Active Room Treatment (ART) by Dirac

[EWL5]

I’ve just received a refund confirmation from my credit card company. Honestly, I’m relieved.

Scammers are getting more sophisticated, though the profession is as old as time!

 

[Oddball]

Scammers are getting more sophisticated, though the profession is as old as time!

I found that the most used scam in recent years was that someone stole your credit/debit card details and used them to do damage. Disputing these is sometimes not easy and takes time and effort.

Couple of ways to get around that is to keep low balances, block the card when not intending to be used, or the latest and greatest to issue new "clone" cards (but with different number - at no fee) through your banking app. I use a combination including the clone card gets canceled and reissued (with new number) every month. Glad that credit card companies finally figured out this is a cheaper way to deal with the problem. It's just few clicks in the system, no real cost involved and your card is dead meat for the thieves after 30 days (or whatever interval you prefer).

EDIT: Clones are just electronic cards, so can't use them as physical cards.

 

[DutchJay]

ART doesn’t really use a traditional “crossover” concept.
It works more like a shared low-frequency workload among all bass-capable speakers.
So, your mains don’t “lose” bass the way they would with Bass Control.
They still contribute, but the system may let the support speakers take more of the load if that gives smoother response.
It’s not an equal 25/25/25/25 split or anything like that, though.
Just let ART do its thing and don’t worry too much about that part.

If you want to add subwoofers, the best cost-effective option is to start with one high-performance sub.
You can always add more later, but a single strong unit already makes a huge improvement with ART.

FYI: I went from ART with 1 to 2 subs and 2 subs is a clear improvement with ART as well

Post in thread 'Dirac ART is now running on beta FW for Denon Xx800H AVRs!' https://www.audiosciencereview.com/...a-fw-for-denon-xx800h-avrs.46616/post-2477175

 

[DutchJay]

The measurements are quite unambivalent - ART achieves things that otherwise can only be achieved through architectural design of a dedicated listening, and/or through extensive room treatment, which is often completely impractical for many of us.

ART shows that it is possible to electronically improve decay time in the room, and by eliminating a large part of that reflected sound overhang, improve the clarity of our system.

It certainly isn't cheap, but the type and level of improvement that it provides would otherwise cost multiple times that amount, so if you value that improvement, it is fantastic value. People spend tens of thousands on acoustic treatments, bass traps, constructing dedicated listening / theatre rooms - and that level of improvement can potentially be achieved for $600... while keeping the setup in your shared purpose living space.

Completely agree, this is what art has done in my living room....

 

[NanoGuidingStar]

The measurements are quite unambivalent - ART achieves things that otherwise can only be achieved through architectural design of a dedicated listening, and/or through extensive room treatment, which is often completely impractical for many of us.

ART shows that it is possible to electronically improve decay time in the room, and by eliminating a large part of that reflected sound overhang, improve the clarity of our system.

It certainly isn't cheap, but the type and level of improvement that it provides would otherwise cost multiple times that amount, so if you value that improvement, it is fantastic value. People spend tens of thousands on acoustic treatments, bass traps, constructing dedicated listening / theatre rooms - and that level of improvement can potentially be achieved for $600... while keeping the setup in your shared purpose living space.

Also, don’t forget how much time ART saves you.

With traditional methods, most people need to spend a huge amount of time, six months, or even more than a year, studying how to treat the acoustics of their room and figuring out how to handle the bass experience without Dirac’s tools. But with ART combined with Bass Control, you only need to lift a finger to instantly achieve a “passing grade” result. From there, you can gradually fine-tune it to 90 points, or even 100, depending on your preferences and available time.

Without these tools, however, just reaching that passing grade baseline is already a tough challenge for ordinary people (especially for office workers like me)

 

[srekal34]

I have been using ART for the last couple of weeks and its a big improvement, but you have to be careful with it - it can easily drive the amp to clipping (Peerless XLS10 sub and Hypex 251 amp - I had to add a very aggressive high pass from 18Hz)

Did anyone notice a strange bug - it seems that ART only properly works when implemented just after measurements - if I do it with measurements, I get flat in room response to 25Hz. However, if I close the app, load the project and fiddle with it- I never get the same bass extension. Feels like a bug in the app.

 

[DutchJay]

I have been using ART for the last couple of weeks and its a big improvement, but you have to be careful with it - it can easily drive the amp to clipping (Peerless XLS10 sub and Hypex 251 amp - I had to add a very aggressive high pass from 18Hz)

Did anyone notice a strange bug - it seems that ART only properly works when implemented just after measurements - if I do it with measurements, I get flat in room response to 25Hz. However, if I close the app, load the project and fiddle with it- I never get the same bass extension. Feels like a bug in the app.

