If you are handy, building some rockwool/mineral wool panels will be a good solution to reducing the overall liveliness of the room. You can experiment with positioning and even alternate heights on them to give it a more pleasing appearance and it helps to spread the absorption out a bit. For reference, my space is 11.5x9.5x8ft. I have 12 4" panels (completely broadband), 4 on each sidewall, 2 on the front wall between the speakers and 2 on the rear wall. I also have 4 corner tri traps (got them from GIK acoustics) and they are placed in the 2 front corners (I listen in the nearfield, if you were in the far field you'd probably reverse them). Treating the ceiling is personal preference, IMO. Really depends where you are in relation to the speakers, but it's worth taking some measurements. For me, it didn't do much.

The most important thing will always be symmetry between the speakers and any boundaries. This is especially true in smaller spaces where reflections can have huge a huge effect on the over tonality of the speakers.

Even with the subs, you'll want to experiment with locations, but corner loading them is usually a decent option...that's how mine are right now. You may want to experiment with crossover frequency to taste. Anywhere from 60-120hz is fine. I prefer a slightly higher XO, about 120. It helps with 2 room modes...one at 100hz and one at 120.

 

[pollock0424]

@ozzy9832001

I don't have a a few tools to make the panels but I could ask someone else to make them for me. Here is the product that I have, I have stacked 3 of these. Look at their NRC plots.

TroyStudio Thick Acoustic Foam Panels, 12 X 12 X 2 Inch 36 Pcs Broadband Sound Absorbing Foam, Dense Soundproof Padding Tile, Recording Studio Foam Absorber, Groove Decorative 3D Wall Ceiling Panel https://a.co/d/cPEJU7i

Also, I found this online:

https://www.thefoamfactory.com/acousticfoam/productcomparison.html

 

[neRok]

TroyStudio Thick Acoustic Foam Panels, 12 X 12 X 2 Inch 36 Pcs Broadband Sound Absorbing Foam, Dense Soundproof Padding Tile, Recording Studio Foam Absorber, Groove Decorative 3D Wall Ceiling Panel https://a.co/d/cPEJU7i

Thin foam like this doesn't work at low frequencies, which is where the major room reflection problems occur (modes, SBIR, etc). The Amazon link shows the declining efficiency in the bass range. Looking at graphs from your first post you can see your room has pretty even decay above 1kHz, so the foam is probably helping there. Also 200ms for -40dB is not so bad (my room is >400ms lol). Once you have decided upon your speaker positions, then you might consider moving the foam to a few specific positions, to target specific problems. I indicated some peaks and nulls that might be improvable this way, but the ones <500Hz might not work (because the foam gets bad <500Hz).


4" thick panels filled with dense insulation will probably do better than the foam, but still won't do enough in that problematic ~<250Hz region. I've been reading up on room treatments and it seems you can get speciality products to treat bass ($$$) or use a deep amount of "fluffy" insulation as many forums call it. You could do like 1' of it on your whole front wall for example, which would really make a dent in that length mode peak at ~110Hz. Here's a thread I was reading the other day, and the graphs are from this calculator.

 

[pollock0424]

@neRok

I have 3 of these foam panels stacked, my guess is that stacking probably has something to do with the even decay above 1khz. Also, 1 foot of rockwool sounds like a good idea lol! I'm not very good with carpentry but I have acquaintances that can help for some money.

I will definitely document the process here as I try different things. I would like to get the room under control <200ms upto 100hz and then I will use subs below that but for the mean time I can move the speakers slightly backwards to push the SPL loss up into the frequency and use treatments for that. Will see..

 

[pollock0424]

A quick update:

1. I traced the distortion around 300 - 400hz to a loose driver (12 inch ). Sometime ago, I removed all the drivers in the left speaker to trace for mute mid range. I found a cold solder but I didn't tighten the screws all the way I guess. Same issue in the right speaker but not as prominent. The distortion was faintly audible actually from the left speaker.

2. Also, I noticed that by placing the heavy crocheted blanket on my desk (on the left side) FR blip in 200 to 250HZ cleaned up too.

