Active Room Treatment (ART) by Dirac

[dlaloum]

Given that ART is still relatively new to me, I can't say that sometimes I don't get that feeling - where is my ribeye fat greasy bass steak. Find myself usually with +8 or 10dB bass curves and unfortunately craving more . But then quickly realize that new diet is so much better than the old one.

Also, think that ART finally made my system obsolete. I would not recommend to go with big speakers all around any more. I did some tests and not much difference if I roll off my 6 surrounds with 10" woofers at 50hz or 80hz. Pretty much the same thing. Number of subs is room dependent and seems that it is the best bang for the buck. I would also keep LCR beefy as they do most of the work.

So what you are saying is more subs less sats?

Drop full range speakers and get more subs? - but isn't that exactly what people have been doing the last 20 years?

Seems to me that ART is showing the way to achieve great results with full range base layer, and a minimum number of subs...

on the other hand, what ART is achieving could be said to be mainly getting the best out of what you have there - then the debate becomes, when revamping your setup, what is the optimal configuration? (looking at all the various value parameters... $, WAF, Performance, longevity etc...)

 

[dlaloum]

I don't think the table shows that at all. According to the article, it is meant to "show the subwoofer’s clean peak SPL before heavy distortion sets in". Had the SPL been held constant at, say, 83dB, we would likely see a precipitous drop in distortion as the frequency goes up. There may still be sweet spots with lower distortion, but we would likely see an overall drop.

Yes - I would like to see more detailed sub THD measurements... will need to search further - especially given I never expect to need 110db in my listening space! My normal listening level is 72db, peak 92db, bass peak with +10db allowance 102db... I guess I do want 110db figures given I want assurance that a 102db peak would be clean and distortion free... but none of the sub data I have seen shows distortion at the 110db peak within what I would consider acceptable levels!

And I would very much like to see what the THD figures for subs are at around my 3 key figures... 72/92/102db (or close thereto)

My subjective experiences tend towards sealed subs and less is more in the bass area... (not a fan of the +10db Harman curves... I guess that means I'm part of the 40% that don't prefer the Harman curve) - I have a feeling the THD is a key parameter that is given far too little attention in subs...

And when Dirac is "forming" the wave environment in the room, what is the impact of THD on that process?
Certainly it is an imperfection/flaw... but how serious is its impact, and to what level can its performance be improved by using lower distortion subs?

With things like noise cancelling (yes I know Dirac doesn't do noise cancelling... but it is a similar area of wave tech) - THD can substantially reduce the effectiveness of the end result... and create some unwanted audible artifacts. I have a feeling this may also be the case with Dirac. (could this be related also to the reason Dirac have limited ART to 150Hz... ?)

 

[OverTheAir]

Yes - I would like to see more detailed sub THD measurements... will need to search further - especially given I never expect to need 110db in my listening space! My normal listening level is 72db, peak 92db, bass peak with +10db allowance 102db... I guess I do want 110db figures given I want assurance that a 102db peak would be clean and distortion free... but none of the sub data I have seen shows distortion at the 110db peak within what I would consider acceptable levels!

And I would very much like to see what the THD figures for subs are at around my 3 key figures... 72/92/102db (or close thereto)

My subjective experiences tend towards sealed subs and less is more in the bass area... (not a fan of the +10db Harman curves... I guess that means I'm part of the 40% that don't prefer the Harman curve) - I have a feeling the THD is a key parameter that is given far too little attention in subs...

And when Dirac is "forming" the wave environment in the room, what is the impact of THD on that process?
Certainly it is an imperfection/flaw... but how serious is its impact, and to what level can its performance be improved by using lower distortion subs?

With things like noise cancelling (yes I know Dirac doesn't do noise cancelling... but it is a similar area of wave tech) - THD can substantially reduce the effectiveness of the end result... and create some unwanted audible artifacts. I have a feeling this may also be the case with Dirac. (could this be related also to the reason Dirac have limited ART to 150Hz... ?)

The old Data-Bass web site included many tests of retail subwoofers and also raw drivers that Josh mounted in sealed enclosures for testing purposes. He stopped updating with new tests many years ago but what remains is still relevant in showing how subs differ in performance across a number of different metrics including THD. For starters a key differentiator is what the driver was designed for and then how it is used in an enclosure.

In any event, this link should take you to a test he did of the 21" B&C 21DS115-4 driver in a sealed enclosure. If you scroll down you will see that that this driver, at 105dB output is at 1.5% THD or less from 40-200Hz, he didn't take the test any higher than 200Hz. 110dB output isn't much higher.

