An attempt to build K-402-MEH clone in wood

[StabMe]

3 minutes ago, Marvel said:

It's a diffuser, and doesn't need to be acoustic foam. They make them out of polystyrene, too. They can be smooth and painted, or whatever. The wood always looks cool, but weighs too much. I looked at aircraft balsa, but it was going to be too expensive.

 

Did you use QRDude to calculate the design for the frequencies you wanted to diffuse?

 

From what i know, it should be made out of material dense enough not to be absorptive or worth 'transparent' in the needed frequency range, rather it should be reflective which is achieved by it being dense.

 

Don't have any calculations on hand. Was going to use some tried and proven designs which are not terribly complicated. Just need to choose the right technology to build them. If i am to cover 50% of the space between listening position and front wall, i'll have to fabricate 12-15 pieces.

[Marvel]

QRDude is pretty easy to use to calculate what you need. That's a lot of space to cover...

[Chris A]

10 minutes ago, Marvel said:

Did you use QRDude to calculate the design for the frequencies you wanted to diffuse?

 

 

2 minutes ago, StabMe said:

Don't have any calculations on hand.

 

FYI only:  it's been my experience that the frequency band 600-2500 Hz is where I see a rise in T₃₀ values when measured at 1 m in front of the loudspeaker in a predominantly untreated room with loudspeakers having fairly good full-range directivity control, like the following:

 

 

If you plug in 600-2500 Hz frequencies into the "QRDude" application, it might be interesting to see what results.

 

Chris

[StabMe]

Chris,

 

I was going to use Leanfuser for my side walls and back wall as shown on the picture a few posts above. It is not very deep (only 7.5cm) and starts working at about 1KHz. Here is a simulation of 7 diffusers put together in an array:

 

Will need to explore QRDude and educate myself on how to design 2D diffusers in this application.

[Khornukopia]

On 9/15/2021 at 5:28 AM, StabMe said:

 

Oh, i meant making a 2D diffuser out of high density XPS rather than out of wood. A variation of this kind of diffuser:

 

 

Check out the diffusers on this thread by @Tarheel TJ, 

 

 

[StabMe]

Babies are home:

 

 

Quality of this image is awful, but here are a few more photos of the process and the MEHs. 

 

All the walls (vertical and horizontal) have a spongy material that gets squeezed to about 0.2-0.5mm which is used to prevent air leaking. Now, In retrospect i understand that i should have designed the MEH so that i would have a prebuilt box with all the walls tightly glued to each other into which i would submerge the horn with drivers already attached. This would better prevent air leakage.

 

Each speaker weighs about 60 kilos. We had to bring them into my apartment on the 4th floor - it was hell  

 

I haven't tested the system yet. Have to bring amplifiers, a soundcard and a PC which would serve as a DSP and the source.

 

So, how do you like the looks?

 

Really excited about the fact that the main part of the project is now finished. 

 

How do i tune it?

[NBPK402]

Don't be disappointed if when you first hook them up if they do not sound good. The dial in process will take a while. Chris had mine sounding pretty good after a few times and then after a few more they just popped in to where they sounded. Magical. 

[Marvel]

On 9/15/2021 at 3:38 PM, StabMe said:

I was going to use Leanfuser for my side walls and back wall as shown on the picture a few posts above. It is not very deep (only 7.5cm) and starts working at about 1KHz.

The arqen.com site has a lot of great material.

 

I assume you will be putting some other furniture in the room and raising up the horns. Four floors up, without an elevator?

[StabMe]

6 hours ago, NBPK402 said:

Don't be disappointed if when you first hook them up if fhey do not sound good. The dial in process will take a while. Chris had mine sounding pretty good after a few times and then after a few more they just popped in to where they soonded. Magical. 

 

Yeah, i suppose tuning it will take some time. Hopefully Chris may find some time to give me a few pieces of advice based on measurements.

2 hours ago, Marvel said:

The arqen.com site has a lot of great material.

 

I assume you will be putting some other furniture in the room and raising up the horns. Four floors up, without an elevator?

 

Yes, i am waiting for the TV stand which i ordered in a local wood workshop. They take forever to finish it. It is 3 meters wide and 60cm tall. Speaker stands are made of metal. Will post a pic of the whole setup with stands and furniture in future posts.

