• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

We don't talk about Dirac enough in this forum.

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
After using the 'focused imaging' arrangement on Dirac 3.0 for a week or so, I'm very very impressed by how well it performs. My bass response sounds better than anything I could achieve (or have achieved) myself manually (or anyone without a proper education on acoustics for that matter). Also, apparently there is something called a sound stage that I didn't have before.

I really thought it was expectation bias, but I can ABX the Dirac filters (MiniDSP device) and this thing is really doing God's work on my system.

I mean it won't fix any speakers that are inherently flawed, but I feel it did a good job separating the effect of the room and the response of my speakers, even though it took like only 10 measurements in total.

It was honestly so good, that I firmly believe now that unless you're a therapist or doing a job that only a human can do, a robot will replace your job eventually.

I feel like among the things that are talked about in this forum, these sort of solutions (that arguably has the most impact) are rarely talked about or recommended, and it's mostly about buying the 'best gear' or having the 'prettiest toy' with very little evidence (excluding speakers and headphones) on how audible that investment will be.

So feel free to share your experiences with Dirac and this thread, perhaps we can inspire other people to invest in it.
 
Last edited:

napilopez

Major Contributor
Forum Donor
Joined
Oct 17, 2018
Messages
2,111
Likes
8,433
Location
NYC
After using the 'focused imaging' arrangement on Dirac 3.0 for a week or so, I'm very very impressed by how well it performs. My bass response sounds better than anything I could achieve (or have achieved) myself manually (or anyone without a proper education on acoustics for that matter). Also, apparently there is something called a sound stage that I didn't have before.

I really thought it was expectation bias, but I can ABX the Dirac filters (MiniDSP device) and this thing is really doing God's work on my system.

I mean it won't fix any speakers that are inherently flawed, but I feel it did a good job separating the effect of the room and the response of my speakers, even though it took like 10 measurements in total.

It was honestly so good, that I firmly believe now that unless you're a therapist or doing a job that only a human can do, a robot will replace your job eventually.

I feel like among the things that are talked about in this forum, these sort of solutions (that arguably has the most impact) are rarely talked about or recommended, and it's mostly about buying the 'best gear' or having the 'prettiest toy' with very little evidence (excluding speakers and headphones) on how audible that investment will be.

So feel free to share your experiences with Dirac and this thread, perhaps we can inspire other people to invest in it.

Dirac is good stuff, and one of the few automated room correction systems that actually knows what it's doing. But I feel like we talk about it all the time! Maybe that's just me.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,201
Likes
16,983
Location
Riverview FL
I try not to talk about things with which I have no experience, other than an aborted trial a few years ago, where I didn't care to submit to what it required of me, in terms of using the PC as the convolver, multiple measurements, and lots of money.

In my case, beheading the bass peaks manually will get me most of the way there, and I rarely listen critically to what the PC emits through its USB or HDMI outputs anyway.

There, I've talked about Dirac enough.

Unless you mean Paul Dirac, in which case more conversation might be interesting to me.
 
OP
A

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
I try not to talk about things with which I have no experience, other than an aborted trial a few years ago, and didn't care to submit to what it required of me, in terms of using the PC as the convolver, multiple measurements, and lots of money.

In my case, beheading the bass peaks manually will get me most of the way there, and I rarely listen critically to what the PC emits through its USB or HDMI outputs.

There, I've talked about Dirac enough.

Unless you mean Paul Dirac, in which case more conversation might be interesting to me.

Do you tend to listen at near-field or far-field?

I recommend you give it another try if you do far-field listening. It’s now in version 3.0 so it might have gone a long way since then.

I already own a miniDSP so i just paid 200$ for the ‘Dirac upgrade’. In comparison, the Windows desktop version is 400$.

I think it really shines in dealing with room modes and frequencies below 500 Hz. Afterwards, your measurement mic gets too directive and you need to pay some attention if you have a speaker with a questionable Spinorama.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Do you tend to listen at near-field or far-field?

I recommend you give it another try if you do far-field listening. It’s now in version 3.0 so it might have gone a long way since then.

I already own a miniDSP so i just paid 200$ for the ‘Dirac upgrade’. In comparison, the Windows desktop version is 400$.

I think it really shines in dealing with room modes and frequencies below 500 Hz. Afterwards, your measurement mic gets too directive and you need to pay some attention if you have a speaker with a questionable Spinorama.