When I was running ART with one sub supported by my mains the amp i use for my mains (hypex nilai stereo) went into protection during the most intense scenes at higher volumes, so yes be carefully

The other bug i don't recognize

 

[Oddball]

ART does not really know the gear you have and it's ultimate limits. When doing sweep it will do it at whatever you set it at, if doing it according to instructions (mid bar) it should come around 70-80dB which is neither low or high volume and will hopefully not test the limits of the gear for SPL. That sweep is not really to determine the limits of the system (which is an interesting point), but more to center AVR/AVP trimms around zero which would provide highest headroom for the system (cuts, boosts). Surprised that ART would not determine support range for the sub correctly. I have much larger subs and they do cut off at 20hz, which ART acknowledges.

Looks like some people are hitting the limits so should in that case probably dial back in terms of range and intensity of the support. There is limit to everything so adjusting to those limits seems appropriate. Don't really have these issues, but lots of drivers and amps in my system and ART seems to share the load smartly so that none of them get into distortion or clipping.

Not sure about the bug either. There was a bug that every time you load the finalized project it will need to recalculate the filters. From what I understand might be resolved now, but did not try to run Dirac in a while. If new bug, best to report to Dirac customer service. They will for sure take it into consideration.

 

[kawauso]

FYI: I went from ART with 1 to 2 subs and 2 subs is a clear improvement with ART as well

Post in thread 'Dirac ART is now running on beta FW for Denon Xx800H AVRs!' https://www.audiosciencereview.com/...a-fw-for-denon-xx800h-avrs.46616/post-2477175

It seems that adding a second Arendal 1S subwoofer produced very good results. In my environment, even a single sub works well, but I would rate it roughly as 90 points with one unit, 95 points with two, 98 points with three, and close to 100 points with four. In a DLBC setup, I would say one sub is around 50 points and two subs about 80 points. The difference between one and two subs was significant there, whereas with ART the difference is not as large.
It is great that you are enjoying playback at such high levels that the Hypex Nilai amplifier even goes into protection mode. Compared to a conventional fixed crossover such as 80 Hz, ART’s behavior may be a bit more demanding on the amplifier. You could either raise the support range slightly to be gentler on the amp, or use BTL mode to increase the available output power.

 

[kawauso]

Also, don’t forget how much time ART saves you.

With traditional methods, most people need to spend a huge amount of time, six months, or even more than a year, studying how to treat the acoustics of their room and figuring out how to handle the bass experience without Dirac’s tools. But with ART combined with Bass Control, you only need to lift a finger to instantly achieve a “passing grade” result. From there, you can gradually fine-tune it to 90 points, or even 100, depending on your preferences and available time.

Without these tools, however, just reaching that passing grade baseline is already a tough challenge for ordinary people (especially for office workers like me)

It looks like NanoGuidingStar is enjoying ART. If you don’t mind,
could you share the environment you’re using it in and some of your experiences with it?

 

[kawauso]

I have been using ART for the last couple of weeks and its a big improvement, but you have to be careful with it - it can easily drive the amp to clipping (Peerless XLS10 sub and Hypex 251 amp - I had to add a very aggressive high pass from 18Hz)

Did anyone notice a strange bug - it seems that ART only properly works when implemented just after measurements - if I do it with measurements, I get flat in room response to 25Hz. However, if I close the app, load the project and fiddle with it- I never get the same bass extension. Feels like a bug in the app.

Regarding the HPF, you may not need to adjust it separately with an external DSP. By adjusting the subwoofer support range instead, for example changing it to 30–150 Hz, you may be able to address the issue more elegantly, and the sound may also become cleaner.
I personally have not encountered a bug where ART only sounds good immediately after measurements. However, I have experienced cases where ART’s effectiveness was significantly degraded when the acoustic information of the space changed after measurement, such as by opening or closing doors.

 

[dlaloum]

It seems that adding a second Arendal 1S subwoofer produced very good results. In my environment, even a single sub works well, but I would rate it roughly as 90 points with one unit, 95 points with two, 98 points with three, and close to 100 points with four. In a DLBC setup, I would say one sub is around 50 points and two subs about 80 points. The difference between one and two subs was significant there, whereas with ART the difference is not as large.
It is great that you are enjoying playback at such high levels that the Hypex Nilai amplifier even goes into protection mode. Compared to a conventional fixed crossover such as 80 Hz, ART’s behavior may be a bit more demanding on the amplifier. You could either raise the support range slightly to be gentler on the amp, or use BTL mode to increase the available output power.

