Dirac ART is now running on beta FW for Denon Xx800H AVRs!

[Oddball]

Cool, I am REALLY interested to test this with ART:

1 Sub and two decent towers, vs 2 subs and the same towers. Also because some people in this threath report that the bass with ART is anchored to the front stage, so all around you might change then? And the smoothness and impact has already improved SO MUCH in my set up.

In the end I'll prob test this myself as it can easily be done with the 60 days trial from Arendal... Let's see how long it will take me to scratch that itch.... Hahaha...

I'd go even more granular. Suspect that 1 vs 2 subs will be a bigger difference. Then would add towers to 1 and 2 sub setup and see where that leads. Not with my setup as traveling but it is an interesting thing to try.

ART can be anchored to any speaker in your setup. C seems to make most sense and yes it is front stage. But if you fancy, could link it to back surrounds as sell. Not sure how would that sound though.

 

[cbelizon]

Just an advice, the important graph to follow is the spread one, the lesses the spread you achieve, the better the sound. I had to even lower the support of my subwoofer to -6db on several of my speakers to ensure less spread with the audioadvice 5db curve that works like charm in HT, and for music the 3db one.

Also, for music, I only engage L/R supporting each other and my subwoofer, better imagining.

 

[Oddball]

Just an advice, the important graph to follow is the spread one, the lesses the spread you achieve, the better the sound. I had to even lower the support of my subwoofer to -6db on several of my speakers to ensure less spread with the audioadvice 5db curve that works like charm in HT, and for music the 3db one.

Also, for music, I only engage L/R supporting each other and my subwoofer, better imagining.

Can we measure that? I get the point, but not sure how it would look like in the graph. Away and traveling so can't use my system.

 

[cbelizon]

Can we measure that? I get the point, but not sure how it would look like in the graph. Away and traveling so can't use my system.

The corrected spread on dirac

 

[DutchJay]

The corrected spread on dirac

Can you post a screenshot of good spread? And maybe even of a worse one, like when you use -18 on those speakers?

And you say that adjusting the support will effect the spread? Only sub 150hz or also above?

Thanks!

 

[gorman]

And you say that adjusting the support will effect the spread? Only sub 150hz or also above?

ART settings influence only frequencies <150Hz. Above you have regular Dirac Live operating.

 

[DutchJay]

Well pretty much what everybody is reporting when they buy a second sub Going dual did a few different things for me, mostly what everybody tells you; smoother frequency response, more headroom (= more deep bass output and less distortion), less localisation (the bass feels more like it's coming from "all around") and the biggest thing for me: more sense of pressure at lower volumes (this increased even more with a third and a fourth sub).

One more question, did your feeling of "bass coming from all around" change with ART as some folks here mention it is more anchored to the front?

 

[AlArenal]

I just "discovered" the corrected spread recently as a tool in Dirac. Of course, there was not much to do with it in the before ART times...
But I'd also like to put it in perspective. I measured my MLP small with two additional points - so 11 points total. The corrected spread shows me the assumed by Dirac levels over the frequency band at my 11 measured points.
But right now I see no way of knowing which of these points attributes which part of the spread or how the spread looks at the heart of my measured 3d sapce, the MLP.

So yes, I see the corrected spread as a tool but not as the one thing that shows me what to expect. In real life, I could not care less about a narrower spread that is the result of tighter ranges at two additional measurement points I'll never have my ears at and it would not tell me if the sound is expected to be somewhat better where my ears are really at.

 

[fogs]

Greetings! Reaching out to check if anyone has had the same issue as me: I'm running a combined system used for both 2-channel music (2.3) and home theater (7.3.4). Main unit for both usage scenarios is Denon AVC-X3800H with Dirac ART, mains are Dutch&Dutch 8c w/ Bacch plugin. And here's the crux: For music, I would love to run the ART-filters with the Bacch filter on my 8c's active, i.e. a fully calibrated 2.3 system.

However, it's turned out to be anything but straight forward to run the Dirac measurements with that in mind. If having the Bacch filter active during the measurements, Dirac stops the process with some kind of error message (don't remember the wording exactly, but I tend to think that's a bit beside the point).

To try to circumvent this issue (i.e. simulate the Bacch plugin latency), I tried to pipe the mains through a MiniDSP SHD I have kept from before, setting a delay there of 10ms, i.e. equivalent to the latency introduced by Bacch (found that snippet of info from an online review of the speakers). My idea was to remove the delay set in the SHD post measurements. For some reason, this also gave hick-ups during the Dirac measurements and it didn't complete.

Running the Bacch plugin without its' latency taken into account during the Dirac measurements is as expected utterly chaotic, with the 10ms latency from the Bacch-plugin totally ruining the timing (it's NOT subtle! ).

The 8c's are put in low-latency mode, i.e. 3ms latency.

I've tried customer service at both Dutch&Dutch and Dirac to no avail (don't blame them though, as my problem lies in the interface between 2 independent products).

Anyone? @slaweks ? @Martijn Mensink ? @Flak ?