 

As for the other furniture - there will be a book shelf on a side wall near the back wall. Currently deciding on putting Leanfusers on the backwall as well. I have 8 of them made out of plywood that i don't use currently, so maybe i'll paint them to better fit the room design and hang on the back wall.

 

Ordered XPS insulation panels of 35kg/m3 density. Will be making 2D diffusers for the ceiling between couch and front wall and Leanfusers for the side wall and front walls near the horns (between each horn flange and TV). Hopefully i'll be able to paint them so they fit the room.

 

Maybe i'll put thick fiberglass panels behind the horns to handle the mid bass reflections.

 

And yes - no elevator in this 4 floor building

[StabMe]

Ok, i hooked everything up and CD on the left speaker sounds normal, but this is what i get from the CD on the right speaker:

 

 

I tried switching channels on the amp and channels on the soundcard. Still get the same sound on both speakers. What could've happened to this CD? I am using FaitalPro HF204.

[Khornukopia]

On 9/22/2021 at 2:50 AM, StabMe said:

but here are a few more photos of the process and the MEHs.

 

I copied and re-sized (smaller for posting here) some of your pictures. These look good.

 

 

 

 

[StabMe]

3 hours ago, Khornukopia said:

 

I copied and re-sized (smaller for posting here) some of your pictures. These look good.

 

 

 

 

Hey, thank you!

 

BTW, the issue with CD freq response is resolved now. I actually messed up with the wiring. Had first listen to them simply using a 4th order Linkwitz Riley filters at 475Hz and setting HF/LF levels by ear. Wow, they are so LOUD. Will start tuning them with a mic tomorrow. 

[NBPK402]

On 4/16/2021 at 2:53 AM, StabMe said:

Deleted

[StabMe]

Had a chance to run more a less listenable setup of the system. 

 

Used 6db slopes on each drivers, set levels roughly matching and then did a series of 1m measurements. 

 

Used some 2" rockwool absorbers, too. Here is how it looked:

 

 

These panels are only for measurements and won't stay.

 

Raw measurements looked like this:

 

 

I then used a 6dB crossover: 525Hz on LF and 425Hz on HF:

 

 

Full signal:

 

 

Looks very ragged.

 

I then EQed full signal flat down to 200Hz using software called APL - it generates FIR filters and uses its own VST plugin as convolver. After that, i moved to listening position, did some moving mic measurements using pink noise and tried to tame everything from 200 and down. Dips cannot be EQed - this i know, so i tried to manage the peaks.

 

Here is the response using pink noise:

 

 

Some songs sound very good. Imaging is great. But the sound is still too boomy, mostly in the bass frequencies. I can hear a lot of echo. Here is the spectrogram and RT60 time of unEQed measurements:

 

 

 

RT60:

 

 

 

And waterfall. Had to extend time range to 2000:

 

 

Are those times to big?

 

Currently, i only have a big sofa in the room and a TV. The room is big. I ordered some diffusers that i will put on the side walls to the left and right of my ears when sitting on the sofa (can't put the diffusers on the side walls in between LP and front wall - have a big window). I will also put diffusers to the front wall next to the horns. Some diffusers will also be put on the ceiling. Will have to buy a rug. 

 

Probably, can make basstraps that Crhis recommends and place them in the corners to tame midbass region.

 

I notice some peak on the RT60 graph in the 100-150Hz region.

 

So, what do you guys say i should do first? Any advice on tuning? 

Edited October 1, 2021 by StabMe

[StabMe]

BTW, does the response look like i need a sub? First, to extend down to maybe 20-25Hz and to fill the dips?

[Khornukopia]

9 hours ago, StabMe said:

I can hear a lot of echo.

 

Are you saying the room has echo, such as when you clap your hands?

[Chris A]

On 10/1/2021 at 3:27 AM, StabMe said:

RT60:

 

 

I'd say you've got a lot of work to do on controlling your reverberation times (RTs) unless you have an extremely large listening room.  I'd aim for at least reducing the RT30 values to below 1 ms, with 0.4-0.5 ms being much more workable. 

 

On 10/1/2021 at 3:27 AM, StabMe said:

Full signal:

 

 

Looks very ragged.

