Active Horns Acoustical Treatment & DSP

[vego99]

I’m hoping some Horn / DSP / Acoustical guru’s can chime in to help give some guidance with my room / system.

Room: 14’ x 20’ x 8.5’ : ( Using short wall for speaker placement)
Speakers : Underground Jubilees / K402 & 1502 Subwoofer(behind listening position)
Acoustic Treatment : DIY 6” panels in corners and 2” panels on early / 1st reflection point. ( nothing on ceiling)

I built this room from the ground up and is a dedicated listening room. The Jubilees have been in the room since it was built and I’ve gone through a few different ways of processing ( active dsp) and am currently using Audiolense for all Xover/EQ/Delay, etc..

The system sounds good. Though I question if I need to rethink my acoustical treatment / speaker / listening position vs. letting Audiolense to work it’s magic.
I’m a firm believer that DSP is a great tool for achieving those last percentages of audio bliss but should not be used as a crutch to fix an acoustically poor room.
In regards my acoustical treatments, I just used info I found online and placed where I thought was correct.

In lies the problem. How do I go about measuring and acoustically treating the room with just the raw responses from the loudspeakers? How should I appropriately adjust speaker / listening position with the raw loudspeaker responses?

I’ve attached the Mdat file of raw responses from Audiolense and final corrected reponses played through REW using a Earthworks M30 mic.

Any help / guidance would be greatly apprecated!

 

[vego99]

Any suggestions on how to upload the Mdat file?

 

[Flaesh]

upload the Mdat file?

zip file and upload

 

[vego99]

Here are both the corrected and raw measuremeents.

Sorry had to split into two zip files.

Thanks

 

[Keith_W]

Thank you for uploading your .mdat. I had a look. Some comments.

These are the corrected curves for left (green) and right (purple) speakers. There is a 10dB slope between 20Hz to 20kHz so IMO this would be way too much bass. I would personally go for a 5dB slope, but that's up to you. The peaks and dips between 40Hz - 400Hz are normal and I would not be too concerned with it. It looks like Audiolense has done a good job.

The step response of the left speaker looks almost the same as the right. It looks as if the subwoofer may not be time aligned correctly, there is a peak about -23.5ms prior to the main impulse.

This is the first 30ms of your Energy-Time Curve for your "Corrected Left Speaker". I have labelled in the first 5 reflections for you. These arrive at 1.64ms (1.8ft), 3.20ms (3.6ft), 4.83ms (5.4ft), and 7.81ms (8.8ft) respectively. All are below the -15dB goal.

This is the T20/T30 of the left speaker. It shows about 300ms rising to about 350ms for bass. This is smack bang in the middle of the target response, so it is pretty good.

The spectro looks excellent. Nice uniform decay through the frequency range.

If I were you, I would be pretty happy with the result. If you think it sounds good, then IMO there is nothing more you need to do. There is a strong likelihood that adding room treatment will mess up the smooth looking spectro so I would leave it well alone.

 

[Flaesh]

I couldn't find the sub in the photos.

 

[vego99]

zip file and upload

I couldn't find the sub in the photos.

The big black box with records on top of it….

 

[Flaesh]

I first asked if there was a subwoofer on the back wall, but then I deleted it. I don't believe in rear subwoofers!))

p.s.: Are these measurements taken with HF20AT?

Subwoofer(behind listening position)

mentioned in the start post. Sorry, I didn't read carefully)

 

[Ruhled]

I'm not in front of REW but based on the graph above I would lift the tweeters 4-6db at the crossover and just fix the three peaks under 100hz in the bass department.

I'm a horn guy too running some 4560s with emilar 175s and aluminum Renkus-Heinz 500hz CD horns on the top. I'm also working on refurbing some 4520 scoops to possibly replace the 4560s if they seem to work better in my room. I haven't worked up the courage to move one in the house yet to measure.it though.

 

[Mr. Widget]

Room: 14’ x 20’ x 8.5’ : ( Using short wall for speaker placement)

That is the common layout, and is good because you get to have plenty of space behind the listening position. That said, I usually prefer rotating the room 90° and placing the speakers on the long wall. This helps with horizontal first reflections and may help with the bass null typical of listening positions near the middle of the room.

 

[vego99]

I first asked if there was a subwoofer on the back wall, but then I deleted it. I don't believe in rear subwoofers!))

p.s.: Are these measurements taken with HF20AT?


mentioned in the start post. Sorry, I didn't read carefully)

These are with the HF20AT Drivers

 

[vego99]

Thank you for uploading your .mdat. I had a look. Some comments.

View attachment 393647

These are the corrected curves for left (green) and right (purple) speakers. There is a 10dB slope between 20Hz to 20kHz so IMO this would be way too much bass. I would personally go for a 5dB slope, but that's up to you. The peaks and dips between 40Hz - 400Hz are normal and I would not be too concerned with it. It looks like Audiolense has done a good job.

View attachment 393651

The step response of the left speaker looks almost the same as the right. It looks as if the subwoofer may not be time aligned correctly, there is a peak about -23.5ms prior to the main impulse.

View attachment 393648

This is the first 30ms of your Energy-Time Curve for your "Corrected Left Speaker". I have labelled in the first 5 reflections for you. These arrive at 1.64ms (1.8ft), 3.20ms (3.6ft), 4.83ms (5.4ft), and 7.81ms (8.8ft) respectively. All are below the -15dB goal.

View attachment 393650

This is the T20/T30 of the left speaker. It shows about 300ms rising to about 350ms for bass. This is smack bang in the middle of the target response, so it is pretty good.

View attachment 393654

The spectro looks excellent. Nice uniform decay through the frequency range.

If I were you, I would be pretty happy with the result. If you think it sounds good, then IMO there is nothing more you need to do. There is a strong likelihood that adding room treatment will mess up the smooth looking spectro so I would leave it well alone.

Thanks for reviewing the files!!

Yes the target curve has about a 10dB slope from where the room starts to drop FR to 20khz. I’ve tried a flatter curve even 5db slope among others and so far this has been the most neutral.

Step Response : Keen eye…. I’m not sure if that’s the subwoofer or if its the filter pre-ringing.

Audiolense is a powerful tool but there is alot more that goes into a good filter than just drawing a curve and clicking compute. After a year of use and hours of playing with the program Im finally getting somewhere.. Things do sound fantastic right now.

Again, thanks for reviewing the file.

Though, I still question are my current room treatments helpful or is Audiolense overcorrecting for improper acoustic treatment.

I should really strip the room of all treatment and see what happens.

 

[vego99]

I'm not in front of REW but based on the graph above I would lift the tweeters 4-6db at the crossover and just fix the three peaks under 100hz in the bass department.

I'm a horn guy too running some 4560s with emilar 175s and aluminum Renkus-Heinz 500hz CD horns on the top. I'm also working on refurbing some 4520 scoops to possibly replace the 4560s if they seem to work better in my room. I haven't worked up the courage to move one in the house yet to measure.it though.

Ill try modifying the cuve to lift the HF. Thanks for the tip.

Not sure if I can get under 100hz much better, the pic is in 1db increments so the dips and peaks seem a little excessive..

 

[vego99]

That is the common layout, and is good because you get to have plenty of space behind the listening position. That said, I usually prefer rotating the room 90° and placing the speakers on the long wall. This helps with horizontal first reflections and may help with the bass null typical of listening positions near the middle of the room.

I’ve tried this but quickley changed things back before I could really try dialing stuff in.. With speakers this big, space runs out quickly.. On the long wall with 14’ depth, things get a little up close and personal, and youre also literally on the back wall.

 

[DWPress]

I should really strip the room of all treatment and see what happens.

If it's an easy job then I'd do it if you don't have previous measurements to compare with. Only way you'll know if your placement could be optimized or if they're doing something negative. I have the same size room and big speakers and listen toward the 14' wall though back a bit further at around 11 feet, my ceilings are 11.5' but I see a lot of similarity with your room measurements. I'll look at your mdat later and try to be more constructive.

edit: I'd also try tilting the upper frequencies up a bit just looking at the above graph.

 

[vego99]

If it's an easy job then I'd do it if you don't have previous measurements to compare with. Only way you'll know if your placement could be optimized or if they're doing something negative. I have the same size room and big speakers and listen toward the 14' wall though back a bit further at around 11 feet, my ceilings are 11.5' but I see a lot of similarity with your room measurements. I'll look at your mdat later and try to be more constructive.

Happy to hear this…. I’m about 11-12’ back from the front wall / face of the bass bins..

I’ve contemplated putting the sub on the front wall with the mains.

 

[Keith_W]

Step Response : Keen eye…. I’m not sure if that’s the subwoofer or if its the filter pre-ringing.

An appearance like that on the step response is definitely in the bass frequencies. It could be one of three things - (1) subwoofer not time aligned, (2) excess phase not adequately compensated for, (3) low frequency pre-ringing. You need to do a bit of experimentation if you want to get rid of it. Regardless, I suspect that the audibility would be pretty low and the only reason you should clean it up is to make the step response look nicer.

Audiolense is a powerful tool but there is alot more that goes into a good filter than just drawing a curve and clicking compute. After a year of use and hours of playing with the program Im finally getting somewhere.. Things do sound fantastic right now.

Indeed it is. I don't use Audiolense, but I have friends who do. I am very impressed by what it does.

Though, I still question are my current room treatments helpful or is Audiolense overcorrecting for improper acoustic treatment.

I should really strip the room of all treatment and see what happens.

You could of course try different correction strategies with Audiolense. Bear in mind that you are looking at a single point measurement with an omnidirectional microphone, which Toole keeps reminding us - is NOT what we hear. I have no idea if Audiolense lets you do this, but you could try making your timing adjustments with a single point sweep, and your amplitude adjustments with a MMM or an average of a few sweeps in different positions. That would be closer to what you would hear.

 

[Flaesh]

I’ve contemplated putting the sub on the front wall with the mains.

Try this!
And you can see the link in the signature about my subwoofer system. (BTW speakers are SEOS24 with DCX464 and 15" in small 40 l boxes in front corners, 45º toed in)

This is weird:

You may want D50 more, say, 90% in the entire range.

 

[Bjorn]

You need to treat specular reflections in a more broadband way. 2" panels should be removed and replaced by minimum 6" if it's DIY (commercial absorbers like Broadsorbor can go lower with less depth). A 6" with porous in the corner isn't doing much in the bass by the way.

Treating the ceiling reflections is important here, because the K-402 horn looses its directivity at around 1000 Hz. The lower the treatment can go, the better it is.

On the rear wall, quality diffusers that work to the highest frequencies is likely the best choice for treating the reflections if the distance is sufficient (hard to see from the picture). Combined with low frequency absorption.

You should also treat part of that rear beam, which is causing specular reflections.

 

[Tim Link]

the K-402 horn looses its directivity at around 1000 Hz.

It looks to me like it has significant vertical pattern control down to 300Hz based on the measurements posted here: https://data-bass.ipbhost.com/topic/1381-measurements-from-big-cds-and-horns/ I would expect it to because it's huge!