 

[Oddball]

Greetings! Reaching out to check if anyone has had the same issue as me: I'm running a combined system used for both 2-channel music (2.3) and home theater (7.3.4). Main unit for both usage scenarios is Denon AVC-X3800H with Dirac ART, mains are Dutch&Dutch 8c w/ Bacch plugin. And here's the crux: For music, I would love to run the ART-filters with the Bacch filter on my 8c's active, i.e. a fully calibrated 2.3 system.

However, it's turned out to be anything but straight forward to run the Dirac measurements with that in mind. If having the Bacch filter active during the measurements, Dirac stops the process with some kind of error message (don't remember the wording exactly, but I tend to think that's a bit beside the point).

To try to circumvent this issue (i.e. simulate the Bacch plugin latency), I tried to pipe the mains through a MiniDSP SHD I have kept from before, setting a delay there of 10ms, i.e. equivalent to the latency introduced by Bacch (found that snippet of info from an online review of the speakers). My idea was to remove the delay set in the SHD post measurements. For some reason, this also gave hick-ups during the Dirac measurements and it didn't complete.

Running the Bacch plugin without its' latency taken into account during the Dirac measurements is as expected utterly chaotic, with the 10ms latency from the Bacch-plugin totally ruining the timing (it's NOT subtle! ).

The 8c's are put in low-latency mode, i.e. 3ms latency.

I've tried customer service at both Dutch&Dutch and Dirac to no avail (don't blame them though, as my problem lies in the interface between 2 independent products).

Anyone? @slaweks ? @Martijn Mensink ? @Flak ?

Very interesting point. ART with actives. Can it be done?

 

[cbelizon]

I just "discovered" the corrected spread recently as a tool in Dirac. Of course, there was not much to do with it in the before ART times...
But I'd also like to put it in perspective. I measured my MLP small with two additional points - so 11 points total. The corrected spread shows me the assumed by Dirac levels over the frequency band at my 11 measured points.
But right now I see no way of knowing which of these points attributes which part of the spread or how the spread looks at the heart of my measured 3d sapce, the MLP.

So yes, I see the corrected spread as a tool but not as the one thing that shows me what to expect. In real life, I could not care less about a narrower spread that is the result of tighter ranges at two additional measurement points I'll never have my ears at and it would not tell me if the sound is expected to be somewhat better where my ears are really at.

Take into account that actually ART is basically made for reducing decay, so, even in your situation reducing the spread would bring benefits to clarity.

 

[fogs]

Take into account that actually ART is basically made for reducing decay, so, even in your situation reducing the spread would bring benefits to clarity.

Not quite sure I follow you - are you referring to "spread" as used/referred to in ART? Anyways, ART and Bacch work in different FR ranges so I imagine that together they would potentially yield more than twice the fun!

 

[cbelizon]

Decay and spread are not the same, but if you reduce the spread it is likely the decay will also be better.

 

[fogs]

Spread=to what degree the response varies across the different microphone positions, right? (ref. ART user manual). Not quite sure how that plays into what I'm trying to achieve here, but feel free to educate me. To clarify what I'm after: The 8c's have really good bass by themselves - but in my room, my preferences etc, they do still benefit from an even more solid bass foundation from well integrated subs. IMO. Bacch on top of that could possibly be the proverbial icing on the cake. Possibly.

 

[Oddball]

Spread=to what degree the response varies across the different microphone positions, right? (ref. ART user manual). Not quite sure how that plays into what I'm trying to achieve here, but feel free to educate me. To clarify what I'm after: The 8c's have really good bass by themselves - but in my room, my preferences etc, they do still benefit from an even more solid bass foundation from well integrated subs. IMO. Bacch on top of that could possibly be the proverbial icing on the cake. Possibly.

Possibly. Engaging 3 different room correction systems at the same time is pretty new concept though. To much processing could potentially turn out as too much processing?

 

[EWL5]

Possibly. Engaging 3 different room correction systems at the same time is pretty new concept though. To much processing could potentially turn out as too much processing?

Yeah, too many chefs in the kitchen!

 

[ban25]

Greetings! Reaching out to check if anyone has had the same issue as me: I'm running a combined system used for both 2-channel music (2.3) and home theater (7.3.4). Main unit for both usage scenarios is Denon AVC-X3800H with Dirac ART, mains are Dutch&Dutch 8c w/ Bacch plugin. And here's the crux: For music, I would love to run the ART-filters with the Bacch filter on my 8c's active, i.e. a fully calibrated 2.3 system.

However, it's turned out to be anything but straight forward to run the Dirac measurements with that in mind. If having the Bacch filter active during the measurements, Dirac stops the process with some kind of error message (don't remember the wording exactly, but I tend to think that's a bit beside the point).

To try to circumvent this issue (i.e. simulate the Bacch plugin latency), I tried to pipe the mains through a MiniDSP SHD I have kept from before, setting a delay there of 10ms, i.e. equivalent to the latency introduced by Bacch (found that snippet of info from an online review of the speakers). My idea was to remove the delay set in the SHD post measurements. For some reason, this also gave hick-ups during the Dirac measurements and it didn't complete.