I don't think Dirac would be able to fix @RayDunzl's bass (phase cancellation? can't remember but I think it's in the sub-bass) since there isn't enough FIR filter taps in the miniDSP -- if it's even possible to fix it effectively just that way.
 

bluefuzz

Major Contributor
Joined
Jan 17, 2020
Messages
1,043
Likes
1,774
this thing is really doing God's work on my system
I agree, Dirac is life changing. However, I feel it is talked about plenty on this forum. It is pretty much universally recommended to try Dirac first before buying a new DAC, amp or speakers if you're not already using Dirac.

It was honestly so good, that I firmly believe now that unless you're a therapist or doing a job that only a human can do, a robot will replace your job eventually.
One of the first 'AI' programmes back in the '60s was a computer 'therapist' that simply echoed peoples problems back at them as a question. It was very successful and many people found it much more 'understanding' than a real human therapist ... ;-)
 
OP
A

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
I don't think Dirac would be able to fix @RayDunzl's bass (phase cancellation? can't remember but I think it's in the sub-bass) since there isn't enough FIR filter taps in the miniDSP -- if it's even possible to fix it effectively just that way.

Perhaps, but he said that he's using the windows PC convolver. So he has infinite taps.

Also some MiniDSP products can do 10024 taps per channel, wouldn't that be enough for anything?
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,201
Likes
16,983
Location
Riverview FL
lso some MiniDSP products can do 10024 taps per channel, wouldn't that be enough for anything?

Open a copy of rePhase and set it for 10,024 taps and see.

It will give frequency and phase lines to compare what you want and what you will get in the electrical realm.

Whatever low number of taps you have, the the low frequency performance of the FIR filter is diminished as the sample rate rises.

My OpenDRC-DI has 6144 taps for each channel and runs at 48kHz. It can make a broad phase adjustment in the lows, but not focus on the phase of some particular problem low frequency or narrow band.

My little cancellation problem is at 48Hz, It is measureable (that's how I discovered it), but doesn't seem to be audible with musical content.

I don't have the education in digital signal processing to comment any more than that (and may be all wrong already).
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,201
Likes
16,983
Location
Riverview FL
Perhaps, but he said that he's using the windows PC convolver. So he has infinite taps.

No, I don't use the PC.

Maybe I could now, but at the time, I didn't have a way to route the sound from CD Transport, HDRadio, nor TV through the PC and back to the rack.

I probably could now, via a Focusrite 4Pre USB which I use as a mixer, but have other immediate interests consuming my passions and wads of cash.
 
OP
A

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
Open a copy of rePhase and set it for 10,024 taps and see.

It will give frequency and phase lines to compare what you want and what you will get in the electrical realm.

Whatever low number of taps you have, the the low frequency performance of the FIR filter is diminished as the sample rate rises.

My OpenDRC-DI has 6144 taps for each channel and runs at 48kHz. It can make a broad phase adjustment in the lows, but not focus on the phase of some particular problem frequency.

My little cancellation problem is at 48Hz, It is measureable (that's how I discovered it), but doesn't seem to be audible with musical content.

I don't have the education in digital signal processing to comment any more than that (and may be all wrong already).

Dirac uses mixed phase correction, so that number of taps can be dynamically allocated exclusively to correct the phase problem you have and it uses PEQ filters to deal with amplitude issues. Well, that's the theoretical concept behind that approach as far as I know.

It's honestly very much worth the try if you're out of options. but try the windows PC version first so you would give it a fair shot with unlimited taps.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Also some MiniDSP products can do 10024 taps per channel, wouldn't that be enough for anything?

According to @mitchco, accurate FIR correction in the bass could require 65536 or twice that amount. I don't bother as I want the lowest latency so my miniDSP's 1024 taps is actually enough for me.
 

nerdstrike

Active Member
Joined
Mar 1, 2021
Messages
257
Likes
309
Location
Cambs, UK
Unless you mean Paul Dirac, in which case more conversation might be interesting to me.

Arguably the poster-child of the forum ought to be Joseph Fourier, but I'm more of a Laplace man.

I mentioned in my Hello Forum post that Dirac Live made a hefty improvement in the mid-range for my old budget speakers and complicated lounge shape. I used the wide target, and boy did it take a while to put the mic in all the right places.
 

