Active Room Treatment (ART) by Dirac

[KenMasters]

I'm pretty sure everyone using ART has noticed this or perhaps not everyone has noticed it because not everyone measures their system post ART calibration. I always measure post calibration as RCS don't always get it right.

In my experience ART does an excellent job with phase and time alignment. This particular example is for 12 point measurement calibration for a 6ft 3 seater sofa. ART's speaker time alignment is impeccable, all speakers are within 0.53 inches of each other and phase alignment is flat and my system sounds great.

I'm an enthusiast and not expert but I think ART makes my subs start super early, 15ms early(Circled in red) to accomplish its goal of optimizing based on a flat phase. It is my understanding that for subs the start of the impulse response which in my case is 15ms should be aligned at t=0 as this is when energy from the subs starts to arrive at our microphone(Ears). But since Dirac locks us out of delay adjustments post calibrations there is nothing we can do about this early start. I wish Dirac would reconsider and allow for post EQ time alignment at least so that we can correct for this.

I believe it's the sub's impulse peak that should be aligned at t=0. Dirac currently has a bug resulting in a delay of 5ms for the LFE channel relative to the other channels, so your graph checks out in that regard. Nothing to be done but wait for the fix.

 

[Oddball]

I believe it's the sub's impulse peak that should be aligned at t=0. Dirac currently has a bug resulting in a delay of 5ms for the LFE channel relative to the other channels, so your graph checks out in that regard. Nothing to be done but wait for the fix.

I don't think delay is limited to 5 or 15ms. It is there and I pointed out to that earlier in the thread. It will just depend on the room and system. When I call for better handover between the ART and Live - this is one of the points. Not sure how it really works with Live - but IMO they should repurpose more filters for 150hz to 8khz range. What happens beyond could be managed with lower filter resolution.

 

[Oddball]

That's just Nolan being Nolan. He deliberately mixes loud music and SFX over top of dialog, describing it as "adventurous and creative." He also refuses to use ADR, so if the dialog recorded by the boom mic over a loud IMAX camera is muddy on the shoot that day, that's what the audience gets. And before I forget, he refuses to adopt Atmos mixes, the best you will get from his films is DTS Master HD 7.1.

Great films, but a very opinionated director.

I love his films and some do have great soundtracks. He just went a bit too far with this one.

 

[KenMasters]

I don't think delay is limited to 5 or 15ms. It is there and I pointed out to that earlier in the thread. It will just depend on the room and system. When I call for better handover between the ART and Live - this is one of the points. Not sure how it really works with Live - but IMO they should repurpose more filters for 150hz to 8khz range. What happens beyond could be managed with lower filter resolution.

Known issues

• 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.

Software Changelog

helpdesk.dirac.com

You can see this evidenced with the peak being situated at +5ms on @mmares056's posted graph.

 

[Oddball]

Known issues

• 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.

Software Changelog

helpdesk.dirac.com

You can see this evidenced with the peak being situated at +5ms on @mmares056's posted graph.

That is a know issue as you say, but the real impact is spread across the Dirac Live board and is definitively not limited to 5ms. I have posted my grahs previously suggesting the bigger delay, as I recall Timo has as well. 5ms is pretty much inconsequential, but it is not really limited to 5ms.

 

[KenMasters]

That is a know issue as you say, but the real impact is spread across the Dirac Live board and is definitively not limited to 5ms. I have posted my grahs previously suggesting the bigger delay, as I recall Timo has as well. 5ms is pretty much inconsequential, but it is not really limited to 5ms.

Well, it's audible in my system when running a sweep (output slumps at 120Hz), so it would be nice to have it attended to. Otherwise I have no complaints.

 

[janbth]

I believe it's the sub's impulse peak that should be aligned at t=0. Dirac currently has a bug resulting in a delay of 5ms for the LFE channel relative to the other channels, so your graph checks out in that regard. Nothing to be done but wait for the fix.

A common belief, but it is wrong. It’s probably unintuitive, but if you split a perfect - i.e. broadband - impulse response into a sum of a low passed and high passed part, then the low passed (the «sub») will start at t=0 when the signal starts to rise. It will not have the top of the peak at t=0.

 

[Oddball]

Well, it's audible in my system when running a sweep (output slumps at 120Hz), so it would be nice to have it attended to. Otherwise I have no complaints.

Not sure I understand - output slumps at 120hz in what respect?

 

[Satanica]

Known issues

• 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.

This also effects DLBC and I've reverted to version 13.12.2 that doesn't have the issue.