 

[theREALdotnet]

And when Dirac is "forming" the wave environment in the room, what is the impact of THD on that process?
Certainly it is an imperfection/flaw... but how serious is its impact, and to what level can its performance be improved by using lower distortion subs?

Yes, that's a good question. Somebody with ART might want to measure this. In the case of plain DLBC, where only time alignment and power distribution based on capability is applied, I would expect distortion to drop as more subwoofers are added.

 

[dlaloum]

The old Data-Bass web site included many tests of retail subwoofers and also raw drivers that Josh mounted in sealed enclosures for testing purposes. He stopped updating with new tests many years ago but what remains is still relevant in showing how subs differ in performance across a number of different metrics including THD. For starters a key differentiator is what the driver was designed for and then how it is used in an enclosure.

In any event, this link should take you to a test he did of the 21" B&C 21DS115-4 driver in a sealed enclosure. If you scroll down you will see that that this driver, at 105dB output is at 1.5% THD or less from 40-200Hz, he didn't take the test any higher than 200Hz. 110dB output isn't much higher.

Very nice - thank you for posting that.

Indeed that level of distortion would be perfectly acceptable...

However on looking at the details of the measurements there...

at 105db, 20Hz - the chart shows roughly 10% distortion, it's only above 50Hz that the distortion drops to around that 1.5% figure

Given the main distortion is 2nd harmonic... the distortion at 40Hz and below may not be audible - and therefore the sub may not be localisable (and above 50Hz the distortion may be low enough not to be an issue...)

Lower distortion designs are certainly possible (and available) - but most aren't low distortion designs.

Here is one that is:

REL Acoustics HT/1510 Predator II Sealed Subwoofer Review

The REL Acoustics HT/1510 Predator II is a 1,000 watt, 15” woofer, sealed bruiser. REL touts their subwoofers as being fast & musical. We evaluate that claim with measurements & associated trade offs.

www.audioholics.com

But not unexpectedly... it sacrifices high SPL to achieve low THD

 

[dlaloum]

Yes, that's a good question. Somebody with ART might want to measure this. In the case of plain DLBC, where only time alignment and power distribution based on capability is applied, I would expect distortion to drop as more subwoofers are added.

Good point... add more subs, so each is individually less strained, and therefore running within its low distortion (or lower...) envelope...

 

[Oddball]

So what you are saying is more subs less sats?

Drop full range speakers and get more subs? - but isn't that exactly what people have been doing the last 20 years?

Seems to me that ART is showing the way to achieve great results with full range base layer, and a minimum number of subs...

on the other hand, what ART is achieving could be said to be mainly getting the best out of what you have there - then the debate becomes, when revamping your setup, what is the optimal configuration? (looking at all the various value parameters... $, WAF, Performance, longevity etc...)

Based on my experience, subs are the most important ART ingredient. They contribute the most to all other speaker groups and none of the "full range" speakers (aka not active) can't really support significantly in 20-50hz range. Looks like support beyond 50hz range is also well handled by subs so really no need to reinforce with full range speakers - provided sufficient sub power.

One should still "size" the speakers based on the SPL and distortion expectations - which might loop back to full range towers for bed channels, but not for support reasons. So, if people are shopping, I would recommend sub department as the first stop. That might not be feasible for some due to variety of reasons, so with 1 or 2 subs one might still hear/measure improvement from full range speakers in the bed channel.

I am not going to scrape my setup just because it is overkill for ART. It was put together before ART and with different considerations - and still works great. But you generally don't really need 6x13.8" drivers and 4.5kW of apms on the subs (unless you have a large room and indeed aim for low THD) and then on top of it 6x10" surround drivers to support above 50hz. Disclaimer: every room is different so this is not a specific statement.

 

[EWL5]

Based on my experience, subs are the most important ART ingredient. They contribute the most to all other speaker groups and none of the "full range" speakers (aka not active) can't really support significantly in 20-50hz range. Looks like support beyond 50hz range is also well handled by subs so really no need to reinforce with full range speakers - provided sufficient sub power.

One should still "size" the speakers based on the SPL and distortion expectations - which might loop back to full range towers for bed channels, but not for support reasons. So, if people are shopping, I would recommend sub department as the first stop. That might not be feasible for some due to variety of reasons, so with 1 or 2 subs one might still hear/measure improvement from full range speakers in the bed channel.

