Is Dirac useful for music?

[fludimir]

>Thanks, Dirac Full, will give you the ability to customize beyond the house curve. There is really not much more you will gain.

Thats why I didn't already bought it, I dont wanna try to "improve" sound of my speakers  with different curve

>Looking at the REW that you posted, I don't see a ton of difference.
My eyes also don't see tons of difference, but ears hear it.. However graphs are very different if look without cal file or with EmotivaMicCalOA.txt from DiracLE or with cal files from your sig - which I should trust more?
>In fact the direc LE run with the standard calibration has less bass then direct.
Actually dirac reduced bass on some freqs - hit in bass drum is too heavy in direct but ok with dirac. But on dynamic tracks with lots of bass(Metallica songs for example) Dirac is adding huge amount of moody bass   On less heavy tracks I cant say it sounds awful , but really different from direct mode - more extensional but less crisp..
Here is picture from dirac i63.tinypic.com/2zp5lyf.jpg  - I see its trying to adjust response from 20Hz (my Dynaudio is cool speakers but they cant play so low)
So may be issue is there, for some stupid reason Dirac thinks that it should make flat response from 20Hz, no matter that it pushed 7" woofers to do that?..
There is also noticeable bass increase on 80-300Hz, but my speakers should play that without any issues.

>Check your trims, whatever you are hearing is not making it into the measurements.
Level / Tone Trims are zeros, is there any other thing that can affect? 

[ansat]

There is one other advantage of dirac full that might interest you, you can utilize the curtains on the sides to pick where dirac needs to stop making corrections. Looking at the graph, there is nothing wrong with the microphone. In the cases I found, it looked like the tweeter and woofer was blown.

There is nothing further to check with the trims.

According to the microphone and the direct vs dirac, it looks like all advantages being had. (with one really odd frequency that is hanging around in the middle of a null)



I'm kinda at a loss where things are going wrong. Unless that 20hz boost is really hurting you. Try this:

1. Tell the XMC that you have a subwoofer,

2. set the xover to 40hz and see if that fixes your problem (make sure that your subwoofer slope is 12db)


Tony

[fludimir]

Tried set crossover - sounds less boomy, but still worse than ref-stereo.

I made records with emotiva mic, maybe this will help to understand issue? Here is the files drive.google.com/folderview?id=0B1Q55deL45peM0JoWHdnLUVHa0E&usp=sharing , recorded part of Metallica's song. Mic cant perfectly reproduce what I hear, but it shows how dirac changes music in my system.

[ansat]

Jan 25, 2016 13:16:31 GMT -5 fludimir said:

Tried set crossover - sounds less boomy, but still worse than ref-stereo.

I made records with emotiva mic, maybe this will help to understand issue? Here is the files drive.google.com/folderview?id=0B1Q55deL45peM0JoWHdnLUVHa0E&usp=sharing , recorded part of Metallica's song. Mic cant perfectly reproduce what I hear, but it shows how dirac changes music in my system.

I can hear the difference, When I look at the waveform, there is quite a bit more energy in the Dirac comparing the two. I have a suspicion what the problem is. Have one more test to confirm, in direct mode, (not reference stereo) set the distance on one of the speakers greater (tweak then listen, repeat) while playing that song. I suspect that you will hear the sound start to sound more like the Dirac as it approaches time alignment.

Tony

[fludimir]

Tried that - when I'm increasing distance to left speaker sound becomes less solid and "virtual singer" is moving left. I cant say that it sounds similar to dirac.
IMO dirac is ok with time alignment, if singer appears in center in direct mode he's also in center in dirac mode. And sound when adjusting left speaker distance doesn't becomes boomy like in dirac mode.

[drtrey3]

Sadly, I do not own Dirac. But I have heard it and the effect ranged from slightly different but not necessarily better to WOW THAT SOUNDS SO MUCH BETTER! More tunes showed the benefit than just showed a difference.

Trey

[Lsc]

The boomy sound went away after I adjusted the phase on my sub. I have my sub near-field so adjusting the phase to 90 synced it better with my mains.

It's a huge improvement in sound and something that everyone should do after running Dirac. I thought Diracs impulse response would take care of phase but it didn't.

Also, all my speakers are full range with the center good to about 50Hz. I set all my speakers to 60Hz. Still testing this out.

