A1 Evo AcoustiX vs Dirac Live vs DLBC vs DIRAC ART in a 2.2 front setup (Denon X3800h, KEF R3 meta, SVS 2000 pro)

[peng]

I only corrected up to 500 Hz.\
Sub placement elsewhere is difficult due to the other functions of the room

If you only corrected to 500 Hz, why would DLBC and ART change the response above 1 kHz so much?

 

[Mort]

If you only corrected to 500 Hz, why would DLBC and ART change the response above 1 kHz so much?

View attachment 512878

Measurement inconsistency is my guess.

 

[Geertidow]

If you only corrected to 500 Hz, why would DLBC and ART change the response above 1 kHz so much?

View attachment 512878

That's the weird part. I use a mic stand and switch presets so it's all the same position. That's why I don't get the phase and FR difference.

For different sets (sub back or front of course it's slightly differen).
Dirac Live is different measurement set. But DLBC and ART is the same for example.

 

[peng]

That's the weird part. I use a mic stand and switch presets so it's all the same position. That's why I don't get the phase and FR difference.

For different sets (sub back or front of course it's slightly differen).
Dirac Live is different measurement set. But DLBC and ART is the same for example.

It is worth investigating to find out what's happening because the significant FR differences in that mid/high range would be highly audible.

 

[Oddball]

Can you get used to it (for music purposes)?
I've read some similar stories on ASR. But perhaps because the subwoofer is placed behind me (on the open side of the U shaped room) that skews the percieved sound.
If anything, I'll buy an extra sub for that position and leave the 2 subs in front.
But I'm not entirely sure about the measurements, becuase of the phase and the high frequency responses (I usually ignore them).
This is ART up to 1000 Hz. With the sub behind graph the phase is normal, with the sub in front it's not. That shouldn't be.

So still a bit of searching to do with this program. But luckily it already sounds great.
I think in both graphs all the deviations from the freq response are at most 3 dB when looking at my scale.

Bonus question: should I unplug the ports of my R3 meta?

Sub in the front/back arrangement is optimal for ART, in some or perhaps most rooms, as cancellations work best in front-back setup. Goes back to double bass array that was around for ages, and is also essential for Trinnov waveforming. Just gives ART more to work with so one can expect better results in systems with somewhat limited resources.

My sub positioning is also 2 front - 2 back. 2 back are actually nearfield and I don't mind them or notice them as they work great with the fronts.

Your graphs are relatively good so I would focus more on how it sounds. Perhaps try a bit different support ranges and intensity and see where that leads you.

Probably should try to unplug the ports and see if that would fill in some of the dips in the lower end. ART will likely not run them too low, but even modest level of support could help a bit.

 

[Oddball]

It is worth investigating to find out what's happening because the significant FR differences in that mid/high range would be highly audible.

For sure worth looking into it and perhaps trying higher level of support between L and R. This causes issues for some people and they note that it muds the soundstage, but might work for some. It is a bit difficult to get pristine response with only 2 speakers. If multichannel, other speakers would likely smooth out the response a bit.

 

[Geertidow]

Yeah it sounds perfect in the highs so I'm not too worried about it.
I actually really enjoy the sound, the difference is stunning.
But I will unplug the ports.

i gotta read on the support ranges.
Excited to test some more with this.

Sub in the front/back arrangement is optimal for ART, in some or perhaps most rooms, as cancellations work best in front-back setup. Goes back to double bass array that was around for ages, and is also essential for Trinnov waveforming. Just gives ART more to work with so one can expect better results in systems with somewhat limited resources.

My sub positioning is also 2 front - 2 back. 2 back are actually nearfield and I don't mind them or notice them as they work great with the fronts.

Your graphs are relatively good so I would focus more on how it sounds. Perhaps try a bit different support ranges and intensity and see where that leads you.

Probably should try to unplug the ports and see if that would fill in some of the dips in the lower end. ART will likely not run them too low, but even modest level of support could help a bit.

It can't be a broken mic or something? It's an old UMIK1. But I guess it works or not.

First I'll play a bit more with the crossover, the support range et cetera.
Then Ill upload some more measurements.

 

[OCA]

That's the weird part. I use a mic stand and switch presets so it's all the same position. That's why I don't get the phase and FR difference.

For different sets (sub back or front of course it's slightly differen).
Dirac Live is different measurement set. But DLBC and ART is the same for example.

You are measuring L+R together, and the mic is not perfectly centered (it looks ~5 cm off). That small offset means each speaker arrives at slightly different times, so the HF response shows strong comb filtering and phase cancellations and is not meaningful.

Use an acoustic timing reference and measure each speaker individually. After that, you can time-align their impulse peaks in REW and use A+B to see the summed response that will better represent what you actually hear.

Also note that in the last couple of measurements the output is no longer using WASAPI drivers to the Denon, while the UMIK seems still on WASAPI. That mismatch can introduce additional clocking/synchronization issues and further corrupt the impulse and phase data.

 

[Geertidow]

You are measuring L+R together, and the mic is not perfectly centered (it looks ~5 cm off). That small offset means each speaker arrives at slightly different times, so the HF response shows strong comb filtering and phase cancellations and is not meaningful.

