Active Room Treatment (ART) by Dirac

[JeffH]

Just want to add to the importance of Directional Bass from a few posts ago. I'm sure some of you know this but bass channels should be preserved, not summed to mono as is common practice. It's true that bass below 80(ish)Hz isn't directional as far as dispersion is concerned but its phase is. In the '90s I set up an experiment with a rather nice system that had two 20 Hz capable subs that could be switched between stereo and mono. I had a CD that was well recorded in a large concert hall. In that recording when they switched on the microphones but before the orchestra started playing you has a very real feeling of the space, the volume of the room, super cool. When you switched the subs to mono it collapsed. I demoed this to a few people that were arguing that it didn't matter and changed their minds because it was so obvious. I was recently at an AES meeting with JJ Johnston speaking and he, rather snarkily, talked about the importance of this. This guy is a heavy weight in audio perception research, IEEE Fellow, AES Fellow, many awards.

 

[EngineerNate]

My intention is to eventually have mains that are flat to 20hz at reference level and then add 2-4 additional subs for ART to work with. Should be best of both worlds.

 

[Oddball]

If you guys are insisting on even more directional bass, might be worth it to send request to Dirac.

Individual "directional" speakers will be challenged by placement so unlikely to have uniform FQ response without support, plus will wreak havoc in decay even if they can deal with SPL. Once you start supporting them you are into ART support matrix. If done properly, this should be done in ART algo. From what I understand, it works differently now.

Adding "prefer the source channel" option might help with pulling more weight to the specific capable channel and not necessarily to what is the most capable support or most convenient support for ART. I am quite happy with what I have now and kind of busy to split hairs, but this would be my suggestion if you want to do it right. Could be quite easy for Dirac to accomodate in best case scenario.

 

[thcdru2k]

Low bass may not be point localizable but I do wonder if our hearing is sensitive to the phase and timing relationships in the crossover where fronts and subs overlap. Multiple subs distributed around the room can give excellent modal control and seat consistency but it may trade away some of the sense that the bass in that transition band is coupled to the mains which could reduce spaciousness. If that's true it seems like a potential tradeoff with ART style correction that redistributes bass across multiple sources.

 

[EngineerNate]

If you guys are insisting on even more directional bass, might be worth it to send request to Dirac.

Individual "directional" speakers will be challenged by placement so unlikely to have uniform FQ response without support, plus will wreak havoc in decay even if they can deal with SPL. Once you start supporting them you are into ART support matrix. If done properly, this should be done in ART algo. From what I understand, it works differently now.

Adding "prefer the source channel" option might help with pulling more weight to the specific capable channel and not necessarily to what is the most capable support or most convenient support for ART. I am quite happy with what I have now and kind of busy to split hairs, but this would be my suggestion if you want to do it right. Could be quite easy for Dirac to accomodate in best case scenario.

To be clear my intent is mostly to make my mains support the subs as effectively as they can. Guess that mainly means flat to 50hz with low distortion.

 

[thcdru2k]

Filter type and slope are important to consider when having mains support subs or vice versa. Too shallow a slope (12db/octave) and there is too much overlap where both radiate strongly and room dominates. Steeper slopes (48db/octave) can reduce blending and sound like they are taking turns. I haven't personally heard ART yet but I am very curious how it affects things. Kind of makes sense it optimizes the room its called active room treatment after all where as traditional crossover alignment optimizes source coherence. I would guess the benefits of ART (modal control and seat to seat consistency) out weigh the benefit of maintaining source coherence except for very edge cases or hyper focus optimization for a single listener.

 

[Oddball]

Filter type and slope are important to consider when having mains support subs or vice versa. Too shallow a slope (12db/octave) and there is too much overlap where both radiate strongly and room dominates. Steeper slopes (48db/octave) can reduce blending and sound like they are taking turns. I haven't personally heard ART yet but I am very curious how it affects things. Kind of makes sense it optimizes the room its called active room treatment after all where as traditional crossover alignment optimizes source coherence. I would guess the benefits of ART (modal control and seat to seat consistency) out weigh the benefit of maintaining source coherence except for very edge cases or hyper focus optimization for a single listener.

There are no crossovers you can set with ART - so you get what is based on you support groups, range, intensity and what ART thinks of the whole setup. So there are no slopes. You get all supporting speakers contributing to some extent that then summons into overall channel response. So in my case, my fronts would play part of what they are supposed to, other parts played by subs, FW and surrounds. All together you get the front channel response. You can look up Dirac filters and see what they are doing in that respect as approximation of what is really happening.