In my use case - with 4 base layer full range speakers in the mix - 1 sub was 50%, 2 subs looks to me like 95% (based on looking at the waterfall charts)... I think a further couple of % could be eked out with another sub (or perhaps with height speakers contributing in the 70Hz to 150Hz range?) - and potentially if my subs or full rangers were changed to speakers that were infrasonic capable, then the exemplary results could potentially be extended to 20hz rather than the current 24hz to 30Hz...

But that is just niggling issues - taking into account the dreaded combination of diminishing returns and exponentially increasing costs for each tiny 1% increment in objective performance - I now feel pretty damn close to where I wanted to be!

Slightly larger height speakers are on my shopping list.... there is a little bit of unevenness in the 70Hz to 150Hz range, which could potentially be improved with more capable height speakers contributing... but given how small that measured unevenness is, that is more of an "I can see the issue on the graph" dissatisfaction, rather than "I can hear a problem" - real issue.

In any case - assuming capable full range base layer speakers, then 2 subs are ample - I really doubt a 3rd sub would change the results perceptibly!

Without full range mains contributing, I think 3 or even 4 subs would be beneficial... (and yes I didn't think that to be the case before I tried ART "hands on").

Furthermore, this is coming from an anti-bass head - who has a target curve of +0db bass, -3db treble (using shelves/handles)

 

[kawauso]

In my use case - with 4 base layer full range speakers in the mix - 1 sub was 50%, 2 subs looks to me like 95% (based on looking at the waterfall charts)... I think a further couple of % could be eked out with another sub (or perhaps with height speakers contributing in the 70Hz to 150Hz range?) - and potentially if my subs or full rangers were changed to speakers that were infrasonic capable, then the exemplary results could potentially be extended to 20hz rather than the current 24hz to 30Hz...

But that is just niggling issues - taking into account the dreaded combination of diminishing returns and exponentially increasing costs for each tiny 1% increment in objective performance - I now feel pretty damn close to where I wanted to be!

Slightly larger height speakers are on my shopping list.... there is a little bit of unevenness in the 70Hz to 150Hz range, which could potentially be improved with more capable height speakers contributing... but given how small that measured unevenness is, that is more of an "I can see the issue on the graph" dissatisfaction, rather than "I can hear a problem" - real issue.

In any case - assuming capable full range base layer speakers, then 2 subs are ample - I really doubt a 3rd sub would change the results perceptibly!

Without full range mains contributing, I think 3 or even 4 subs would be beneficial... (and yes I didn't think that to be the case before I tried ART "hands on").

Furthermore, this is coming from an anti-bass head - who has a target curve of +0db bass, -3db treble (using shelves/handles)

Looking at dlaloum’s graph with only the first subwoofer enabled and many support speakers active, which appears close to a near-default configuration, it certainly looks to be around 50 points. Have you tried an experiment where you turn off only the second subwoofer while keeping your current settings?
In my environment, there is not much difference whether I set the support range of the main front K2 speakers to 30–150 Hz or to 50–150 Hz. Therefore, I believe that only a moderate level of low-frequency capability is sufficient for the base layer speakers.

 

[dlaloum]

Looking at dlaloum’s graph with only the first subwoofer enabled and many support speakers active, which appears close to a near-default configuration, it certainly looks to be around 50 points. Have you tried an experiment where you turn off only the second subwoofer while keeping your current settings?
In my environment, there is not much difference whether I set the support range of the main front K2 speakers to 30–150 Hz or to 50–150 Hz. Therefore, I believe that only a moderate level of low-frequency capability is sufficient for the base layer speakers.

I am planning on doing some further testing when I have the "space" - what with the holiday season, space is at a premium!

My next set of tests, I want to boost both the subs gain - I have them currently at around 50% on their internal amp gain - I estimate that I can get a 10db boost in there without stressing the subs too much, and that will extend their bottom end potentially... (to maintain the target, there will be more cut through the main frequencies, and the roll off will be relatively raised... ) - what I have to watch out for is an increase in THD... keep the THD under 1.5% and I should be OK.

