Active Room Treatment (ART) by Dirac

[JRowland]

Not pre-EQued in SVS LFE
View attachment 534641

Pre-EQued in SVS LFE
View attachment 534640

Same same, pre eq looks barely better. I’d do a preset a vs b good enough comparison and pick the one that sounds better to you.

 

[smygolf]

Same same, pre eq looks barely better. I’d do a preset a vs b good enough comparison and pick the one that sounds better to you.

Decay and waterfall with the combined result- do you have that to ?


If one are not going nuts with pre-correction ART should be able to handle it.

 

[RenPa]

Decay and waterfall with the combined result- do you have that to ?


If one are not going nuts with pre-correction ART should be able to handle it.

Front left subs not pr-equed in SVS

Same channel, same Art settings, same TC 2 subs pre-equed in SVS

Front left + subs Art Waterfall same channel NOT pre-equed in SVS . (Note: usually I prefer to measure L+R+subs for waterfall but I didn't do it this time)

FL+subs pre-equed +Art: (ringing at 60Hz is the refrigerator don't worry!)

My listening tests continue. All placebo effect???????? Can't say for now! Maybe...
There is a noticeable difference between the two approaches. In film soundtracks, the musical instruments are more prominent and detailed, the sound stage of the music goes further away the front wall (deeper) the surround sound is slightly more pronounced and expansive, and the bass is stronger. I’m talking about nuances that are fairly easy to pick up on, but it’s not a drastic change. I haven’t listened to any music in stereo yet. I’m still surprised to hear such subtle differences when analysing the various criteria in REW.

I obviously know the response of my subwoofers inside out. Before Dirac, I had integrated my subwoofers with a miniDsp 2x4hd and then manually mixed them at the crossover with the satellites in my AV receiver using 10 filters per channel. I got the best results if I reversed the polarity of one subwoofer. We know that neither DLBC nor Art do this. But my impression is that by first equalising the subwoofers in SVS, it’s as if Art were reversing something like the polarity of one of the subwoofers in the filters between 20Hz and 60Hz. Perhaps I’m just letting my imagination and my OCD get the better of me.

 

[smygolf]

Front left subs not pr-equed in SVS
View attachment 534720


Same channel, same Art settings, same TC 2 subs pre-equed in SVS
View attachment 534721

Front left + subs Art Waterfall same channel NOT pre-equed in SVS . (Note: usually I prefer to measure L+R+subs for waterfall but I didn't do it this time)
View attachment 534723


FL+subs pre-equed +Art: (ringing at 60Hz is the refrigerator don't worry!)
View attachment 534724

My listening tests continue. All placebo effect???????? Can't say for now! Maybe...
There is a noticeable difference between the two approaches. In film soundtracks, the musical instruments are more prominent and detailed, the sound stage of the music goes further away the front wall (deeper) the surround sound is slightly more pronounced and expansive, and the bass is stronger. I’m talking about nuances that are fairly easy to pick up on, but it’s not a drastic change. I haven’t listened to any music in stereo yet. I’m still surprised to hear such subtle differences when analysing the various criteria in REW.

I obviously know the response of my subwoofers inside out. Before Dirac, I had integrated my subwoofers with a miniDsp 2x4hd and then manually mixed them at the crossover with the satellites in my AV receiver using 10 filters per channel. I got the best results if I reversed the polarity of one subwoofer. We know that neither DLBC nor Art do this. But my impression is that by first equalising the subwoofers in SVS, it’s as if Art were reversing something like the polarity of one of the subwoofers in the filters between 20Hz and 60Hz. Perhaps I’m just letting my imagination and my OCD get the better of me.

Thank You

 

[Oddball]

This looks pretty good either way. I would not obsess over it - just pick the one that sounds better to you and don't look for a reason other than pure subjective preference.

BTW did you use "normal" SVS EQ or the new Auto EQ? Not familiar with the Auto EQ - just came out.

 

[RenPa]

BTW did you use "normal" SVS EQ or the new Auto EQ?

Don't no anything about Automatic for SVS peq.
In the app I can set manually 3 filters for each sub, selecting Q, gain and FR for each filter. I choose to always cut gain. When I will have time, I may try to reduce filter to one or 2 filters only with a larger Q (< 1).