Running the Bacch plugin without its' latency taken into account during the Dirac measurements is as expected utterly chaotic, with the 10ms latency from the Bacch-plugin totally ruining the timing (it's NOT subtle! ).

The 8c's are put in low-latency mode, i.e. 3ms latency.

I've tried customer service at both Dutch&Dutch and Dirac to no avail (don't blame them though, as my problem lies in the interface between 2 independent products).

Anyone? @slaweks ? @Martijn Mensink ? @Flak ?

I don't think it's a good idea to stack DSP in that manner, at least not during the calibration phase. Consider that during calibration, Dirac is going to try to correct for any of the "corrections" Bacch has applied, effectively un-doing them. If you really want to stack them, then disable Bacch during Dirac calibration and re-enable it after (during playback).

The other issue is that Dirac has a maximum amount of latency it can tolerate before it gives up. The default is actually 20 ms, but some manufacturers like D+M have relaxed that limit to 50 ms. That said, ART may be more restrictive in regards to latency.

 

[gorman]

Spent the better part of the day fiddling with settings and measuring every combination I have tried.

I had a problem with bass on surround speakers, tried everything, added more support, added more speakers to support... it got better but not to the level of the front speakers...

Well, as an afterthought I looked at the fronts and saw that I had opted to have them supported just by the subwoofers. Thought "why not do the same for the surrounds?". Lo and behold, far better results.

I went from this

to this

I have saved all measurements, hoping that knowledgeable people could comment on my results. @Keith_W are you still active here these days? Peak energy delay in Wavelet spectrogram looks great now, correct?

 

[Keith_W]

Well @gorman my aim isn't to read measurements for people, what I am trying to do is show which measurements are relevant so that you can read them yourself. If I did the former, I would be here all day. It's a muscle you have to practice, if you look at my posts on ASR from a couple of years ago I was hopeless. You only get better by looking at dozens of measurements and wondering what it all means.

Anyway, it is difficult to say what ART has done for you since "before" measurements were not included. So there is nothing to compare. But there are still a few interesting things in your MDAT.

All the little squiggly lines at the bottom are individual speakers, and in green we have the vector sum of all of them. What we want to see is that the squiggly lines do not deviate too much (indicating that all your speakers are tonally similar) - and they don't. Good.

We can also see some really surprising things. Like, your centre channel and surround speakers are flat down to 30Hz and individually have identical measurements below 150Hz, despite different placements in the room, and despite pesky room acoustics. Like, all the peaks and dips are identical. This is unbelievable and it's super fishy. It has to be an error in measurement, like your AVR grafting the bass onto multiple speakers and only measuring the upper frequencies individually. It's impossible for individual speakers to behave like that. Maybe go repeat the measurement and check that the AVR isn't also sending signal to the subwoofers or something like that.

The other thing to see is that each individual speaker is adjusted to a -10dB downwards treble tilt. When summed, the tilt becomes almost -20dB. If that is true (and I suspect that it isn't!) ... that's a huge amount of bass in your room. What does it sound like?

The other thing to see is the unbelievably flat spectrogram from 40Hz to 20kHz. The RT60 on average is <200ms from 300Hz up. It's obvious that this is a highly treated room, possibly overtreated. Without knowing the room dimensions I don't know if this is within target or not.

This is the step response of "SL". I am assuming this is the front left speaker. Here we see the same behaviour I observed in this post where I looked at @just1n 's MDAT of Dirac ART. I wondered if this behaviour was a one-off. But this more or less confirms it. The step response of all your speakers looks similar - the woofer plays about 15ms before the tweeter. I really don't like that and I wonder whether this would be audible or not.

 

[gorman]

Well @gorman my aim isn't to read measurements for people, what I am trying to do is show which measurements are relevant so that you can read them yourself. If I did the former, I would be here all day. It's a muscle you have to practice, if you look at my posts on ASR from a couple of years ago I was hopeless. You only get better by looking at dozens of measurements and wondering what it all means.

Anyway, it is difficult to say what ART has done for you since "before" measurements were not included.

I'm sorry. You are much more active here than I am, so of course I remembered your past help but you couldn't possibky remember me.

Past measurements were here (and later in that discussion):

New REW measurements, hope knowledgeable people will help me analyse them.

After what's been for me a long journey, I've now more or less settled on what sounds good to me. I'm fully aware that sound can always be better, as such I've taken measurements as best as I could, trying to follow the guidelines kindly provided here by @Keith_W . Measurements were taken at MLP...

www.audiosciencereview.com

I will reply later to everything else. But have extensively listened to this system and it sounds like a regular +6 - 4 target curve. A 20db difference from 20 to 20,000 would be unlistenable to me.

EDIT: some of the things you wrote make me ask: are you familiar with how Dirac ART works? Because from 20Hz to 150 Hz all speakers work to provide good bass. They emit sounds that cancel unwanted waves to bounce around the room (if I understand correctly).