 

I''d make sure that the curve smoothing is set to "psychoacoustic" before trying to EQ anything.  Overall, the closer you can get the SPL response to plus/minus 3 db using psychoacoustic smoothing, the better it will sound. Plus/minus 2 dB is even better.  I recommend using REW's EQ facility to do this.

 

It looks like the gain of the bass bin is set too low, and you're trying to boost the 100 Hz notch in response.  I'd instead recommend boosting the overall bass bin channel, then attenuate the 200-500 Hz band, then boost the 70-130 Hz portion using no more than 3-4 dB.  That way, you'll be avoiding using too much PEQ gain, which can cause internal issues with a DSP crossover.

 

Chris

[StabMe]

On 10/1/2021 at 10:20 PM, Chris A said:

I'd say you've got a lot of work to do on controlling your reverberation times (RTs) unless you have an extremely large listening room.  I'd aim for at least reducing the RT30 values to below 1 ms, with 0.4-0.5 ms being much more workable. 

 

I am sure you meant 1 s, but yes - i have a lot of work to do. Will report back when i apply some treatments and measure the result. But the RT is too big at this moment.

 

On 10/1/2021 at 10:20 PM, Chris A said:

It looks like the gain of the bass bin is set too low, and you're trying to boost the 100 Hz notch in response.  I'd instead recommend boosting the overall bass bin channel, then attenuate the 200-500 Hz band, then boost the 70-130 Hz portion using no more than 3-4 dB.  That way, you'll be avoiding using too much PEQ gain, which can cause internal issues with a DSP crossover.

 

That was a very good advice, actually.

 

A raised bass bin volume so that its levels matched the level of compression driver at x-over frequency. After that i attenuated the 200-500 band so that its level also matches CD at x-over. It required -17db cut at about 300Hz with Q equal to 1.

 

Took measurement at 1M distance:

 

 

Then ran APL correction to about 180hz and upwards to achieve this on both channels:

 

 

This is at 1/6 oct smoothing so one can see the difference. APL takes its own measurement, I used REW to see what's happening pre-post. This very flat response is exactly the same mic position as was used in APL software.

 

Then i took some measurements at listening position and only dealt with 20-200Hz region. This is what i was able to achieve:

 

 

Sorry for the picture from the screen from my laptop - forgot to save the REW file.

 

Overall - very good imaging, very deep bass. Reverberation ruins the detail, though. Singing, acoustic versions of some of favourite songs, strings (Richters version of Vivaldi's 4 Seasons, for example) - all sound good and excessive reverberation add some interesting taste to it, but listening to music with many layers of intricate detail doesn't sound good. 

 

Need. More. Room. Treatment.

 

 

Chris, can you please tell me what is the thickness and density of the material used in the bass traps you placed behind your speakers? Since I'll be adding some treatment, why not build the traps as well  And by the way - you recommend treating sidewalls near the edges of the horns - what thickness and density would you recommend in these traps?

Edited October 3, 2021 by StabMe

[Chris A]

21 hours ago, StabMe said:

I am sure you meant 1 s

Yes.

 

21 hours ago, StabMe said:

can you please tell me what is the thickness and density of the material used in the bass traps you placed behind your speakers?

I use double Owens-Corning 703 fiberglass panels (total 4" thick each) in ATS Acoustic bags. They are placed across the room corners and are closed on one end (sitting on top of the TH-SPUD DIY subwoofers in each front corner). 

 

It's important that one end is closed off in order to make it a 1/4 wavelength trap (down to ~70 Hz based on the panel length).  If you leave both ends open, then it becomes a half wavelength trap (i.e., 140 Hz).

 

Chris

[Chris A]

22 hours ago, StabMe said:

And by the way - you recommend treating sidewalls near the edges of the horns - what thickness and density would you recommend in these traps?

If you place the horn mouths within 2 feet (0.6 m) of a side wall, I then recommend a line of 1" (2.5 cm) thick laundry-lint-type squares of 2 feet by 2 feet dimensions (0.6 x 0.6 m) placed just adjacent to the horn mouth exits.  This will dramatically help to recapture the phantom center stereo image, in my experience.  You can experiment using more squares just around the horn mouths--out to perhaps 6 feet (2 m) radially.  You may need to place more absorption along the side walls, carpet on the floor, etc. to help control the nearfield reflections to the loudspeakers--and the listening positions.

 

Chris