Why does DIRAC completely remove bass response (KEF R3 + sub + minidsp flex)

[Holdt]

Not really, if the crossover is not summing to zero anechoicly then one of the two drivers is receiving a significantly higher voltage than the other. there is no other way around it.

The key word in your comment here being "anechoic" which definitely not is at listening position.

 

[abdo123]

The key word in your comment here being "anechoic" which definitely not is at listening position.

I don’t think you’re comprehending what i’m saying to be honest.

 

[Tangband]

Not really, if the crossover is not summing to zero anechoicly then one of the two drivers is receiving a significantly higher voltage than the other. there is no other way around it.

The correct crossover slopes are a combination of the room, and the frequency response of the loudspeakers . A closed box loudspeaker has a certain roll off below the box resonance frequency . A ported loudspeaker falls 24 dB/ oct below tuning frequency . Its very complicated because we also have the room gain in a real room - all I can say , you can go ahead and only rely on measurements but if that sound less good you have to combine this with listening .

Different orders of crossover slopes sounds different even if the frequency response looks the same .

We always ( as you know ) listens to the acoustical slopes .

 

[Holdt]

I don’t think you’re comprehending what i’m saying to be honest.

Electrically your are absolutely right. Then theory is correct. But the best sounding and measuring crossover config isn't necessarily the same as the theoretically correct one. And as we (I speak for my self) do not live in an anechoic chamber but listen to the music in a living room, at a distance, then it really isn't so important what the electrical and theoretical correct xover is but rather the one that sums the flattest around the point of xover at the main listening pos. and sounds/blends the best.

 

[Geertidow]

However, won't this put a major burden on your speakers if you use natural rolloff as part of your xo?

 

[abdo123]

However, won't this put a major burden on your speakers if you use natural rolloff as part of your xo?

Not at all, and you kind of have to with most speakers. If you don’t use it (and cross at 40Hz or so) then you’re putting a major burden on the the speakers.

 

[Geertidow]

I will test and measure different xo's.
I did like @abdo123 's suggestion of 12/db oct rolloff at 60 Hz for the KEF R3 and a LR 24 db/oct rolloff for the sub.
It was nice to be able to remove that 80 hz huge dip by placing the speakers 25 cm (rear side of speaker) from the front wall.

 

[Holdt]

I will test and measure different xo's.
I did like @abdo123 's suggestion of 12/db oct rolloff at 60 Hz for the KEF R3 and a LR 24 db/oct rolloff for the sub.
It was nice to be able to remove that 80 hz huge dip by placing the speakers 25 cm (rear side of speaker) from the front wall.

I would try a higher crossover at 80 Hz. Then you could probably place the speakers where they were and get more headroom. It would be nice to see some measurements from your experiments.

 

[Geertidow]

Oke gentlemen, I have spent a lot of time positioning the speakers.
I likte them 30 cm from the back wall and 50 cm from the side walls.
If you like to load measurements in REW and see the results: https://we.tl/t-4hqtiPYaH7.
The 100 hz crossover is actually not sounding so bad, but neither is no sub, only the KEF R3

However, the sub does really add that extra layer.
A 100 hz 12 db/oct xo also seems a good choice.

But, more listening is needed to determine the best xo in this specific room.

 

[Geertidow]

I think it has to do that if I move farther away from the front wall that 80 hz dip increases greatly. The price to pay is the 130 hz dip

 

[Geertidow]

As requested by the tech support, redid everything. Green no dirac, red dirac.
What is going on here (regardless of minor details).
XO at 70 hz, 12 db/oct

 

[Holdt]

As requested by the tech support, redid everything. Green no dirac, red dirac.
What is going on here (regardless of minor details).
XO at 70 hz, 12 db/oct

Your screenshot from Dirac. Is those two lines R+L?
I like to have the "corrected" box checked so I can see what the predicted response is. Maybe that could reveal a clue about What's going on.

I also don't use the pre defined curves. I should not make a difference, but you could try to do a manual curve by points. Imitating the standard curve if you like it. But try to check the predicted box..