[ansat]

Jan 26, 2016 10:30:38 GMT -5 fludimir said:

Tried that - when I'm increasing distance to left speaker sound becomes less solid and "virtual singer" is moving left. I cant say that it sounds similar to dirac.
IMO dirac is ok with time alignment, if singer appears in center in direct mode he's also in center in dirac mode. And sound when adjusting left speaker distance doesn't becomes boomy like in dirac mode.

I believe that dirac is ok as well, based on the rew measurements, there is nothing that should be boomy. I am thinking that the boomy only exists when tour right and left play bass at the same time, and in a perfect world, you would see a plus 6 todb to a speaker array. I was hoping that tmall time alignment changes would confirm this as i think either the problem is placement related, or you are hearing issues in the recording that was masked by the previous room interactions. One of the steps in determining placement of the main (and center) that i added in the last 2 years was to measure speaker interactions with each other. I use various sound modes and different channels to achieve this. (All channel stereo).

Tony

[fludimir]

>The boomy sound went away after I adjusted the phase on my sub.
I have no sub

> I am thinking that the boomy only exists when tour right and left play bass at the same time,
I turned off one monoblock but sound is less but still boomy in dirac mode

ansat, you had listen to files that I recorded - can you confirm that sound in dirac mode is issue, surely it should not sound like that? May it be that with default DiracLE house curve, where bass is increased, it should be boomy?

>based on the rew measurements
may the issue be in rew measurements? I made them with emotiva mic but dont know what cal file better to use. So I have another question - can I use xmc1 mic to make room measurements, adjust PEQ and do some acoustic treatment in my room, or need to buy another better mic like UMIK? Will also need to buy DiracFull to use that mic..

[ansat]

Jan 30, 2016 8:42:04 GMT -5 fludimir said:

>The boomy sound went away after I adjusted the phase on my sub.
I have no sub

> I am thinking that the boomy only exists when tour right and left play bass at the same time,
I turned off one monoblock but sound is less but still boomy in dirac mode

ansat, you had listen to files that I recorded - can you confirm that sound in dirac mode is issue, surely it should not sound like that? May it be that with default DiracLE house curve, where bass is increased, it should be boomy?

>based on the rew measurements
may the issue be in rew measurements? I made them with emotiva mic but dont know what cal file better to use. So I have another question - can I use xmc1 mic to make room measurements, adjust PEQ and do some acoustic treatment in my room, or need to buy another better mic like UMIK? Will also need to buy DiracFull to use that mic..

You should not need another mic. My calibration file will yeild accurate enough measurements for what you need. But i dont see many issues with your rool looking at the rew file. The only think i could see that was an issue is the amount of energy difference in the two recordings.

If your up for one more trial using just the front two speakers, delete the math cal and open dirac. It will generate a new calibration. Do a calibrated run on the front 2 speakers using the microphone positions recommended by dirac. Have the mic pointed forward instead of pointing it towards the ceiling. This should change some of the interactions on the initial sample.

As for the house curve, a +3 at 20hz and a -3 at 20k should barely be noticable to flat.

Tony

[fludimir]

So I can fully trust xmc mic plus "Matched by math" cal file for spl reading, waterfall and impulse response? I don't need 0.1db precision, but at least 1db should be. When I compared two xmc mics there were 3db difference..

It will be hard to make 9 meausements with mic pointed forward, my custom mic stand built for ceiling mic point.. will it be useful to made only 1point calibration?

[minthral]

You don't point it forward...you point it up.

[Talley]

Feb 1, 2016 12:04:23 GMT -5 minthral said:

You don't point it forward...you point it up.

I have to tell her the same darn thing.....

[fludimir]

Guys, have you read ansat's message that is above mine?

[klinemj]

I believe Tony (ansat) was referring to use of a specific cal file that is intended for using a mic facing forward. The standard file from Emotiva and if one wants to create filters for multichannel is associated with a mic facing upward (makes sense for speakers in front of, beside, and being the listening position). But, In this case,it's different.

That is why fludimir is talking about forward facing.

Mark

[ansat]

Jan 31, 2016 18:30:56 GMT -5 fludimir said:

So I can fully trust xmc mic plus "Matched by math" cal file for spl reading, waterfall and impulse response? I don't need 0.1db precision, but at least 1db should be. When I compared two xmc mics there were 3db difference..

It will be hard to make 9 meausements with mic pointed forward, my custom mic stand built for ceiling mic point.. will it be useful to made only 1point calibration?