Optimizing the response for the seating area is same is in previous room correction systems. It will depend on the measured area. If you have a sofa e.g. you can take measurements for all 3 or 4 seats, or just for and around MLP.

 

[thcdru2k]

There are no crossovers you can set with ART - so you get what is based on you support groups, range, intensity and what ART thinks of the whole setup. So there are no slopes. You get all supporting speakers contributing to some extent that then summons into overall channel response. So in my case, my fronts would play part of what they are supposed to, other parts played by subs, FW and surrounds. All together you get the front channel response. You can look up Dirac filters and see what they are doing in that respect as approximation of what is really happening.

Optimizing the response for the seating area is same is in previous room correction systems. It will depend on the measured area. If you have a sofa e.g. you can take measurements for all 3 or 4 seats, or just for and around MLP.

Yeah that makes sense. ART is likely the better solution for most situations. Where as tight crossover exact coupled integration might sound more natural in very specific situations but this would require proper A/B control with multiple listeners to prove.

 

[middlemarch]

Just thought I'd throw in a cautionary note for any newly arrived Dirac ART users. Ensure your measurement environment is identical to your listening environment!

My basement listening room is adjacent to my shop area, separated by a door. The shop has my HVAC system, so I close that door during calibration to remove that noise source. At all other times it is open (because kitty litter box...). I've never been completely satisfied with the bass during my listening sessions, tried lots of things.

Last night I closed that door.

Instant Nirvana. No more bass issues, best it's ever been, by a lot! I listened for over 5 hours, loving every minute. Wife came down several times expressing concern about kitty potty needs. Told her to tell him to hold it.

So I'm getting ready to re-calibrate with the door open. It's that or move the kitty litter box. Hmmmm...

In the meantime I ordered a new 15" HSU VTF-3 MK5 to replace my 25 year old 12" VTF-3 Mk1, thinking that might be the weak point. Oh well, newer is better, right?

 

[Mort]

15" HSU VTF-3 MK5 to replace my 25 year old 12" VTF-3 Mk1, thinking that might be the weak point. Oh well, newer is better, right?

Bigger is definitely better! Well so they say.

 

[kawauso]

If you guys are insisting on even more directional bass, might be worth it to send request to Dirac.

Individual "directional" speakers will be challenged by placement so unlikely to have uniform FQ response without support, plus will wreak havoc in decay even if they can deal with SPL. Once you start supporting them you are into ART support matrix. If done properly, this should be done in ART algo. From what I understand, it works differently now.

Adding "prefer the source channel" option might help with pulling more weight to the specific capable channel and not necessarily to what is the most capable support or most convenient support for ART. I am quite happy with what I have now and kind of busy to split hairs, but this would be my suggestion if you want to do it right. Could be quite easy for Dirac to accomodate in best case scenario.

In kawauso’s experience, ART does not weaken the sense of bass directionality; if anything, it tends to strengthen it.
No special settings are required to perceive this. I think this matches Oddball’s impression too.

There are no crossovers you can set with ART - so you get what is based on you support groups, range, intensity and what ART thinks of the whole setup. So there are no slopes. You get all supporting speakers contributing to some extent that then summons into overall channel response. So in my case, my fronts would play part of what they are supposed to, other parts played by subs, FW and surrounds. All together you get the front channel response. You can look up Dirac filters and see what they are doing in that respect as approximation of what is really happening.

Optimizing the response for the seating area is same is in previous room correction systems. It will depend on the measured area. If you have a sofa e.g. you can take measurements for all 3 or 4 seats, or just for and around MLP.

Since my listening position is usually more or less the same, I calibrate over a relatively small area (about 30–50 cm from the center, 13 measurement points). Even so, it sounds good over a surprisingly wide area. The sound remains good even in extreme positions, such as close to the walls.

 

[kawauso]

Just thought I'd throw in a cautionary note for any newly arrived Dirac ART users. Ensure your measurement environment is identical to your listening environment!

My basement listening room is adjacent to my shop area, separated by a door. The shop has my HVAC system, so I close that door during calibration to remove that noise source. At all other times it is open (because kitty litter box...). I've never been completely satisfied with the bass during my listening sessions, tried lots of things.

Last night I closed that door.

Instant Nirvana. No more bass issues, best it's ever been, by a lot! I listened for over 5 hours, loving every minute. Wife came down several times expressing concern about kitty potty needs. Told her to tell him to hold it.

So I'm getting ready to re-calibrate with the door open. It's that or move the kitty litter box. Hmmmm...