Use an acoustic timing reference and measure each speaker individually. After that, you can time-align their impulse peaks in REW and use A+B to see the summed response that will better represent what you actually hear.

Also note that in the last couple of measurements the output is no longer using WASAPI drivers to the Denon, while the UMIK seems still on WASAPI. That mismatch can introduce additional clocking/synchronization issues and further corrupt the impulse and phase data.

I have no idea how to change WASAPI drivers but I will try to find out out to fix that. Would acoustic timing ref fix this as well?

 

[OCA]

I have no idea how to change WASAPI drivers but I will try to find out out to fix that. Would acoustic timing ref fix this as well?

Java Exclusive drivers (EXCL) are wasapi drivers and have much lower latency than normal java drivers. Always prefer them when available. ASIO drivers are even better. You will see them in the input/output lists in preferences/sound card.

 

[OCA]

Always 0-24kHz for 48kHz sampling rate measurements, always timing reference, almost never L+R

 

[Geertidow]

Tomorrow I'll measure.
You sir, are an absolute wizard of audio measuring.

 

[Chromatischism]

What target curve are you using?

I've just done a Dirac Live calibration with the Magic Beans 2 pass method and I am going to be doing some post EQ measurements soon, I'll upload them here so you can compare.

This is what Dirac Live did to my speakers using a Flat Target curve.

View attachment 512716

Please adjust your graph limits so that the vertical axis ranges from 45 to 105 dB.

 

[Geertidow]

So I've changed my seating according to AI's advice on listening distances and room sim in REW (double checked AI).
I will remeasure as soon as I find the time and post some graphs, hopefully with the right drivers and with acoustic time reference.
I'll use the right vertical axis range and hope to be able to show a proper graph (more in line with the great sound I'm hearing).
For the mean time with the proper axis.

 

[Geertidow]

Still working on the summation (unsuccesfully).
Still something weird with the right speaker if you look at the phase, but can't really hear anything strange.
If I get the summation graphs I'll post some plots

 

[Chromatischism]

If SPL is even, don't worry about phase, IMO.

Tip: most people post from the "All SPL" tab.

Also, you can use the Camera icon to capture the image.

 

[thcdru2k]

Looking at the ART calibration by itself, the decay and timing behavior suggest it is tuned more for even bass across the room rather than the fastest decay. In the 25 to 45 Hz range the energy hangs around for several hundred milliseconds, which points to the room modes staying active instead of fading quickly. Above about 100 Hz the decay looks generally controlled, but there is clearly more low frequency energy stored over time. This is a measurable trait and likely reflects a choice to favor consistency across seats rather than the quickest bass decay. Personally, I would probably lean toward a fully active DSP speaker approach since it allows tighter control of both the speaker and the room, but that kind of setup is simply out of my budget right now.

 

[thcdru2k]

I might be nitpicking a little but also look at your SPL graphs right main and sub, left main and sub. It has downward tilt which is fine (strong bass extension) but it dips around crossover region. For right main and sub that presents as a slight valley between 60-100hz ; it's also there on left main and sub but very very minor/subtle some would just call it flat. I have a 2.2 setup myself so that is where I've paid close attention to my system. I prefer less of a downward tilt but that's just preference however; I do not like to see a valley at the crossover region. If I have two subs and two speakers playing shared frequencies I would like to see a flat even small hill to indicate that there is proper summation. I listen to content that has rapid bass transients so it's something I notice. I think Dirac ART might accept a small dip or compromise in 80-120hz if it helps avoid bigger peaks and nulls elsewhere is my guess. I think there is just no way to have everything you want unless you have a lot of money to throw at fully active DSP speakers. Dirac ART is balancing crossover behavior against spatial uniformity, rather than optimizing purely for the most perfect local blend between speakers and subwoofers.

 

[Geertidow]

I might be nitpicking a little but also look at your SPL graphs right main and sub, left main and sub. It has downward tilt which is fine (strong bass extension) but it dips around crossover region. For right main and sub that presents as a slight valley between 60-100hz ; it's also there on left main and sub but very very minor/subtle some would just call it flat. I have a 2.2 setup myself so that is where I've paid close attention to my system. I prefer less of a downward tilt but that's just preference however; I do not like to see a valley at the crossover region. If I have two subs and two speakers playing shared frequencies I would like to see a flat even small hill to indicate that there is proper summation. I listen to content that has rapid bass transients so it's something I notice. I think Dirac ART might accept a small dip or compromise in 80-120hz if it helps avoid bigger peaks and nulls elsewhere is my guess. I think there is just no way to have everything you want unless you have a lot of money to throw at fully active DSP speakers. Dirac ART is balancing crossover behavior against spatial uniformity, rather than optimizing purely for the most perfect local blend between speakers and subwoofers.

Maybe in the future I'll put 2 subs behind the coach to help in the xo region,but for now I'm satisfied. What kind of music do you listen to?

 

[thcdru2k]

Oh yeah by no means was I trying to say your graphs or system does not look good. I do a lot of single listener classical piano cello heavy or EDM and hip hop. From your graphs It does not look like you need more subs. The bass is very consistent And I do not see any cancellations that are deep