[Giogio]

Hello guys.
 
Lately I was given a test unit of the XTZ Headphones Divine, and a license for the Windows version of Dirac, so that I could test and review them.
Well, I am definitely impressed by how Dirac changes the sound of these headphones, which are good by themselves but become amazing with Dirac.
 
And I would like to understand more about Dirac.
First of all, I had understood that Dirac does not apply generic effects and is not meant to be used on any device, because it applies FR and IR correction and for that you need to record the sound of a specific headphone or speakers with a dummy head or room analyser.
So I do not understand well some comments of some people who seem to be using Dirac on any device, just like that.
 
Second, I am not familiar with what IR (impulse response) really is and how it affects the sound. I am not interested in mathematical formulas, I just want to understand it in practical terms, and I have found no explanation for this.
And about FR (frequency response) I find confusing that some people use it to indicate the range of frequencies (e.g. 20-20000) which a headphone can play, while other people use it to indicate HOW those frequencies are played, how loud, how much of them, like "eq curve". Or at least this is what I understood, and also what I think that Dirac refers to when they talk of FR correction.
So, without understanding well how FR and IR affect the sound, I also cannot understand how their correction in Dirac affects the sound.
Maybe you can help me with this.
 
But the fact is, the XTZ Divine have the best highs I have ever head on anything.
They've used a Dummy Head and created a Filter which make the EQ curve neutral, so, what was too much went down, what was too few went up.
But I am sure that Dirac is NOT just correcting the EQ curve, or I could get those highs on the Divine with just some EQ. And I definitely can NOT.
Besides, I have tried Dirac with the Divine filters on my ATH-WS99BT and the highs became wonderful there too. Which I also can NOT get with just some EQ.
So, Dirac is doing something else. WHAT???
Now I became so addicted to these clear highs that I cannot use my beloved ATH (which I still prefer over the XTZ) without Dirac, no matter if the Divine filters are not optimised for it.
 
So I am curious. What does Dirac really do?
 
(btw, such a question is more for Computer Audio or for Sound Science?)

 

[overyond]

i'm a fresh in this forum, and i saw your tread carefully, though i'm sorry for your question that i can't answer for you, i have also learnt a lot form your articles. thanks.

 

[Joe Bloggs]

How did you find the highs of the Divine improve? Also is there more highs or less highs in quantity, with the filter active?

Also, are you aware of the XTZ player app for iPhone and Android? Would you download one of those and tell us how the sound from that compares with the PC app?

If you can't tell, I also own a pair of these and am interested in your thoughts

 

[Giogio]

How did you find the highs of the Divine improve? Also is there more highs or less highs in quantity, with the filter active?

Also, are you aware of the XTZ player app for iPhone and Android? Would you download one of those and tell us how the sound from that compares with the PC app?

If you can't tell, I also own a pair of these and am interested in your thoughts

I do not have any iOS device, so I am afraid I cannot try the iOS XTZ App, but I was told from XTZ that it is based on the Dirac HD Player which was/is to be found on Apple Store but which is optimised for Apple Earbuds.
In other words, the XTZ iOS App is Dirac and should not sound much different than the PC/Mac software, although I was also told that the latter has a newer algorithm.
The Android App is not Dirac. They are working on it. They want to find the right programmer. Apparently working in Android is more tricky.
But it still sounds very good. Somehow a tight more warm. The highs are still very good, and you still can see that it is not something achievable with just EQ, but they are a bit less "super detailed" than in the Dirac for Desktop.
 
Which answers to your first question. With Dirac I have found that somehow the focus of the high range is shifted on the very high frequencies. Without Dirac all the push is on the mid-highs, which are well forward. With Dirac the mid-highs are still there and may even be more detailed (as Dirac makes all more detailed) but are now no more so forward, while suddenly the upper-highs are forward in a glorious yet not aggressive not harsh way, even those you did not even know they were there, those you maybe had never listened in that song with any other headphones (unless you have a VERY good headphone, surely not with any BT Headphone).
I cannot imagine why anybody who like jazz, for example, would ever possibly want to use again anything else, after trying Dirac.
 
And, how do you like your Divine?
 
@overyond, you're welcome

 

[Giogio]

I had posted a shortcut to this thread on the Sound Science section but forgot to put a link to here, so there were replies there, which I did not mean to happen.
Now it will look strange but I am going to ask here my doubts about what answered there:
 
I was told that Dirac does FIR, IIR, and parametric, by analysing the system through recordings made in several spots of the room (so we are talking of speakers now, because for headphones it would be used a dummy head I suppose). The result are some FIR and PEQ parameters which are made into a filter which can be applied to the sound through virtually any DSP (so in this case the DSP would just be a host, and it would not affect the sound with its own proprietary algorithm but just with the filter created with Dirac).
Now, first of all I would like to know if what I have just said sound correct.
Then I would like if you can give me a not formulaic, not mathematical, not complicated technical engineering scientific definition, but a practical explanation which can make understand what they (FIR, IIR, "parametric", PEQ) are, what they do to the sound and who does the sound improve when Dirac messes with them.
Thanks

 

[Joe Bloggs]

The stuff they said apply mostly to loudspeaker system room correction, not the Divine headphones you are hearing.

Also check your pm (in a few minutes' time)

 

[Giogio]

Not even if a Dummy Head is used to record the Headphones performance?
I mean, what can be the difference between analysing a speaker system recording it with a room analyser or analysing headphones recording them with a dummy head?

 

[Joe Bloggs]

Oh, I missed the part about dummy heads. Yes, in that case it applies.

The tutorial I posted today deals with practical use of parametric EQ
http://www.head-fi.org/t/794467/how-to-equalize-your-headphones-2016-update

 

[RRod]

Originally Posted by Giogio /img/forum/go_quote.gif

I was told that Dirac does FIR, IIR, and parametric, by analysing the system through recordings made in several spots of the room (so we are talking of speakers now, because for headphones it would be used a dummy head I suppose). The result are some FIR and PEQ parameters which are made into a filter which can be applied to the sound through virtually any DSP (so in this case the DSP would just be a host, and it would not affect the sound with its own proprietary algorithm but just with the filter created with Dirac).
Now, first of all I would like to know if what I have just said sound correct.
Then I would like if you can give me a not formulaic, not mathematical, not complicated technical engineering scientific definition, but a practical explanation which can make understand what they (FIR, IIR, "parametric", PEQ) are, what they do to the sound and who does the sound improve when Dirac messes with them.
Thanks

 
Yes that sounds correct. FIR and IIR are two different classes of filters. Parametric EQ is a type of EQ that allows you to easily adjust the parameters (options) of each individual filter. The goal of this kind of software is to measure the response of your speaker system and then match it up to a specified response by using the filters. In cases where you want to adjust more than one listening position, you can make measurements at each position and then average them (what the exact algorithm Dirac uses, I can't say). The other 1/2 of Dirac appears to be their own audio back-end that can apply the adjustments system-wide.
 
The filters improve the sound by adjusting:
a) how much amplitude each frequency gets as it goes through the system (the frequency response)
b) how much offset/delay each frequency gets as it goes through the system (the phase)
 
Finding out the needed corrections is done by computing an impulse response of the system. Theoretically you would just measure the response to an actual unit impulse, but that has many real-world issues (e.g. noise, distortion), so there are a variety of techniques for how do get a "best" measurement. From what I've seen Dirac has a couple of methods it can use to make the measurement. Once you know the corrections, the filters you use depend upon them, but how Dirac chooses a "best" set, I don't know.