I am not going to scrape my setup just because it is overkill for ART. It was put together before ART and with different considerations - and still works great. But you generally don't really need 6x13.8" drivers and 4.5kW of apms on the subs (unless you have a large room and indeed aim for low THD) and then on top of it 6x10" surround drivers to support above 50hz. Disclaimer: every room is different so this is not a specific statement.

Much as I'd like to dip my feet into ART, I feel my setup would not benefit. I have only one SW with no plans to buy more (wife is apparently allergic to LFE) and most of my speakers are ceiling speakers and most likely not good candidates to support each other. If I had more conventional speakers (even bookshelves), then it would be more ART-friendly, I think.

 

[peng]

Much as I'd like to dip my feet into ART, I feel my setup would not benefit. I have only one SW with no plans to buy more (wife is apparently allergic to LFE) and most of my speakers are ceiling speakers and most likely not good candidates to support each other. If I had more conventional speakers (even bookshelves), then it would be more ART-friendly, I think.

I feel the same, and it obviously depends mainly on one's room. Some rooms don't even benefit much from RC/EQ, so it all depends. For one to decide, I really think the freeware REW would be their friends, before spending the money and time that they may not have to. I only use the PC version so I can't even try ART, but if they eventually offer the option, I would do the free trial (if they offer it) to see if the improvements are worth it. Regardless, I believe in the theoretical benefits.

 

[Oddball]

Much as I'd like to dip my feet into ART, I feel my setup would not benefit. I have only one SW with no plans to buy more (wife is apparently allergic to LFE) and most of my speakers are ceiling speakers and most likely not good candidates to support each other. If I had more conventional speakers (even bookshelves), then it would be more ART-friendly, I think.

Valid point.

Just as a side-note, perhaps wife never heard ART LFE. My oldest daughter hated the LFE before ART so had a special present for her. She is now much happier with ART and does not need her own preset.

 

[smygolf]

Best practice with Dirac has always been to measure the result with REW.
Now with Art and if you have medium size or smaller fronts it feels more important then ever.
My guess is that Dirac "knows" what it is doing, but as user you can screw it up pretty bad.

As a test i added support down to 30hz / -18dB on my fronts.
They handle 105dB @ ~25hz without any issues.
But now they got overloaded close to ref volume and beyond.

So if you own something that start to fall at 80hz - i cant see that Dirac itself ads support below, but you can do it.
And things might go Puff.....

 

[kawauso]

Based on my experience, subs are the most important ART ingredient. They contribute the most to all other speaker groups and none of the "full range" speakers (aka not active) can't really support significantly in 20-50hz range. Looks like support beyond 50hz range is also well handled by subs so really no need to reinforce with full range speakers - provided sufficient sub power.

One should still "size" the speakers based on the SPL and distortion expectations - which might loop back to full range towers for bed channels, but not for support reasons. So, if people are shopping, I would recommend sub department as the first stop. That might not be feasible for some due to variety of reasons, so with 1 or 2 subs one might still hear/measure improvement from full range speakers in the bed channel.

I am not going to scrape my setup just because it is overkill for ART. It was put together before ART and with different considerations - and still works great. But you generally don't really need 6x13.8" drivers and 4.5kW of apms on the subs (unless you have a large room and indeed aim for low THD) and then on top of it 6x10" surround drivers to support above 50hz. Disclaimer: every room is different so this is not a specific statement.

ART integrates the subwoofers remarkably well.

In my current mixed setup with one JTR and two HSU subs, ART performs far better than Bass Control ever did — it’s a huge step up.

That alone gives me motivation to expand the system.

I already know ART can handle additional subs with lower performance without issues,
but since I’m in deep anyway, I’m planning to add another JTR next year.

 

[Flak]

『When Dirac ART Freed My KEF Room』

Oct 25, 2025 · User ReviewTable of Contents Measurements & First Impressions Li…

ameblo.jp

 

[kawauso]

So what you are saying is more subs less sats?

Drop full range speakers and get more subs? - but isn't that exactly what people have been doing the last 20 years?

Seems to me that ART is showing the way to achieve great results with full range base layer, and a minimum number of subs...

on the other hand, what ART is achieving could be said to be mainly getting the best out of what you have there - then the debate becomes, when revamping your setup, what is the optimal configuration? (looking at all the various value parameters... $, WAF, Performance, longevity etc...)

I’ve been testing ART in four different rooms.
In my experience, using full-range speakers as support sometimes introduces a bit of “smearing” or tonal coloration.
But when I use subwoofers as support, even with all of them active and the support range extended up to 150 Hz, there are no audible side effects.