Art is doing amazing things anyway.

 

[gepgep]

With an ART calibration using Harman 10db curve (which I used extensively with DLBC with no issues) it struggles to hit the curve below 50hz
When listening to movies there is a sound of something being overdriven on bass peaks in the soundtrack

I'm running 4 SVS PB3000 in corners in a single ART group supported by Centre and Mains at -18, 50Hzx

Digging into it with REW the overall picture shows the same effect as a classic crossover, in this case at 48Hz
Note the support is flat from 45-130Hz despite FSL at 50Hz

The individual subs however show wide variation in filters applied which makes me think I'm losing significant headroom

 

[Oddball]

Don't no anything about Automatic for SVS peq.
In the app I can set manually 3 filters for each sub, selecting Q, gain and FR for each filter. I choose to always cut gain. When I will have time, I may try to reduce filter to one or 2 filters only with a larger Q (< 1).

Art is doing amazing things anyway.

Right - just did some reading on SVS Auto EQ and works only with new R/Evolution series. So for the older ones, only manual PQ.

 

[Oddball]

With an ART calibration using Harman 10db curve (which I used extensively with DLBC with no issues) it struggles to hit the curve below 50hz
When listening to movies there is a sound of something being overdriven on bass peaks in the soundtrack

I'm running 4 SVS PB3000 in corners in a single ART group supported by Centre and Mains at -18, 50Hzx

Digging into it with REW the overall picture shows the same effect as a classic crossover, in this case at 48Hz
Note the support is flat from 45-130Hz despite FSL at 50Hz

The individual subs however show wide variation in filters applied which makes me think I'm losing significant headroom


View attachment 534753View attachment 534754View attachment 534755View attachment 534756View attachment 534757

There is absolutely no reason why ART would, in general be inferior to DLBC. It is supposed to be same or better in FQ response and ways better in decay control.

While you have subs that are pretty good, +10dB curve is a challenge (with +12dB being the max). What are your sub levels at AVR level and gain on the subs? I would doubt that they tap out even in large room but worth to ask.

It is normal that ART uses different subs with different filters, so that is not indication of losing headroom. Figuring out the ART filters is a bit complex with REW, but you can pull the ART screen that will show you what is the response for each channel and will be easier to comprehend as contribution from each channel will be overlaid.

Not sure why ART is running your LCR support this loud. I have LCR to F3 of 30hz, and also set to 50-120hz, but they start kicking in only at around 80-100hz.

 

[JRowland]

I'm running 4 SVS PB3000 in corners in a single ART group supported by Centre and Mains at -18, 50Hzx

What does each subs individual response look like? I’d prefer to see the Dirac measurement screen shots of their measured response but whatever image works for you.

 

[azzy_mazzy]

With an ART calibration using Harman 10db curve (which I used extensively with DLBC with no issues) it struggles to hit the curve below 50hz
When listening to movies there is a sound of something being overdriven on bass peaks in the soundtrack

I'm running 4 SVS PB3000 in corners in a single ART group supported by Centre and Mains at -18, 50Hzx

Digging into it with REW the overall picture shows the same effect as a classic crossover, in this case at 48Hz
Note the support is flat from 45-130Hz despite FSL at 50Hz

The individual subs however show wide variation in filters applied which makes me think I'm losing significant headroom


View attachment 534753View attachment 534754View attachment 534755View attachment 534756View attachment 534757

check the filters being applied to the subwoofers if one is being used more than the others, it also could be the speakers reaching their limits. 4 subwoofers are enough to support each other without much contribution from other speakers below 80Hz.

 

[smygolf]

With an ART calibration using Harman 10db curve (which I used extensively with DLBC with no issues) it struggles to hit the curve below 50hz
When listening to movies there is a sound of something being overdriven on bass peaks in the soundtrack

I'm running 4 SVS PB3000 in corners in a single ART group supported by Centre and Mains at -18, 50Hzx

Digging into it with REW the overall picture shows the same effect as a classic crossover, in this case at 48Hz
Note the support is flat from 45-130Hz despite FSL at 50Hz

The individual subs however show wide variation in filters applied which makes me think I'm losing significant headroom


View attachment 534753View attachment 534754View attachment 534755View attachment 534756View attachment 534757

Some questions

1#
Is any wall length or ceiling height 2 or 4 meters ?