 

[KenMasters]

Not sure I understand - output slumps at 120hz in what respect?

On my system, with ART engaged, if I listen to a sweep such as this, response sounds level until it reaches 120Hz, at which point it drops out, then picks up again.

 

[KenMasters]

A common belief, but it is wrong. It’s probably unintuitive, but if you split a perfect - i.e. broadband - impulse response into a sum of a low passed and high passed part, then the low passed (the «sub») will start at t=0 when the signal starts to rise. It will not have the top of the peak at t=0.

If you could help me to understand. Here is an example using my three subs without EQ. First I have the two Arendals:


Then I overlay my Buchardt:


Are you saying this would not be correctly aligned, that the Buchardt should be placed later in time?

 

[Keith_W]

@KenMasters those three impulses are aligned at their peaks. But that kind of alignment is irrelevant, what matters is whether the phase is aligned and the outcome is a smooth frequency response.

 

[KenMasters]

@KenMasters those three impulses are aligned at their peaks. But that kind of alignment is irrelevant, what matters is whether the phase is aligned and the outcome is a smooth frequency response.

Sure, I agree, but currently I'm speaking purely within the context of this earlier post.

In that posted example, assuming phase is not an issue, would you say his subs should be placed forward or back in time, or be left alone?

 

[Keith_W]

Sure, I agree, but currently I'm speaking purely within the context of this earlier post.

In that posted example, assuming phase is not an issue, would you say his subs should be placed forward or back in time, or be left alone?

ASSUMING phase is not an issue (which it almost always is), that kind of impulse alignment is correct if he is using linear-phase DSP. But if he is using minimum-phase DSP, the impulses should be aligned at impulse start, not impulse peak.

 

[KenMasters]

ASSUMING phase is not an issue (which it almost always is), that kind of impulse alignment is correct if he is using linear-phase DSP. But if he is using minimum-phase DSP, the impulses should be aligned at impulse start, not impulse peak.

Thank you. We're discussing ART, so the assumption is based on the fact that Dirac will have done what it can to correct phase though latency adjustment - so not minimum-phase DSP.

 

[markus]

Well, it's audible in my system when running a sweep (output slumps at 120Hz), so it would be nice to have it attended to. Otherwise I have no complaints.

There’s hope that the 120Hz LFE low-pass filter will become a user-selectable option during calibration in a future release. Ideally, it would be available as a playback setting, allowing users to better match the filter to the specific program material. Some content benefits from a low-pass filter, while other material may sound better without it. In any case, Dirac could address the erroneous delay by applying an all-pass filter to the other channels that matches the LFE filter.

 

[Keith_W]

Thank you. We're discussing ART, so the assumption is based on the fact that Dirac will have done what it can to correct phase though latency adjustment - so not minimum-phase DSP.

Dirac is mixed-phase DSP, but mostly minimum-phase.

 

[KenMasters]

Dirac is mixed-phase DSP, but mostly minimum-phase.

Yes, mixed phase, but it's correcting for the non-minimum phase in-room response. It aligns impulse responses by shifting frequencies in time, which in my room adds 37ms of latency to the audio signal.

Looking at @mmares056's graph and considering Dirac themselves point out that the LFE channel adds an unintentional 5ms of latency, it makes sense that ART is seeking to align impulse peaks.

 

[janbth]

If you could help me to understand. Here is an example using my three subs without EQ. First I have the two Arendals:

View attachment 529730
Then I overlay my Buchardt:
View attachment 529731

Are you saying this would not be correctly aligned, that the Buchardt should be placed later in time?

No, I'm not saying excactly that. If the goal is to time-align different subs with different shapes for their impulse responses, then aligning the peaks is not a reliable method. It may or may not work. Instead focus on the frequency response, and ensure that it is smooth.

If you are using REW, it also makes sense to check the graph for excess group delay. Here is a quote from the REW help pages [https://www.roomeqwizard.com/help/help_en-GB/html/minimumphase.html#top]:

"As an aside, the excess group delay plot also clearly shows there is a time offset between the subwoofer and the main speaker, the sub being about 25ms delayed, which is not so obvious from the overall group delay plot. Excess group delay is a useful plot for time aligning speakers."

 

[Paffi]

ASSUMING phase is not an issue (which it almost always is), that kind of impulse alignment is correct if he is using linear-phase DSP. But if he is using minimum-phase DSP, the impulses should be aligned at impulse start, not impulse peak.

Interesting - when I'm reminding correctly, OCA teached it with the peaks.