That might be because my setups are relatively small rooms with short listening distances, and the subs are well placed and time-aligned.
In other environments, I can imagine that using satellites as support and only one—or even zero—subwoofers could still work perfectly fine.

How much low-end quality is “enough” really depends on the listener… and, of course, on the WAF.
In my case, the WAF for audio is easy—because I am the wife (kawauso is female).

 

[masterhw]

One question - why is ART limited to 150Hz. Would it not be advantageous to go up to the room transition point?
Are cancellation waves a problem/audible >150Hz?

I would assume the directionality of the wave becomes an issue above that frequency. You might start to notice where the anti wave is coming from (particularly at the non-MLP) and it might be distracting.

 

[Oddball]

I’ve been testing ART in four different rooms.
In my experience, using full-range speakers as support sometimes introduces a bit of “smearing” or tonal coloration.
But when I use subwoofers as support, even with all of them active and the support range extended up to 150 Hz, there are no audible side effects.

That might be because my setups are relatively small rooms with short listening distances, and the subs are well placed and time-aligned.
In other environments, I can imagine that using satellites as support and only one—or even zero—subwoofers could still work perfectly fine.

How much low-end quality is “enough” really depends on the listener… and, of course, on the WAF.
In my case, the WAF for audio is easy—because I am the wife (kawauso is female).

Well, we are glad to have diversity so all the better.

I have not done any conclusive testing (and might not for a while), but some preliminary results also support the statement that less, but adequate, resources might work "better". I think ART gets a bit confused with my setup where there is so much to tap into. It's not smearing, but it takes a bit longer to plug in the bed channels in the low end chain than the subs. Perhaps still worth it, but under consideration.

Now, whatever setting I have in ART it returns sub curve within +/- 1.5 dB, and for the love of god and all the saints, I can't really "hear" which curve is better. I think that with larger system we have hit the point of completely idiot proof setup.

 

[fzst]

I’ve been testing ART in four different rooms.
In my experience, using full-range speakers as support sometimes introduces a bit of “smearing” or tonal coloration.
But when I use subwoofers as support, even with all of them active and the support range extended up to 150 Hz, there are no audible side effects.

That might be because my setups are relatively small rooms with short listening distances, and the subs are well placed and time-aligned.
In other environments, I can imagine that using satellites as support and only one—or even zero—subwoofers could still work perfectly fine.

How much low-end quality is “enough” really depends on the listener… and, of course, on the WAF.
In my case, the WAF for audio is easy—because I am the wife (kawauso is female).

So you don't even let the LCRs support each other?

 

[dlaloum]

I’ve been testing ART in four different rooms.
In my experience, using full-range speakers as support sometimes introduces a bit of “smearing” or tonal coloration.
But when I use subwoofers as support, even with all of them active and the support range extended up to 150 Hz, there are no audible side effects.

That might be because my setups are relatively small rooms with short listening distances, and the subs are well placed and time-aligned.
In other environments, I can imagine that using satellites as support and only one—or even zero—subwoofers could still work perfectly fine.

How much low-end quality is “enough” really depends on the listener… and, of course, on the WAF.
In my case, the WAF for audio is easy—because I am the wife (kawauso is female).

I apologise for the inherent assumptions in WAF... it should indeed be "Partner Acceptance Factor" ... but common usage makes that eve more awkward!

 

[kawauso]

So you don't even let the LCRs support each other?

I actually prefer having mutual support between the Left and Right channels.
Any speaker that has a paired partner (like L/R or SL/SR) is set up with mutual support in my system.

On the other hand, single-position speakers like the Center channel are supported only by the subwoofers.

Many of my full-range speakers also help support the LFE group.

If you don’t mind sharing, I’d be curious to hear what kind of difference this configuration would make in your setup — and whether it would sound like something you’d prefer.

My guess is that in a larger room, the number of support speakers probably doesn’t change things much, either way.

 

[smygolf]

I actually prefer having mutual support between the Left and Right channels.
Any speaker that has a paired partner (like L/R or SL/SR) is set up with mutual support in my system.

On the other hand, single-position speakers like the Center channel are supported only by the subwoofers.

Many of my full-range speakers also help support the LFE group.

If you don’t mind sharing, I’d be curious to hear what kind of difference this configuration would make in your setup — and whether it would sound like something you’d prefer.

My guess is that in a larger room, the number of support speakers probably doesn’t change things much, either way.

Before ART i always used L/R as fullrange when listen to music.
Even with just a slight raise the lower region 2 x 4645C was to much - not so much with ART.

But for the sake of it i will try again and see how the support looks etc