2#
Room measurements and where is MLP - pic from REW would be great.

3#
What is support range within the subs group, and is the same setting used to support LCR ?


4#
When the overloading occurs, is it when listening to music or movies ?
And what lvl are you playing ?


And for tips - change LFE sub to another might help.
Have one group for each sub might also help
Also, to move the subs might also help

And lastly, my setup can play Crazy loud with Dirac/DLBC but i have loosed a bit of the top end with ART.
To widen the measurement area, using 13 points helped a bit

 

[gepgep]

What does each subs individual response look like? I’d prefer to see the Dirac measurement screen shots of their measured response but whatever image works for you.earlier

See my graphs posted

 

[gepgep]

Some questions

1#
Is any wall length or ceiling height 2 or 4 meters ?


2#
Room measurements and where is MLP - pic from REW would be great.

3#
What is support range within the subs group, and is the same setting used to support LCR ?


4#
When the overloading occurs, is it when listening to music or movies ?
And what lvl are you playing ?


And for tips - change LFE sub to another might help.
Have one group for each sub might also help
Also, to move the subs might also help

And lastly, my setup can play Crazy loud with Dirac/DLBC but i have loosed a bit of the top end.
To widen the measurement area, using 13 points helped a bit

1 width 4m
2 8x4x2.8 MLP centre 3.7m from screen
3 Yes
4 I only listen to movies. High level transient bass sometimes uses the effect

 

[gepgep]

check the filters being applied to the subwoofers if one is being used more than the others, it also could be the speakers reaching their limits. 4 subwoofers are enough to support each other without much contribution from other speakers below 80Hz.

You can see the effect of filters in the with/without graphs.

 

[gepgep]

What does each subs individual response look like? I’d prefer to see the Dirac measurement screen shots of their measured response but whatever image works for you.

See graphs already posted

 

[KenMasters]

@gepgep, have you tried removing the speaker support from the LFE group and playing the same material again to see if it's improved with the subs only supporting themselves?

 

[gepgep]

There is absolutely no reason why ART would, in general be inferior to DLBC. It is supposed to be same or better in FQ response and ways better in decay control.

While you have subs that are pretty good, +10dB curve is a challenge (with +12dB being the max). What are your sub levels at AVR level and gain on the subs? I would doubt that they tap out even in large room but worth to ask.

It is normal that ART uses different subs with different filters, so that is not indication of losing headroom. Figuring out the ART filters is a bit complex with REW, but you can pull the ART screen that will show you what is the response for each channel and will be easier to comprehend as contribution from each channel will be overlaid.

Not sure why ART is running your LCR support this loud. I have LCR to F3 of 30hz, and also set to 50-120hz, but they start kicking in only at around 80-100hz.

Thanks for your input

In this specific instance ART does not hit the curve as fully as DLBC. I'm not implying its inferior in general

Subs gain set at -15db I'll check the extra gain applied in the AVR. The net result of gain+filters shows in the graphs

My understanding is headroom, (for a group of subs), is the max gain+filters that could be applied to any single sub without causing it to audibly distort.
Hence adding gain+filters of 14db on one sub and 6db on another (see graphs) reduces headroom by 8db unless I'm misunderstanding the process.

Yes he support levels actually applied seem much higher than one would expect.

I ran some tests on a different calibration measuring the different support actually applied at support settings from -1 to -18db and FSL from 50-90Hz and there was only a very small variation between any of the settings. In the extreme the gain applied at 65Hz was identical in every case Even with DSL at 90Hz

 

[smygolf]

1 width 4m
2 8x4x2.8 MLP centre 3.7m from screen
3 Yes
4 I only listen to movies. High level transient bass sometimes uses the effect

Yes on nr 4 demands Proper subs and that i would say that you have.
Looking at the rooms measurement you should be fine and graphs are pretty close to the theoretical.
Especially around 60hz

Around 42 - 47 you should av a mix of modes and nulls and that is also a trouble region on your graphs.
And related to gain, my calibrated result have never got beyond the lvl of the uncalibrated - turn up your sub

 

[azzy_mazzy]

Do denon AVRs allow tone controls to function with dirac?