 

[Geertidow]

Reloading everything now results in a normal bass. I am utterly clueless as it seems to be a software error.
The worst part, miniDSP dev team is basically saying I have to allign the sub (already did), remove all loaded PEQ's (already did).
And, to make it worse: "Having never seen this behavior before on all our customers, we wonder what setup you have here. So please provide (..)" followed a few email later by "That's extremely odd as nobody else ever seen this issue on all the setups we have worldwide".

I understand that they want me to measure in DIRAC using a completely flat curve (only an XO), but I kindoff like to PEQ a bit first and then use DIRAC.
The device should be able to do this right?

Furthermore, reloading the same DIRAC filters which then actually works implies it really is the miniDSP.
Maybe it looses all output to the sub?

By the way I ended up with a 12 db/oct xo @70 Hz. I assume that does not overload the KEF's too much.

 

[withoutsuit]

I´ve also mixed experience with subwoofer integration & DIRAC at an minidsp DDRC24. That´s why I´m using it now without DIRAC. I think, the lack of separate subwoofer measurements (like an AVR with independent subouts or the DDRC88 with BM Plugin), can lead to some unexpected behaviour and hence may not work in some setups.

 

[Geertidow]

That's not what they think at miniDSP:

"
Note that it's all fine this being said. It's never been an issue for the past 12years we tune countless Dirac live installs and sell the platforms for all our stereo Dirac to do a 2.1 or multiway with a stereo Dirac live license. As a matter of fact, it's indeed better. Why? Because using only 2 x channels for say a 2.1, you are indeed Sweeping and tuning BOTH the main and the sub = You are correcting the system at the transition/crossover section. If you were using 3ch, you indeed wouldn't get as good of a crossover transition/tuning because Dirac wouldn't tune the sub to main section. Do you understand what we mean?
Here are some app notes that all our customers are following:
https://www.minidsp.com/applications/subwoofer-tuning/sub-integration-dirac

Even for a multiway speaker and even if you're using a DDRC-88A, e.g. 3way, you do "not" want to use multiple Dirac channels. It just doesn't work as well.
https://www.minidsp.com/applications/digital-crossovers/stereo-34way-xover

Same applies for multisub.. Sometimes it's better to a)tune the speaker with basic time alignment b) Apply Dirac live on the overall system.
"

12 years of DIRAC and never has that been an issue. It really feels like an overconfident tech support.
Also since they don't seem to believe me I have really time alligned the sub.

 

[Flak]

As requested by the tech support, redid everything. Green no dirac, red dirac.
What is going on here (regardless of minor details).
XO at 70 hz, 12 db/oct

Are those REW measurements?
Are you measuring one channel only or both together?
Is the subwoofer included or not?
Thanks!

 

[peng]

I just plotted some REW graphs to check how Dirac performs/transforms my tiny Altec Lansing 2.1 computer/satellite system and I was nicely surprised to see results that corelates very well with my subjective measurements (ears/brains).

These are for "as is", no adjustment/customizing target curves yet. Dirac is able to extend the little subwoofer (again, those are just tiny desktop computer sub/sat speakers) all the way to 25 Hz, that's incredible. There are no crossover setting for those speakers, and I only have the 2.0 DL live PC version anyway, so the speakers and the tiny sub were measured together.

 

[withoutsuit]

12 years of DIRAC and never has that been an issue. It really feels like an overconfident tech support.
Also since they don't seem to believe me I have really time alligned the sub.

Thats the same impression I've got. They told me something similar ("never a problem blabla").

 

[Geertidow]

Here you see my dirac results, this with no PEQ whatsoever beforehand (XO 90 hz, 18db/oct speakers, 30 db/oct sub).
I still dont know what to make of this. I like the sound of the full range and the 500 hz, but the 300 hz and 100 hz are off.
I used to use DIRAC only up to 300 hz and this was fine.

If you compare it to manual PEQ or no DIRAC, i seem to loose a lot off bass. What am i missing here.
Measurments are at LP using sub, left and right speaker.

 

[Geertidow]

Are those REW measurements?
Are you measuring one channel only or both together?
Is the subwoofer included or not?
Thanks!

REW yes, 2.1 setup, measured at listening position through all speakers