Daverz

Major Contributor
Joined
Mar 17, 2019
Messages
1,294
Likes
1,451
I'm still using DRC-FIR to good effect. I've been investigating some of the commercial room correction software: MathAudio, Audiolense, and Acourate. Haven't considered Dirac at their prices for the PC version, and I'm not interested in any new hardware like miniDSP. Any others I should consider?
 

Alexanderc

Addicted to Fun and Learning
Forum Donor
Joined
Jun 11, 2019
Messages
640
Likes
1,011
Location
Florida, USA
After using the 'focused imaging' arrangement on Dirac 3.0 for a week or so, I'm very very impressed by how well it performs. My bass response sounds better than anything I could achieve (or have achieved) myself manually (or anyone without a proper education on acoustics for that matter). Also, apparently there is something called a sound stage that I didn't have before.

I really thought it was expectation bias, but I can ABX the Dirac filters (MiniDSP device) and this thing is really doing God's work on my system.

I mean it won't fix any speakers that are inherently flawed, but I feel it did a good job separating the effect of the room and the response of my speakers, even though it took like only 10 measurements in total.

It was honestly so good, that I firmly believe now that unless you're a therapist or doing a job that only a human can do, a robot will replace your job eventually.

I feel like among the things that are talked about in this forum, these sort of solutions (that arguably has the most impact) are rarely talked about or recommended, and it's mostly about buying the 'best gear' or having the 'prettiest toy' with very little evidence (excluding speakers and headphones) on how audible that investment will be.

So feel free to share your experiences with Dirac and this thread, perhaps we can inspire other people to invest in it.
Once you've been around a little while, you'll see everybody gets a, "the thing that will help your system most is room correction," comment. There may not be a lot of threads specifically about Dirac, but it's recommended constantly and the products which have room correction built in often have long digressions about the specifics of the functionality.
 

phoenixdogfan

Major Contributor
Forum Donor
Joined
Nov 6, 2018
Messages
3,297
Likes
5,077
Location
Nashville
I think the thing that would truly be life changing with respect to Dirac for me would be the release of their Multiple Bass Management for the PC version of the software. Anyone know when that's happening?
 

onion

Senior Member
Joined
Mar 5, 2019
Messages
338
Likes
379
I already own a miniDSP so i just paid 200$ for the ‘Dirac upgrade’. In comparison, the Windows desktop version is 400$.
Does the upgrade allow multi-sub adjustment (4 subs)?
 
OP
A

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
Does the upgrade allow multi-sub adjustment (4 subs)?

I haven’t checked, but you can EQ two inputs. If you have a crossover unit then the following four (or more) outputs will get EQed as a result.
 

BullBuchanan

Active Member
Joined
Mar 27, 2019
Messages
114
Likes
27
I just ran through the setup of the Dirac 3 trial tonight and it was an unmitigated nightmare. I'm a Product Manager by trade who's been developing consumer facing software for 15 years and the entire experience left me shaking mad trying to get it to work.

After installing everything I was greeted with a message that I didn't have a license and was prompted by a non direct link to acquire one. After getting that, I still couldn't figure out how to get the license to connect to the software. After a long while of troubleshooting I saw in the bottom left hand corner I was auto signed in to an old account from when I joined the Dirac 2 beta that never worked for me either. Ok, fine. Probably not a primary use case.

I go into the calibration section and I have no idea what do do. No call to action, so I click on everything - nothing happens. I.play the noise tone and nothing. I decide to turn it up, and once it's blaring loud I see a tiny little black line that represents the meter. Ok. It doesn't tell me when to stop so I turn it up as loud as I can handle, which is only halfway in the blue. Fine.

I run the tests and the first one is so loud it's painful. I look at the dots on the screen for where to place the mic and it just says Next. So, I guess at where it wants me to put it. Once I'm all done I accidently mouse over one of the dots and it tells me the position - that would have been useful info to have. Much to my surprise the entire thing is backwards. The right most dot represents "far left". They did the entire thing from the perspective of the person sitting in the chair. Who on earth would do that intuitively?

I view the filter page and again, just a total guess as to what I'm supposed to be doing here. I stumble my way through it and export the filter which loads into the processor without issue. Yay!

This is where the nightmare begins...

