Active Room Treatment (ART) by Dirac

[mmares056]

If Dirac or the support-level user settings determine that subs 2/3/4 contribute more (or less) than sub 1 to the LFE signal, aren't they still just supports subs ?

"Subs 2/3/4 are support" and "the LFE signal is distributed in an optimized manner among all the subs" seem to mean the same thing, with ART


I think you would agree that if circumstances require it (such as suboptimal frequency response), supports subs can play LFE signal louder than sub 1.

It seems to me to be more of a misunderstanding of the terms than anything else this disagreement.

Yes this is how ART algo works with LFE and rest of support channels. In my asymmetrical, open concept living room, I have my 4 subs separated into their own group because I get difference performance from every sub due to compromised and unmovable MLP and limited placement options for my subs. Even though I have placed my subs in the best possible locations for ART to do its magic, room modes and bad room acoustics prevent my subs from reaching their full potential which would be 20Hz to 200Hz measured response.

I have 3 subs on the front, 2 corner loaded and 1 at the 3/4 of the front wall. Sub 4 in the RR corner 7.5' from the MLP. The detected range by Dirac is different from all 4 subs, Sub1(FL corner leaded) detected range 20.5Hz-116Hz, Sub2(3/4 of Front wall) detected range 22.3Hz-105Hz, Sub3 (FR corner loaded) detected range 20.1Hz-83.9Hz, and Sub4 (RR corner loaded) detected range 20Hz-150Hz. So in my use case the best performing sub is Sub 4 and the validated response support the fact Sub 4 contributes more than the other three subs where support is needed in the 20Hz to 150Hz range so that ART can hit my bass target curve.

Even with ART cutting off the high frequency response early for three of my subs, it still manages to nail my 10dB bass rise target. So ART's algo picks the best pieces of the puzzle to linearize the phase and perform precise sub to speakers time alignment in the speaker FSL region.

 

[JRowland]

Have you checked Dirac LFE-channel subwoofer filter responses? In my system, only Sub 1 has response up to 120 Hz, the others do not.
Previously, before the fixed 120 Hz filter was added, it was very obvious that only Sub 1 was handling the LFE channel.

For me sub 1 is the only thing with response up to the 150Hz handoff but all my support speakers go to 120Hz

My sub 1 is definitely the main actor by looking at the filters but this looks different for others in their rooms. Ken’s images show great “unison”.

Raw responses

@KenMasters how does your spread look with just the subs as support? Mine sucks without adding in the rest of my speakers. I still haven’t made time to play with continuous audio correction again but I have left it on without noticing any negatives. Better safe than sorry after listening to your clips!

 

[KenMasters]

@KenMasters how does your spread look with just the subs as support? Mine sucks without adding in the rest of my speakers.

It's not particularly tight:

If I add the speakers the graph might suggest a marginal improvement:

But it doesn't sound good, bass loses punch while simultaneously creating a sense of pressure on the ear - the LS50s and R8s are not suited to bass reproduction. Even with speaker support I limit all speakers to 100Hz.

Not the most effective use of ART, but my system is what it is.

 

[Mark609]

Fixed it!

 

[RenPa]

how does your spread look with just the subs as support?

I have 2 subs only and WAF allow me onnly one configuration is left and right side of the mlp.
Raw response is goos from 20Hz to 200Hz ( even from 17Hz issh...):

Spead with conventional settigs (spl -18 fsl 20Hz fsh 150Hz for both) is:

I found this with a little change: fsl for sub 1 LFE at 22HZ spl at -5dB and for sub 2 spl at -5dB. One could fear to lose basses. Ir was not the case.

 

[JRowland]

I have 2 subs only and WAF allow me onnly one configuration is left and right side of the mlp.
Raw response is goos from 20Hz to 200Hz ( even from 17Hz issh...):

View attachment 541308

Spead with conventional settigs (spl -18 fsl 20Hz fsh 150Hz for both) is:

View attachment 541311

I found this with a little change: fsl for sub 1 LFE at 22HZ spl at -5dB and for sub 2 spl at -5dB. One could fear to lose basses. Ir was not the case.

View attachment 541313

Less overall boost with that tighter spread bottom screenshot as well, sounds better to your ears as well? Curious if you set both subs to default -18db and let them both run 20-150Hz how does the ART algorithm do the filters when you are not forcing its hand by setting either sub to above 20Hz FSL or lowering support settings?

Edit: that was a dumb question, I missed the description of the second screenshot spread denoting what I asked for, sorry.

 

[TimoJ]

All supporting LFE:

Just subs (3 subs, sub1 and sub2 in the same group):

 

[Mort]

Finally connected a second subwoofer to my dual GLM/DiracART system for testing.
I compared with one sub a few months ago. The differences were surprisingly small.
It's a bit of a pain to set up honestly. I hope it's worth it.

 

[JRowland]

All supporting LFE:
View attachment 541315

Just subs (3 subs, sub1 and sub2 in the same group):
View attachment 541316

Nice! How many measurements and how far of a distance from MLP are they on average?

 

[TimoJ]

Nice! How many measurements and how far of a distance from MLP are they on average?

7 measurements, about 45-50cm from MLP.

 

[KenMasters]

7 measurements, about 45-50cm from MLP.