 

[pollock0424]

@ozzy9832001 and @neRok :
A significant development, I positioned the speakers along the long wall in my room(16ft x 12ft). Speaker baffle is about 2.5ft from the wall. Speakers are 8.2ft apart and I am sitting 9 ft away from tweeters.

But the problem is, I have stuff right in between speakers now! The distance between the stuff and the speakers is around the wavelength of mid range/Upper mid, I am concerned that I might be getting some refractions, but so far I like it! Vocals seem to be "forward" a bit! It could be just in my head. Here is a link to my L, R, and L + R measurements if you're interested. I am also using this thread to document my journey with speaker placement.

 

[ozzy9832001]

The vocals would probably seem a bit forward, since you're getting a lift from 500hz to about 1k and the drop off between 1k and 2.5k probably isn't noticeable since we are more sensitive to those frequencies in that range. I would guess it would sound a bit more balanced. Bass looks clean minus the small hump at 50hz which may or may not be noticeable.

I have stuff in-between my speakers as well. A monitor, desk, PC and even the subwoofer which is quite large. If I move all the stuff out, there are some minor differences as far as measurements are concerned, but I really can't hear much of one. It really depends on how big the objects are, where they are and what the directivity of the speakers are. Most of the soundwaves that a small desk or monitor would be causing issues with are not coming off the driver at an angle that would be problematic.

 

[cyjanopan]

For frequency response, you need to do two measurements or at least two settings for the same measurement:

1. From 20 to 200 Hz to determine room modes with 1/12 or 1/24 octave smoothing (or none).

2. From 20 to 20 kHz for overall tonality with 1/6th smoothing.

or just use variable smoothing

 

[ahofer]

I had Thiel CS3.6 for about 25 years. Great speakers. Too bad about the company.

 

[pollock0424]

@ozzy9832001
Okay, I took a few more measurements after using Dirac, it definitely contained that 50Hz mode (became unbearably boomy at higher volumes).

I have noticed that Dirac ON greatly improves sound quality when listening to Billie Eilish. However, if I listen to say Fleetwood Mac then I prefer the Dirac off option. Also notice the increase in distortion with Dirac ON

Also notice the messed up RT60 with Dirac ON <200Hz.

For now, I reduced the Dirac gain at 100hz. I'm thinking that I don't need Dirac at all, just use PEQ to control the bump at 50hz and I should be alright?

 

[radix]

I'm not sure what your system is and if you're using unbal or bal connections with the SHD (or digital in?). If you have analog in, you need to make sure the preamp is not clipping the SHD input. That can be a big source of distortion.

 

[pollock0424]

@radix
I'm using a miniDSP SHD and 2 x Purifi 1ET7040SA Buckeye Amps. I'm using XLR outputs. miniDSP is my preamp, I'm not using anything else. I think the distortion is from Dirac gain that I added to add "weight" to the music, I toned that down by a bit. I haven't taken a measurement to verify by how much that distortion has gone down. I think that I have some loss in SPL at about 100hz because of room interaction and my extra gain combined with Dirac..probably contributed to that distortion. Note that, I couldn't hear any of this distortion.

 

[radix]

Did you download the Harmon target curves? I don't remember if they are bundled or not. In DIRAC, if you go to the File menu, there's an option for "Load target curve" and there should be a +4, +6, +8, and +10 Harmon curve. They are in the $HOME/Dirac/targets directory. Or maybe I added them?

If you are going to have a large boost, you might need to apply something like a -6dB gain or drop the whole curve. I believe if you adjust the DIRAC 1 and DIRAC 2 sliders, those apply pre-DIRAC gain to help prevent boosts clipping. not positive about that.

 

[radix]

I'll also toss out that I usually just pick the Harmon +4 or Harmon +6 (or save those two to different presets). The, if I want to tweak the bass or something, I use the SHD PEQ rather than try and adjust it in the DIRAC software. It's a lot easier to do it in the PEQ. You can also link channels (e.g. 1+2) so your PEQ settings are automatically applied to both. The PEQ is also saved per-preset.

 

[pollock0424]

@radix
Good points,
I will reduce the overall SPL and try, also didn't know that PEQ can be applied to multiple banks.