The other thing I want to try, is with all the speakers on the internal amps, lets see what happens when the X4800 is faced with a 1.6ohm load... - it isn't power that is needed, as the SPL's I am aiming for require no more than 16W (est.) peak - but lots of amps get unstable, with dramatic rises in distortion into such difficult loads.... it was nasty with my previous Integra DRX3.4... hence my setup using external power amps for L/C/R.

That will also require running a fresh set of speaker wires, as my current ones use neutrik connectors at the power amp, and I will need to lay down a set with banana plugs for the X4800....

And once I am up and measuring, I will measure the 2 subs seperately and set up a DL single sub config, as well as ART single and ART double sub.... - but given current results, these are more academic, than anything I expect to provide further improvements... just trying to gain a better understanding of ART behaviour... the latest measurements are so good, it is difficult to imagine subsequent tweaks providing an audible improvement... although there may be a measurable improvement - and that is what I will be aiming for.

So a few things to do when I have the place to myself....

 

[kawauso]

I am planning on doing some further testing when I have the "space" - what with the holiday season, space is at a premium!

My next set of tests, I want to boost both the subs gain - I have them currently at around 50% on their internal amp gain - I estimate that I can get a 10db boost in there without stressing the subs too much, and that will extend their bottom end potentially... (to maintain the target, there will be more cut through the main frequencies, and the roll off will be relatively raised... ) - what I have to watch out for is an increase in THD... keep the THD under 1.5% and I should be OK.

The other thing I want to try, is with all the speakers on the internal amps, lets see what happens when the X4800 is faced with a 1.6ohm load... - it isn't power that is needed, as the SPL's I am aiming for require no more than 16W (est.) peak - but lots of amps get unstable, with dramatic rises in distortion into such difficult loads.... it was nasty with my previous Integra DRX3.4... hence my setup using external power amps for L/C/R.

That will also require running a fresh set of speaker wires, as my current ones use neutrik connectors at the power amp, and I will need to lay down a set with banana plugs for the X4800....

And once I am up and measuring, I will measure the 2 subs seperately and set up a DL single sub config, as well as ART single and ART double sub.... - but given current results, these are more academic, than anything I expect to provide further improvements... just trying to gain a better understanding of ART behaviour... the latest measurements are so good, it is difficult to imagine subsequent tweaks providing an audible improvement... although there may be a measurable improvement - and that is what I will be aiming for.

So a few things to do when I have the place to myself....

With the A1H, I was able to drive the KEF Reference 1 Meta to fairly high levels without audible distortion, so I think there is good reason to expect solid performance from the X4800H’s amplifier as well. If it passes the dlaloum test, please make sure to proudly display the certificate of achievement.
One recent experiment I found particularly interesting was a single-subwoofer comparison between a JTR 18-inch and a JBL 15-inch. There is a significant performance gap between these two subwoofers, but when listening to music at around 100 dB SPL, I did not perceive a large difference in sound quality. The infrasonic option was turned off, and the support range was set to 20–150 Hz for the comparison.
When using DLBC, the sound lacked tightness and punch unless the JTR was used, but ART narrowed that performance gap significantly. The placement was different, so it was not a perfectly fair comparison, but even so, the result was quite surprising.

JTR SW

JBL SW

 

[DutchJay]

It seems that adding a second Arendal 1S subwoofer produced very good results. In my environment, even a single sub works well, but I would rate it roughly as 90 points with one unit, 95 points with two, 98 points with three, and close to 100 points with four. In a DLBC setup, I would say one sub is around 50 points and two subs about 80 points. The difference between one and two subs was significant there, whereas with ART the difference is not as large.
It is great that you are enjoying playback at such high levels that the Hypex Nilai amplifier even goes into protection mode. Compared to a conventional fixed crossover such as 80 Hz, ART’s behavior may be a bit more demanding on the amplifier. You could either raise the support range slightly to be gentler on the amp, or use BTL mode to increase the available output power.

Hey, in the end I didn't need to change the support level as this actually only happened in testing and not in real life scenarios. I nearly never watch movies around reference levels. And after I got the second sub I was able to raise the frequency level of the support for the towers so it does not happen anymore

Funny that you rate 1 sub so high. Does the 2nd not give you significantly more deep bass? That was the biggest improvement for me. The other also noticeable but the deep bass and more pressure in the room was striking....

 

[kawauso]

Hey, in the end I didn't need to change the support level as this actually only happened in testing and not in real life scenarios. I nearly never watch movies around reference levels. And after I got the second sub I was able to raise the frequency level of the support for the towers so it does not happen anymore

Funny that you rate 1 sub so high. Does the 2nd not give you significantly more deep bass? That was the biggest improvement for me. The other also noticeable but the deep bass and more pressure in the room was striking....