It wont matter that much, your comparing 2 results and only looking at the delta. A single point calibration will be a good proof of concept.

Tony

[Boomzilla]

What's the cost difference between running DIRAC on a PC / Mac as a stand-alone solution vs. buying a processor (such as the XMC-1) that has the hardware built in?

To do the stand-alone, one would have to buy the program, the mic, and a hardware box to Implement the signal modification, yes? The XMC-1 would come with its own DSP-hardware & mic?

Is there any sonic advantage in going one way vs. the other? Any price advantage?

Thanks - Boom

[Talley]

Feb 2, 2016 9:29:49 GMT -5 Boomzilla said:

What's the cost difference between running DIRAC on a PC / Mac as a stand-alone solution vs. buying a processor (such as the XMC-1) that has the hardware built in?

To do the stand-alone, one would have to buy the program, the mic, and a hardware box to Implement the signal modification, yes? The XMC-1 would come with its own DSP-hardware & mic?

Is there any sonic advantage in going one way vs. the other? Any price advantage?

Thanks - Boom

XMC-1 using Dirac has to sample at 48k.  The PC solution offers 192k capability.  The standalone mini-dsp 8 channel that just release can do 192k as well.  Not sure if this makes a difference to you but I'd prefer to keep the communication chain to a minimum.

So you haven't tied Dirac yet?

[Boomzilla]

Hi Talley - No, I haven't. But I'm beginning to think that it might be an advantage to me. Despite the fact that other DSP solutions have been ineffective, I'm being told by many here on the Lounge that the DIRAC algorithms are superior and should be able to do what the other DSPs haven't. We'll see...

[Talley]

Feb 2, 2016 13:13:23 GMT -5 Boomzilla said:

Hi Talley - No, I haven't. But I'm beginning to think that it might be an advantage to me. Despite the fact that other DSP solutions have been ineffective, I'm being told by many here on the Lounge that the DIRAC algorithms are superior and should be able to do what the other DSPs haven't. We'll see...

I had used Audyssey XT32 prior and the bass was always lifeless and I had to increase the gain to the sub to make it feel right.

With the XMC-1 and the Dirac the sound out of the box is pretty darn close and a simple curve to the low end, I stick to +2db at 80hz and +4.5db at 30hz and the music is down right perfect.  This is using my ultra towers crossed over at 40hz to my HSU sub.   The Dirac makes the sub integration seemless.  One thing to note though is check your levels w/ the onboard pink noise after dirac as I've found they are not exactly perfect they can be +/- about .5 to 1db between all of them.  Not sure why but I always check and adjust that last bit.

Seriously when it's done right there difference between my reference stereo and Dirac are not super strong amazing but the soundstage is wider w/ Dirac with better pinpointing of instruments but the best thing is the flattening out of the FR and how it handles timing adjustments.  It does something that works and it's definately noticable but if you don't get like some super amazing results it should be an improvement of anykind though.    I've also found that adjusting the sliders on the Dirac software to limit it's ability to work beyond 4khz helps maintain top end.  You can adjust sliders so if you only wanted to work out the 20hz to 600hz areas then adjust the sliders to window only those frequencies and then you can EQ them to your liking.  Out of the box is pretty decent... has a slight slope compared to Audyssey who would try to produce ruler flat FR.  Ruler flat is dull and sounds like crap.  You don't want boomy bass but you need a slight bump for realistic sounds.

I always dial in my sound using a cello and then a drum solo with the kick drum.  always.  This way I can verify proper EQ of the bass.

I've attached two images.  One is my ultra speakers and another is the sub.  The sub is the straight out of the box test audyssey vs. dirac.  You can see audyssey wants to flatten the bass out while Dirac standard has a slight curve yielding better low end bass RIGHT OUT OF THE BOX.... which with dirac you can adjust this curve to your liking.

The speaker you can see the difference between my best ever left channel audyssey results and I think my second or third left channel dirac result only bumping up the bass.  I've since adusted the low boost down to +4.5db instead of the 6.   I overlayed the 1/3 smoothing (which is what we can relate to hearing) over the 1/12 smoothing so you can see the curves.  Pretty close but Dirac does a better job and again... allows the curve.

I recommend when you try your friends Dirac setup that you setup a +2db at 80hz and +4db to maybe a +5db at 30hz for music and let that ride.  I think you'll be happy with those results.

Attachments: