Active Room Treatment (ART) by Dirac

[TimoJ]

HTP-1 is no longer made only option is crazy expensive storm audio atm

HTP-1 is still made and available. Order yours today:

Monolith by Monoprice HTP-1 16-Channel Home Theater Processor with Dolby Atmos, DTS:X, Auro-3D, Roon Ready, and Dirac Live Bass Control Compatibility - Monoprice.com

The Monolith™ HTP‑1 is an "AVR" or "Pre‑Processor" designed for use in home theaters. It supports Dolby Atmos®, DTS:X®, Auro‑3D®, and Dirac Live

www.monoprice.com

 

[Oddball]

That’s perfectly possible of course . I guess my point is that impulse responses are unintuitive and hard. There’s more to sub integration than simply aligning peaks. The algorithm that Dirac ART uses to set relative time delays is not known, but it it probably involves a number of other ingredients as well, like phase, frequency response, etc.

That is also my understanding. There is definitely connection and dependency between all the "ingredients". One of my biggest challenges with Audy (and its best interface MultiEQ-X) was how to compromise between all of them. ART is of course much smarter and powerful algo, but don't think it can make the graphs perfect in all aspects. It also needs to compromise.

I noticed that subs are firing a bit early way back but never really got to fixing it. I think Timo also confirmed that similar thing happens on HTP-1, although his delay was just in the Dirac Live EQ'd area (he uses limited EQ), and my delay was across the entire range (as I use full range EQ). I also spoke to local Storm dealer and they confirmed this happens on Storm with ART as well but that they don't fix it as it would impact other areas of performance.

My thinking to "fix" the problem that subs are firing early was to delay them post ART calibration via their respective DSPs, which would work nicely with Arendals as they have delay on ms scale in the app, but with SVS would need to adjust the phase as no delay setting. Then I gave up on fixing it all together as just did not want to go through complexity of figuring out if anything else got broken.

ART and Dirack pack overall is for sure not perfect, but still the best I ever had. And I actually like the fact that it makes compromises for you.

 

[KenMasters]

I noticed that subs are firing a bit early way back but never really got to fixing it. I think Timo also confirmed that similar thing happens on HTP-1, although his delay was just in the Dirac Live EQ'd area (he uses limited EQ), and my delay was across the entire range (as I use full range EQ). I also spoke to local Storm dealer and they confirmed this happens on Storm with ART as well but that they don't fix it as it would impact other areas of performance.

My thinking to "fix" the problem that subs are firing early was to delay them post ART calibration via their respective DSPs, which would work nicely with Arendals as they have delay on ms scale in the app, but with SVS would need to adjust the phase as no delay setting. Then I gave up on fixing it all together as just did not want to go through complexity of figuring out if anything else got broken.

How do you square the notion that the subs fire early with Dirac's note - The 120 Hz low-pass filter introduced in version 3.13.2 applied to the LFE channel introduces a delay of about 5 milliseconds compared to the other channels.

If I look at my graph and others posted here, it suggests to me alignment should occur at the impulse peak, not the dashed line:

 

[welwynnick]

How do you square the notion that the subs fire early with Dirac's note - The 120 Hz low-pass filter introduced in version 3.13.2 applied to the LFE channel introduces a delay of about 5 milliseconds compared to the other channels.

Maybe they mean the delay is added to the other channels, relative to the LFE?
That would be consistent with the subs firing early.

 

[KenMasters]

Maybe they mean the delay is added to the other channels, relative to the LFE?
That would be consistent with the subs firing early.

Dirac clearly states the delay is to the LFE channel and the impulse peak positioning is consistent with their 5ms approximation.

 

[welwynnick]

Dirac clearly states the delay is to the LFE channel and the impulse peak positioning is consistent with their 5ms approximation.

To the best of my knowledge, it doesn't work like that. Group delay always increases at low frequencies, and time domain correction seeks to reduce or minimise that. That means adding delay to the higher frequencies. I can't think of any exceptions or caveats to that.

 

[KenMasters]

To the best of my knowledge, it doesn't work like that. Group delay always increases at low frequencies, and time domain correction seeks to reduce or minimise that. That means adding delay to the higher frequencies. I can't think of any exceptions or caveats to that.

Yes, and that is what occurs, which is why, in my case, my speakers are delayed by about 37ms.

The issue, as Dirac have stated, is that by introducing a low-pass filter to the LFE channel an additional delay of 5ms is incurred (on that channel only).

 

[welwynnick]

So the LFE is delayed 5ms and the other channels are delayed even more?

 

[KenMasters]

So the LFE is delayed 5ms and the other channels are delayed even more?

No, nothing happens to the other channels, and initially delay is set correctly for the LFE channel as well.

Then a low-pass filter is applied to the LFE channel, this incurs an erroneous 5ms penalty.

 

[janbth]

That is also my understanding. There is definitely connection and dependency between all the "ingredients". One of my biggest challenges with Audy (and its best interface MultiEQ-X) was how to compromise between all of them. ART is of course much smarter and powerful algo, but don't think it can make the graphs perfect in all aspects. It also needs to compromise.

I noticed that subs are firing a bit early way back but never really got to fixing it. I think Timo also confirmed that similar thing happens on HTP-1, although his delay was just in the Dirac Live EQ'd area (he uses limited EQ), and my delay was across the entire range (as I use full range EQ). I also spoke to local Storm dealer and they confirmed this happens on Storm with ART as well but that they don't fix it as it would impact other areas of performance.

My thinking to "fix" the problem that subs are firing early was to delay them post ART calibration via their respective DSPs, which would work nicely with Arendals as they have delay on ms scale in the app, but with SVS would need to adjust the phase as no delay setting. Then I gave up on fixing it all together as just did not want to go through complexity of figuring out if anything else got broken.

ART and Dirack pack overall is for sure not perfect, but still the best I ever had. And I actually like the fact that it makes compromises for you.

I have also observed from REW measurements (channels L+R) that the subs are firing a bit early, maybe 5 ms or so. But is this really a problem? It sounds great!

 

[mmares056]

I understand your point about not having access to the distance settings to correct things. The way my mind works it is moot so what can be done then with what is available? It’s subjective but things are better now that my impulses start together. It was good before, it’s better now. Totally subjective but it makes sense to me that the better “impact” I get from percussion now is a result of better timing. So if/then in my mind what did I do? New measurement so all of the variables involved with that, new subwoofer in a similar location with similar internal delay based off uncorrected impulse responses, I moved my tube trap over to cover the corner of my recessed door frame more, and I made a wireless sub wired all at the same time. I liked Ken’s dark mode so here again is my current set up

Just subs, orange trace is rear and closest physically, blue is sub 1 LFE.

View attachment 529983

So Dirac chooses a little later than the actual start again based on your circle explanation from earlier correct? I notice it hits directly in the trough this time instead of to the left a bit just writing “out loud”.

Total timing

View attachment 529984

Zoomed

View attachment 529985

I went back to some pics from when I only had 2 subs and my initial ART set up. I didn’t know about overlays but I see my subs were 2-3ms early then with no wireless involved.

View attachment 529987

View attachment 529988

The other thing I’m noticing and don’t understand is the numbers on the scale. Initially my first Dirac was low numbers and now they’re higher, why?

@KenMasters I did a some more digging and noticed that my 9 point measurement, MLP focus calibration looks different than my previously posted calibration which is a 12 point measurement covering a 6ft 3 seater sofa. The 9 point measurements MLP focus cal were taken 1ft from each subsequent measurement around the MLP. Nothing changed physically and one thing to note here is that my rear right corner loaded sub is the best performing sub. Second best performing sub is the right front, also corner loaded sub.

It's worth mentioning that even though my 4 subs are all the same, same specs and driver size not all four measure the same because of room modes and bass bleeding out to different areas of the house. Sub 4 RR corner is the best performing with flattest response and 20Hz to 150Hz detected range. The detected FR range for the FR corner loaded sub is also 20Hz-150Hz with the 9 point MLP Focus. The other two subs are the least performing with detected FR range of 22Hz-104Hz and 20.1Hz-84.6Hz respectively.

Please note that for my 12 point 3 seat sofa measurement cal only the RR sub has a detected FR range of 20Hz-150Hz and the other three show detected high frequency response 116Hz and below. So what I can conclude from all this is that Sub position is key and the stronger the position where all subs can extend the high frequency range to 150Hz or higher the better the time alignment ART will accomplish. However, REW verified impulse response does not change with either the 9 point MLP focus or the 12 point 3 seat sofa measurement cal.

9 point MLP focus correct/predicted IR. The dotted lines are almost on top of each other.

9 point MLP focus correct/predicted IR. Subs only

9 point MLP focus REW IR verified

I learn something new about ART every day. ART is intriguing indeed but whatever is doing it works because even in my imperfect room with all the limitations my system sounds very good.

 

[KenMasters]

The dotted lines are almost on top of each other.

And still the impulse peak exhibits the same 5ms delay.

I seem to be the only one that holds the opinion it's the peak that needs paying attention to, not the dashed line. Lets see how things look when Dirac finally addresses the issue.

 

[Oddball]

How do you square the notion that the subs fire early with Dirac's note - The 120 Hz low-pass filter introduced in version 3.13.2 applied to the LFE channel introduces a delay of about 5 milliseconds compared to the other channels.

If I look at my graph and others posted here, it suggests to me alignment should occur at the impulse peak, not the dashed line:

View attachment 530110

I honestly don't know as ART is a bit complex and we don't exactly know all the plumbing. I am still on Oct 2025 Dirac version and older D&M firmware. Subs firing early is definitively present in that setup. I am a bit slow on upgrades as unfortunately no time to verify upgraded setup which would drive me crazy.

Also, I am not a big fan of just figuring out the problem that I can't fix. Ignorance is not exactly a bliss, but at least saves some headache. I hope they continue to develop both Dirac Live and ART and make them better.

 

[JRowland]

@KenMasters I did a some more digging and noticed that my 9 point measurement, MLP focus calibration looks different than my previously posted calibration which is a 12 point measurement covering a 6ft 3 seater sofa. The 9 point measurements MLP focus cal were taken 1ft from each subsequent measurement around the MLP. Nothing changed physically and one thing to note here is that my rear right corner loaded sub is the best performing sub. Second best performing sub is the right front, also corner loaded sub.

It's worth mentioning that even though my 4 subs are all the same, same specs and driver size not all four measure the same because of room modes and bass bleeding out to different areas of the house. Sub 4 RR corner is the best performing with flattest response and 20Hz to 150Hz detected range. The detected FR range for the FR corner loaded sub is also 20Hz-150Hz with the 9 point MLP Focus. The other two subs are the least performing with detected FR range of 22Hz-104Hz and 20.1Hz-84.6Hz respectively.

Please note that for my 12 point 3 seat sofa measurement cal only the RR sub has a detected FR range of 20Hz-150Hz and the other three show detected high frequency response 116Hz and below. So what I can conclude from all this is that Sub position is key and the stronger the position where all subs can extend the high frequency range to 150Hz or higher the better the time alignment ART will accomplish. However, REW verified impulse response does not change with either the 9 point MLP focus or the 12 point 3 seat sofa measurement cal.

9 point MLP focus correct/predicted IR. The dotted lines are almost on top of each other.
View attachment 530157

9 point MLP focus correct/predicted IR. Subs only
View attachment 530158

9 point MLP focus REW IR verified
View attachment 530159

I learn something new about ART every day. ART is intriguing indeed but whatever is doing it works because even in my imperfect room with all the limitations my system sounds very good.

Nice! I notice on the uncorrected sub impulse display it hits 2 of your subs in a trough (pink and orange traces) and 2 on the peaks (green and yellow) where all of mine are hit on the peaks. Any idea what that is about?

Yours

Mine

Also why are my peaks so much more defined and no first dips instead of peaks? Is that room treatment on my end helping, I notice more "control" on my chart?

 

[Keith_W]

I have also observed from REW measurements (channels L+R) that the subs are firing a bit early, maybe 5 ms or so. But is this really a problem? It sounds great!

Short answer: sub firing 5ms early does not matter provided that it was deliberate.

For those who are curious, please read this article on auditory masking published by Siemens. You want section 2, on "temporal masking". In short, a smaller stimulus can be masked by the arrival of a larger stimulus (in this case, the main impulse) provided that the arrival time does not exceed a certain threshold. The article states 20ms, but we know that this also depends on other factors, such as frequency (less sensitive at low frequencies) and SPL relative to the main stimulus.

 

[mmares056]

Nice! I notice on the uncorrected sub impulse display it hits 2 of your subs in a trough (pink and orange traces) and 2 on the peaks (green and yellow) where all of mine are hit on the peaks. Any idea what that is about?

Yours

View attachment 530174

Mine

View attachment 530175

Also why are my peaks so much more defined and no first dips instead of peaks? Is that room treatment on my end helping, I notice more "control" on my chart?

I'm not really sure about the peaks being so much more defined and show more control for your graphs but my guess is probably room, setup, sub placement, and MLP specific. I also have a decent amount of room treatments and the RT60-Decay of my room hovers around 243ms and 400ms in the 100Hz to 11,500Hz. I don't have optimal sub placement and my MLP is somewhat compromised due to the fact my couch sits 1ft from the back wall where there is a 9ftx6ft window with double thick curtains that do a very good job absorbing reflections. The MLP true distance from back wall/window is 24.5". I also have absorbing and acoustical panels on the front and side walls. Sub 1 is located in the front left corner, Sub 2 at the 3/4 of front wall, Sub 3 at the front right corner, and Sub 4, which by the way is the best performing sub sits at the rear right corner of the room about 6ft from the MLP. My room is 16 ft x 13 ft x 8 ft and the left side opens up to the foyer, kitchen and bedroom hallway.

So my take away from all this is that my room sucks specially for bass response and unfortunately sub placement options are limited. So I have to live with what I have available at the moment and I strongly believe current sub placement is the best ART has to do its magic and it truly does magic in my imperfect room. Another option I have to compensate for the bleeding of bass into other areas of the house is to buy more powerful subs to pressurize the room enough so that ART doesn't cut the high frequency below 150Hz for any of the subs. But I don't think is is the right thing to do right now since we may have to move soon.

This is the detected Subs FR range for the 9MP MLP Focus calibration

And this is the Waterfall Graph for L+R+SUBS

 

[JRowland]

I'm not really sure about the peaks being so much more defined and show more control for your graphs but my guess is probably room, setup, sub placement, and MLP specific. I also have a decent amount of room treatments and the RT60-Decay of my room hovers around 243ms and 400ms in the 100Hz to 11,500Hz. I don't have optimal sub placement and my MLP is somewhat compromised due to the fact my couch sits 1ft from the back wall where there is a 9ftx6ft window with double thick curtains that do a very good job absorbing reflections. I also have absorbing and acoustical panels on the front and side walls. Sub 1 is located in the front left corner, Sub 2 at the 3/4 of front wall, Sub 3 at the front right corner, and Sub 4, which by the way is the best performing sub sits at the rear right corner of the room about 6ft from the MLP. My room is 16 ft x 13 ft x 8 ft and the left side opens up to the foyer, kitchen and bedroom hallway.

So my take away from all this is that my room sucks specially for bass response and unfortunately sub placement options are limited. So I have to live with what I have available at the moment. Another option I have to compensate for the bleeding of bass into other areas of the house is to buy more powerful subs to pressurize the room enough so that ART doesn't cut the high frequency below 150Hz for any of the subs. But I don't think is is the right thing to do right now since we may have to move soon.

This is the detected Subs FR range for the 9MP MLP Focus calibration
View attachment 530191

And this is the Waterfall Graph for L+R+SUBS
View attachment 530192

My sub scene isn't all that different from yours. My rear is mid wall, front left is corner loaded, front right (biggest of the 3) is in a hallway and probably half of the output goes into the office, up the stairs, and into the bathroom which are all the opposite direction of my MLP.

Interesting to me about the range Dirac predicts and your findings. My first ART I had 2 subs and the left front was corner loaded but only a 10 incher. It got down to 19Hz no problem at all (good output +10db) but fell off at 100 ish Hz. My impulses were not aligned and off by 2ms. I moved the 10 incher to the rear and added a 12 incher to that front left corner but I used that wireless kit and my impulses were off by 4ms (yet all subs had free range from 20Hz to 150Hz now). Now currently all 3 subs do at least 20Hz to 150Hz and are hard wired and my impulses align. I only went looking into this because when they aligned I felt an immediate difference (for the better) in the feel (specifically - not tone or anything but physical feel) of percussion. I figured it was just in my head but I found the impulses aligned where they weren't before after I started looking for what changed. Might still just be in my head but I like it better.

The other thing I notice and am curious about on your corrected sub impulse graph the first hump is a little prouder than mine and your first and second trough go to -0.004 where mine only goes to -0.002. Is that showing less pre-ringing in my room?

 

[mmares056]

The other thing I notice and am curious about on your corrected sub impulse graph the first hump is a little prouder than mine and your first and second trough go to -0.004 where mine only goes to -0.002. Is that showing less pre-ringing in my room?

I'm not 100% sure as I'm not quite there yet knowledge wise on the subject, if a larger number like mine means more ringing or if it means that the deeper throughs are more aggressively contributing to the cancellation of specific frequencies for ART to accomplish a flat phase. I encourage people with more knowledge on this subject to chime in.

 

[mmares056]

Anyone here has a sub located at the rear of the room, either rear right or rear left that happens to be the best performing sub and is using it as Subwoofer 1/LFE? Reason I'm asking is because my rear right, corner loaded sub that sits 6ft from the MLP is the best performing sub and I'm curious to hear other people's experience with this setup. I plan on experimenting with my rear right sub connected to the Sub 1 out to see how the bass sounds.

 

[JRowland]

I'm not 100% sure as I'm not quite there yet knowledge wise on the subject, if a larger number like mine means more ringing or if it means that the deeper throughs are more aggressively contributing to the cancellation of specific frequencies for ART to accomplish a flat phase. I encourage people with more knowledge on this subject to chime in.

Sorry for asking you so many questions Obviously I'm lacking knowledge and curious, thank you for sharing the knowledge you do have as it has been helpful to me.