In my environment, I don’t experience such a clear change when adding more subwoofers. As the number of subs increases, the sound tends to become lighter, more spacious, and more natural, but with fewer subs there is more of a traditional “audio punch,” so either approach works for me.
In my JBL room, all of the base layer speakers are large (three 15-inch units and four 12-inch units), and all of them are set with a support range of 50–150 Hz. I don’t really use the very low end much.

 

[DutchJay]

In my environment, I don’t experience such a clear change when adding more subwoofers. As the number of subs increases, the sound tends to become lighter, more spacious, and more natural, but with fewer subs there is more of a traditional “audio punch,” so either approach works for me.
In my JBL room, all of the base layer speakers are large (three 15-inch units and four 12-inch units), and all of them are set with a support range of 50–150 Hz. I don’t really use the very low end much.

Do we know for SURE that ART is not using all that cone area to support the primary bass, even when you set the support to 50 and up? Dirac always says they use all available bass capability with ART. So: Does support mean crossover or level when they start cancelling lingering bass? If it is the latter that could be a big factor why more subs dont give you much improvement?

 

[dlaloum]

With the A1H, I was able to drive the KEF Reference 1 Meta to fairly high levels without audible distortion, so I think there is good reason to expect solid performance from the X4800H’s amplifier as well. If it passes the dlaloum test, please make sure to proudly display the certificate of achievement.
One recent experiment I found particularly interesting was a single-subwoofer comparison between a JTR 18-inch and a JBL 15-inch. There is a significant performance gap between these two subwoofers, but when listening to music at around 100 dB SPL, I did not perceive a large difference in sound quality. The infrasonic option was turned off, and the support range was set to 20–150 Hz for the comparison.
When using DLBC, the sound lacked tightness and punch unless the JTR was used, but ART narrowed that performance gap significantly. The placement was different, so it was not a perfectly fair comparison, but even so, the result was quite surprising.
View attachment 499854
JTR SW
View attachment 499855
JBL SW
View attachment 499856

The KEF Ref1 Meta's are another torture test speaker with a dip going below 2ohm (EPDR) - the A1 is a muscular AVR that handles this well! - in the PAC range, the RZ70 would be expected to handle this well too (as its ancestors that I owned, the SR876 and DTR70.4 did for my Gallo's).

The X4800 is however, not in the same overall ball park as the A1 / A10 / RZ70... at a simple "rule of thumb" level the weight of the AVR provides an indication as a large proportion of the weight is in the power supply, and a robust high current power supply is a fundamental requirement for low impedance high current output (at least for traditional power supplies, not so much for switch mode power supplies).

Overall weight of the AVR's :

Denon A1h 32kg (should be OK with difficult speakers)
Denon A10h 23.6kg (should be OK with difficult speakers)
Denon X6800 15.6kg (?? - maybe ok with difficult speakers?)
Denon X4800 13.4kg (?? - maybe ok with difficult speakers?)
Denon X3800 12.5kg (?? - unlikely to be ok with difficult speakers)
Onkyo RZ70 22.5kg (should be OK with difficult speakers)
Integra DRX 3.4 11.0kg (known not to be OK with difficult speakers)

There is a definite step change variation in weight between the flagships and the rest of their respective ranges... a circa 5kg+ difference not at all unsubstantial!

Having said that - one key aspect is the stability of the power amp design into the difficult reactive loads - even with a more limited power supply, if it is stable, it will handle the speakers OK, albeit with power limitations caused by the limit of the power supplies current output (which into 2ohm or less, can be quite limiting!!) - the only way to tell this, is either on a lab test bench (and sadly few reviews test distortion/stability into low impedance reactive loads!) -or by ear with difficult speakers.

Hopefull I will be able to provide feedback (possibly that famous "certificate" ) soon!

I really would like to simplify my setup and remove the power amps - but not at the cost of sound quality...

 

[kawauso]

Do we know for SURE that ART is not using all that cone area to support the primary bass, even when you set the support to 50 and up? Dirac always says they use all available bass capability with ART. So: Does support mean crossover or level when they start cancelling lingering bass? If it is the latter that could be a big factor why more subs dont give you much improvement?

The support range for the full-range speakers is set to 50–150 Hz, while the subwoofers operate over 20–150 Hz. Below the configured lower limit, there is very little direct output or support-related output.