I would imagine the number of measurements makes a difference to the spread. I have 17 measurements, most in about a 40cm bubble, but some extended couch positions as well - with two about 80cm out.

 

[JRowland]

I would imagine the number if measurements makes a difference to the spread. I have 17 measurements, most in about a 40cm bubble, but some extended couch positions as well - with two about 80cm out.

It sure does. 9 measurements and 12” I had a tight spread like TimoJ. I’m at 21 measurements and 2’ diameter now and my spread is much wider. One of these days I’m gonna fire up a tight set of measurements on this configuration to compare maybe. I just like listening now it sounds so good.

 

[Steve Dallas]

Dirac just put ART on sale 15% off, which amounted to a little over $50 savings, so I finally joined the club with my Denon 4800. No idea when I will be able to take measurements and give it a spin as I have to wait for the family to depart the house for a few hours.

 

[EWL5]

Dirac just put ART on sale 15% off, which amounted to a little over $50 savings, so I finally joined the club with my Denon 4800. No idea when I will be able to take measurements and give it a spin as I have to wait for the family to depart the house for a few hours.

Not the best discount from Dirac but finding "quiet time" may sometimes be priceless!

 

[RenPa]

sounds better to your ears as well?

Yes. But I am Always suspicious of placebo effect though!

Now a stupid question about sub 1 LFE and the 120Hz filter applied to it by Dirac after I saw the previous post.
See the blue sub 1 filter continue after the 150Hz doted line and to my eye it seems like a 120Hz LPF. The dark red filter of the sub 2 STOP at 150Hz hitting a wall. Is it possible that the infamous 120Hz filter at LFE channel is a variable of the algo to ovewrite the 150 limit? Or am I delusionnal?

 

[TimoJ]

Is it possible that the infamous 120Hz filter at LFE channel is a variable of the algo to ovewrite the 150 limit?

No. Test with REW.

 

[RenPa]

No. Test with REW.

The force is with you, you are a jedi not a Ninja...

I measured my 2 subs each with front right disconnected and each with LFE channel front right disconnected. Yellow is the sub 1 LFE measurement with LFE channel in REW BLUE are the 3 remaining measurements

CQFD

 

[RenPa]

I measured with an acoustic timing reference . Is using alignment tool could be usefull? tell me.

I sum subD 1 LFE+ subD 1 (front right disconnected) = pink let's call it total subD 1

I summed subD 1 LFE + sub left (fr disconnected) = red let's call it total sub left

Am I wrong to think that all bass should be the results of:
Total subD1 + sub left (fr disconnected) = orange ? not so good...

On the contrary, subD 1LFE + sub left LFE = green nice!
Why? Does it proves that bass are not correctly managed? Am I Am I confused?

 

[Keith_W]

I measured with an acoustic timing reference . Is using alignment tool could be usefull? tell me.

Whenever you want to sum two measurements, you MUST have a timing reference. Given that these are measurements where neither microphone nor speaker have been moved, an acoustic timing reference is perfectly adequate.

HOWEVER - I must warn about inconsistencies with REW's timing measurements. I don't like automatic timing measurements of any kind - basically the algorithm looks for any deviation greater than a set level from the noise floor as the start of the impulse. This is especially problematic with subwoofers, since (1) the rise time is very slow, and (2) the signal-noise ratio is typically poor. If you are taking subwoofer timing measurements, ALWAYS repeat the measurement a few times and check that the result is the same every time. If it isn't, you have a problem. It can usually be fixed by attempting to improve the SNR.

If you attempt to sum two measurements which have wonky timing references, the result will be very strange. So you must have absolute confidence in your timing reference before you attempt to sum them.

As for this set of measurements that you posted, you have to understand how ART works before attempting to interpret the measurement. ART uses multiple speakers to obtain the flattest response possible with the fastest decay. Individual speaker measurements only show you what ART is asking that individual speaker to do. To get an idea of what ART is doing, you need to measure all the speakers together. Or you could manually sum each individual measurement (with a correct timing reference!) to see the response.

 

[Flak]

Here's some good news for standalone PC and Mac Configurations

Dirac has added Dirac Live Active Room Treatment to Dirac Live Processor, the PC and Mac software previously known as Dirac Live Room Correction Suite. The software applies room correction before audio reaches the sound system and can be used in computer-based playback setups, including compatible DAWs and media players through VST, VST3, AAX, and AU plug-in formats.

With this update, Dirac Live Processor supports Dirac Live Room Correction, Dirac Live Bass Control, and Dirac Live Active Room Treatment in a single software platform. Active Room Treatment builds on room correction by using the speaker system to help manage room resonances and sound decay, with the aim of improving clarity, control, and overall listening consistency.

(No Bass Control license is required for ART if no subwoofers are used)

Dirac Live Active Room Treatment is available in Dirac Live Processor from June 30, 2026. Existing Dirac Live Room Correction Suite users will be transitioned to Dirac Live Processor with their current licenses and settings. Licenses are available from Dirac with an optional 14-day free trial.

Existing users of Dirac Live Room Correction Suite will automatically be transitioned to Dirac Live Processor together with their current licenses and settings. Licenses can be purchased with an optional 14-day free trial at www.dirac.com.