Dirac Live's graphs... how accurate?

[flak]

When looking at Dirac Live's graphs (before and after correction) some wonder if they are accurate or not, the reasoning is that the after correction graphs are calculated (or "predicted").

So it is reasonable to wonder... how accurate are those computations? (or predictions if you like)

 

In order to verify that you have to compare the Dirac Live's graph with a measurement by a dedicated application like REW, Fuzzmeasure or others.

Markus767 from AVS forum has taken advantage of the possibility of forcing Dirac Live to use one single measurement for a single point in space... this often will not produce the best listening results but it is a way of double checking how accurate Dirac Live's graphs are by comparing a single point REW measurement with the Dirac computed graph in a single point.

He has tested a test subwoofer and posted the following REW and Dirac Live images for validation purposes:

 

 

 

I'd say that those Dirac's graphs are quite consistent with REW both before and after correction

Other measurements from markus767 like the relative "before" and "after" waterfalls as well as full range measurements with different methods including one with MMM (moving mic measurement) have been posted here:

**miniDSP DDRC-88A Official Thread**8-channel AI/AO Dirac Live in a box - Page 17 - AVS Forum

 

A WBF forumer, Michael Lowe, has used Fuzzmeasure instead with his Mac and a reference Earthworks M23 microphone to validate the full range results.

This is the image he has posted of the two channels as shown by Dirac Live BEFORE and AFTER correction:

 

 

This instead is his image of the eighteen averaged measurements (both channels combined, 1/24 octave smoothing) by Fuzzmeasure BEFORE Dirac Live correction:

 

 

while this is the image of the 18 measurements by Fuzzmeasure AFTER Dirac Live correction:

 

 

Again by comparing them we can see that the Dirac Live's graphs are quite consistent.

The full thread is here: Dirac validation

 

One last fact worth noting is that common frequency response graphs show the frequency response in one single point in space... this literally means one point in space where the mic is placed, but measurements are different in different positions... even a few centimeters make a difference up to the point that what we measure at one ear is different from the other one.

So we want to correct the common behaviour of the different curves at different measurement positions as we do not listen with one ear only in a rigidly fixed single point in space.

 

Here we see a forumer's nine curves of the left channel in the nine points that define the listening area together with the curve (the lighter one) that represents the average.

 

The individual curves are shown only in the full Dirac Live version for PC/MACs while the Stereo version shows the one that is most practically useful as a user tool to modify the target, the lighter curve in this image... this is the behaviour relative to the Left channel BEFORE correction:

 

while this is the Left channel AFTER correction:

 

 

(actually Dirac Live does NOT average the nine measurements for correction but uses them to determine what changes with position and what can be corrected and what not... but that's another story)

 

Ciao, Flavio

[Youreye]

Thanks Flavio for the nice summary and sharing!

 

BTW this is my 2.1 system in my room.

[flak]

Thanks Flavio for the nice summary and sharing!

 

BTW this is my 2.1 system in my room.

 

Thanks Youreye,

 

your graphs seem pretty good, if are willing to experiment you may try dragging down the anchor point at the right of something like 2 dBs and eventually you could also lift the anchor point on the left of 2 dBs... if after careful listenings you have found the tonal balance of your speakers to be the right one for your listening room and music material it is possible that you will appreciate a similar tonal balance also after correction.

 

Good listenings

Flavio

[wgscott]

I did this experiment with Fuzzmeasure a while ago. Be careful. I blew out a tweeter.

[Youreye]

You mean to deviate from the recommended curve? Because that's what I am using now (on the graph). I might try the additional -+2db as you recommend. Thanks for the tip Flavio!

 

Ciao!

[dtb300]

Thanks Youreye,

 

your graphs seem pretty good, if are willing to experiment you may try dragging down the anchor point at the right of something like 2 dBs and eventually you could also lift the anchor point on the left of 2 dBs... if after careful listenings you have found the tonal balance of your speakers to be the right one for your listening room and music material it is possible that you will appreciate a similar tonal balance also after correction.

Flavio

 

Flavio,

 

Curious, where would you set the Anchor Points for the +-2dB gain and drop - Frequency points I am asking.....

[flak]

Flavio,

 

Curious, where would you set the Anchor Points for the +-2dB gain and drop - Frequency points I am asking.....

 

In Youreye's case I'd experiment by simply using the already existing control points

 

Flavio

[dtb300]

In Youreye's case I'd experiment by simply using the already existing control points

In the attached graph you commented on, I do not see any Anchor Points (DOTS). I would guess those would be at 40Hz and 15-16kHz (roll off points) for the +-2dB?

[flak]

In the attached graph you commented on, I do not see any Anchor Points (DOTS). I would guess those would be at 40Hz and 15-16kHz (roll off points) for the +-2dB?

 

We don't see the anchor points because the target curve is unselected and we cannot see it either... but there are two at the extremes which could make the slope sligtly more ripid.

 

Flavio

[Youreye]

It wasn't so hard to understand but for the sake of "visualization" here is the graph I came up with (thanks Flavio again). I have added one more point to it.

[armani-thomas]

How can I get REW sweeps through the DAP? Have been googling and googling, but no luck. It works when REW output in Java. But then I can't sweep anything else than L+R. With ASIO output I can choose center and surrounds, but I get no sweep sounds when I route ASIO to the DAP :-/