• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Expecting too much from Dirac?

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,171
Likes
4,733
Location
UK/Cheshire
Hi

Bit of a Dirac noob here. Just done my second Dirac measurement and filter. I'd assumed direc would give a much smoother response, but though it has tamed a significant mode at about 50Hz (And eliminated Bass Boom as a result) the rest of the response looks similar to Dirac turned off.

Minidsp Flex

Small room (3mx4m) carpeted, but two glass walls.

REW measurements taken from the center of the "tight" measurement box. Umik 1 used for REW and Dirac. Is this what I should be expecting?...

DIracOff.png
DiracOn.png
 

ppataki

Addicted to Fun and Learning
Joined
Aug 7, 2019
Messages
660
Likes
689
Location
Budapest
@tonycollinet Just put the mic into the main listening position and perform like 3 measurements all in that position
Dirac will apply way more aggressive correction, especially below the Schroeder frequency - that shall be visible on the FR when you measure it with REW post-correction.
If still not, you can send here the measurement (L and R) and I would be happy to generate a compensation filter for you based on that
 
Last edited:

PatentLawyer

Active Member
Forum Donor
Joined
Apr 13, 2021
Messages
281
Likes
458
Location
Deep in the Soundstage
Seems like the corrected response remains within a pretty tight margin. How does it sound? I assume you have the full-fat version of Dirac that does full range correction?

FWIW, I used Dirac on my cheap-o desktop setup of a $99 Dayton sub and ~ $150 MarkAudio full range 3" speaker kit, and got this....and it sounds really, really nice. (Sorry, I cut off the X axis, but the yellow bars are 20-20K).

1662644403984.png
 

mglobe

Active Member
Joined
Mar 18, 2022
Messages
169
Likes
223
Should you be comparing average response at listening position before and after instead of left and right? In the end, isn't that essentially what Dirac is correcting? Not sure if that is correct, but that's the way I think of it in simple terms.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
12,218
Likes
14,314
Location
Riverview FL
Bit of a Dirac noob here. Just done my second Dirac measurement and filter. I'd assumed direc would give a much smoother response, but though it has tamed a significant mode at about 50Hz (And eliminated Bass Boom as a result) the rest of the response looks similar to Dirac turned off.



I don't use Dirac, using AcourateDRC (lite version of Acourate for miniDSP OpenDRC-DI).

It corrects full-range.

AcourateDRC measures Left and Right, at the listening position, one sweep each. I put the mic between where my ears would be.

Afterwards, I measure Left, Right, and Both, at the listening position. Both speakers adds 6dB to the trace.

1/6 smoothing applied.



JBL LSR 308 before and after:

1662646840800.png


1662646871267.png



MartinLogan reQuest (electrostatic hybrid dipole)

1662646926387.png


1662646969157.png


Target "curve" was flat.

I make sure I'm centered between the speakers, overlay Left and Right Impulse Response, make sure there is only one peak.

Sometimes what you think is the middle is not.

---

On the corrected MartinLogan, you can see a severe dip at 48Hz with both speakers playing that is not evident in either the left or right traces.

Standing wave in this asymmetrical room cause a null when both speaker play.

It's not noticeable with music, and if the bass is stereo, it disappears entirely (or pops up at a different frequency) as the timing of the bass from the speakers is different.

The minor dip around 220Hz is reflection (and cancellation) from the wall behind the speakers.

Both problems also show on the JBL trace, to a lesser degree.

---

Why one measurement point"

I figure get one (where I listen) as best possible, and the rest falls where it may. and apparently, that's how my software sees it, too.

---


Measurements across the couch - about 6 feet

Trace for the center of the sofa is marked.

1662647719605.png


Impulse Responses across the sofa.

You can see how, except at the center position, there are two impulses due to the difference in distance to the two speakers as you move across the couch.

Just something you have to live with. Physics.

An extreme case of not having the microphone acoustically centered between the speakers.

1662647866218.png


More than you wanted to know, I suppose, but I have the traces and did the experiment a couple of years ago.
 
Last edited:

Peluvius

Senior Member
Forum Donor
Joined
Aug 5, 2021
Messages
498
Likes
490
Hi

Bit of a Dirac noob here. Just done my second Dirac measurement and filter. I'd assumed direc would give a much smoother response, but though it has tamed a significant mode at about 50Hz (And eliminated Bass Boom as a result) the rest of the response looks similar to Dirac turned off.

Minidsp Flex

Small room (3mx4m) carpeted, but two glass walls.

REW measurements taken from the center of the "tight" measurement box. Umik 1 used for REW and Dirac. Is this what I should be expecting?...

I am in a similar situation. Had the Flex for about two weeks now. Following advice from the forum and MiniDSP, I have found the wider focus calibration (middle one of the three) helps to improve the result as well as playing around with the frequency where Dirac applies correction. I am currently listening to a Harman adjusted curve with correction only below 300Hz. There is definitely some "user calibration" required to optimise the results and I am still playing so take my feedback as a work in progress, not an endpoint.
 
OP
tonycollinet

tonycollinet

Major Contributor
Joined
Sep 4, 2021
Messages
3,171
Likes
4,733
Location
UK/Cheshire
Thanks for replies all.

Well I've now adjusted the target curve to damp down some "brightness" and have been listening this evening - it sounds superb now. I'm not expecting that change to smooth the response, but I don't think it matters.

This second measurement was to allow me to set the receiver to analogue direct mode, and removing some port bungs from the speakers. The first measurement a few months back, was a step up, but I think this one is a bigger step again.
 
Top Bottom