In the meantime I ordered a new 15" HSU VTF-3 MK5 to replace my 25 year old 12" VTF-3 Mk1, thinking that might be the weak point. Oh well, newer is better, right?

I think you’ll be surprised how much better it can sound if you calibrate with the door open. If you sometimes listen with the door closed as well, you can save separate calibrations using Dirac slots or DENON presets.
So you ordered the HSU VTF-3 MK5. I’m using two of them myself. And the VTF-3 Mk1 really did an amazing job for 25 years—that’s an impressive run.

 

[gepgep]

Something I just found while messing with REW and one of my ART calibrations

I thought I would investigate the effect of support levels on my subs

My subs are supported by the FL & FR 50-150 -18dB which ART set as default

I turned the subs off to see what the Fronts were contributing at different support levels

I tested -18, -12 and -6 support levels
To my surprise all 3 REW graphs were virtually identical.

In addition the fronts appear to carry almost all the load above 70Hz ie when I turned the subs back on the graph above 70hz was unchanged

Unfortunately disconnecting my Fronts to see what subs only were doing isn't easy so I haven't yet measured sub only

 

[Oddball]

Something I just found while messing with REW and one of my ART calibrations

I thought I would investigate the effect of support levels on my subs

My subs are supported by the FL & FR 50-150 -18dB which ART set as default

I turned the subs off to see what the Fronts were contributing at different support levels

I tested -18, -12 and -6 support levels
To my surprise all 3 REW graphs were virtually identical.

In addition the fronts appear to carry almost all the load above 70Hz ie when I turned the subs back on the graph above 70hz was unchanged

Unfortunately disconnecting my Fronts to see what subs only were doing isn't easy so I haven't yet measured sub only

Not surprised at all and my experience is consistent with yours. From what I understand, the levels of support will tell ART how hard to try. If it does not need to try hard based on setup it might not be fundamentally different - but on the other hand it might in another setup.

Did you check what is ART showing in terms of filters vs what you got out of REW?

 

[JRowland]

@Oddball I must thank you for lighting the fire under me to re-try a rear sub in my system. Best outlay of cash I’ve made since the Dirac license.

 

[Oddball]

@Oddball I must thank you for lighting the fire under me to re-try a rear sub in my system. Best outlay of cash I’ve made since the Dirac license.

I am sorry for the cash outlay, but very glad you got your money's worth. Older solutions to address bass decay like double bass array were based on physics and still are. While ART defies physics to some extent, it will still gladly take help from the old principles.

 

[Flak]

ART, use case definition, and set up guidelines:

https://cdn.sanity.io/files/tq9n76g3/production/1a283645e35a349971fc06ae08d1b53cd93afcb8.pdf

 

[jhaider]

The shop has my HVAC system, so I close that door during calibration to remove that noise source. At all other times it is open (because kitty litter box...). …

So I'm getting ready to re-calibrate with the door open. It's that or move the kitty litter box. Hmmmm...

Cut a cat door into the door between your shop and listening room, e.g. https://a.co/d/0eCIUg2Y.

 

[90sNintendoKid]

I thought it would be helpful to share something I think I was doing wrong in my initial Dirac measurements and I suspect was affecting Dirac ART in particular.

I wasn't following the instructions in the Dirac app about level matching the speakers in subwoofers.

This time, in the Dirac app, I set the mic gain to max and the avr volume to -15db and adjusted the volume of the speakers to -18db and adjusted the volume on my subwoofers (not in the Dirac app) to -12db.

Before, DLBC had better bass and was following my lfe target curve better than ART.

Now, ART has similar bass performance than DLBC.

I've been obsessing too much about getting things perfect so I don't have any post EQ measurements to verify this and I can provide them at a later time, I've been listening to music and watching movies.

 

[Vincentponcet]

No doubt that these Focal subs will sound very resolving but their design choice needs extra attention. That 18Hz HPF is absolutely necessary to protect the woofers from bottoming out. Focal should at least offer a matching power amp with built in EQ to avoid clicking and popping problems.

My 18yo Focal Scala Utopia also sound awesome but they're only allowed to do 50Hz support.

Why 50Hz ? The datasheet says 27Hz at -3dB. I understood the rule was +10Hz above -3dB if properly powered. Recommend amp is 50-500W.
Did you tried with f low at 40hz ?

Scala Utopia Evo - 3-way floor-standing speaker | Focal

Scala Utopia Evo, 3-way floor-standing speaker

www.focal.com