I soon realize that the filter isn't having any effect. I'm a long time user of EAPO, so I try a bunch of known troubleshooting. I restart the computer, uninstall all the software, reinstall - nothing. I just can't get the virtual device to show up. I reread the docs 5 times, scour the net for clues, try to contact support only to be respected to the same knowledge base and quick start guide and I'm ready to throw my keyboard through the television.

Finally I find a doc that talks about using vsthost and voicemeeter. Voicemeeter is a total dealbreaker because it massively degrades sound quality when it decides to and chews through processor cycles, but I figure I can use vb cable. I set up the VST host they recommend - the first one deadends on the link from the docs and come to find out you either need to subscribe to them or build not from a github repo. More chores for a $500 app.

I find a different free one they recommend and the instructions are wrong in several places. I get it all setup to find that in order to work, the Dirac processor has to run inside this janky vst app that seems like it's from 1994. If I close it Dirac stops. Whatever.

How does it sound? Hard to say. I was so pissed off by now that any impartiality went out the window. That said, the dialogue resonances I was trying to correct were still audible, so it's not even like it paid off in the end.

Very long story short - I wouldn't use this for free in its current incarnation, let alone shell out $500 to do some personal home theater EQ. I'll stick with REW until they figure out how to make good a user experience with results worth the investment.
 
Last edited:
D

Deleted member 65

Guest
Installed Dirac Live Stereo 3.0 a few weeks back. Installation was smooth sailing, just followed this Quick guide: https://confluence.dirac.services/pages/viewpage.action?pageId=142863214

Very happy with result, best audio investment I’ve done in years. Tried earlier to correct using REW and manually implement in Roon. Failed due to lack of talent & patience. Now I’m all set, just needs more power then my Purifi amp can deliver since Dirac taxed it quite a lot.
 

Coach_Kaarlo

Active Member
Joined
Apr 27, 2020
Messages
196
Likes
222
Location
Sydney
I just ran through the setup of the Dirac 3 trial tonight and it was an unmitigated nightmare. I'm a Product Manager by trade who's been developing consumer having software for 15 years and the entire experience left me shaking mad trying to get it to work.

After installing everything I was greeted with a message that I didn't have a license and was prompted by a non direct link to acquire one. After getting that, I still couldn't figure out how to get the license to connect to the software. After a long while of troubleshooting I saw in the bottom left hand corner I was auto signed in to an old account from when I joined the Dirac 2 beta that never worked for me either. Ok, fine. Probably not a primary use case.

I go into the calibration section and I have no idea what do do. No call to action, so I click on everything - nothing happens. I.play the noise tone and nothing. I decide to turn it up, and once it's blaring loud I see a tiny little black line that represents the meter. Ok. It doesn't tell me when to stop so I turn it up as loud as I can handle, which is only halfway in the blue. Fine.

I run the tests and the first one is so loud it's painful. I look at the dots on the screen for where to place the mic and it just says Next. So, I guess at where it wants me to put it. Once I'm all done I accidently mouse over one of the dots and it tells me the position - that would have been useful info to have. Much to my surprise the entire thing is backwards. The right most dot represents "far left". They did the entire thing from the perspective of the person sitting in the chair. Who on earth would do that intuitively?

I view the filter page and again, just a total guess as to what I'm supposed to be doing here. I dumble my way through it and export the filter which loads into the processor without issue. Yay!

This is where the nightmare begins...

I soon realize that the filter isn't having any effect. I'm a long time user of EAPO, so I try a bunch of known troubleshooting. I restart the computer, uninstall all the software, reinstall - nothing. I just can't get the virtual device to show up. I reread the docs 5 times, scour the net for clues, try to contact support only to be respected to the same knowledge base and quick start guide and I'm ready to throw my keyboard through the television.

Finally I find a doc that talks about using vsthost and voicemeeter. Voicemeeter is a total dealbreaker because it massively degrades sound quality when it decides to and chews through processor cycles, but I figure I can use vb cable. I set up the VST host they recommend - the first one deadends on the link from the docs and come to find out you either need to subscribe to them or build not from a github repo. More chores for a $500 app.

I find a different free one they recommend and the instructions are wrong in several places. I get it all setup to find that in order to work, the disc processor has to run inside this janky vst app that seems like it's from 1994. If I close it Dirac stops. Whatever.

How does it sound? Hard to say. I was so pissed off by now that any impartiality went out the window. That said, the dialogue resonances I was trying to correct were still audible, so it's not even like it paid off in the end.

Very long story short - I wouldn't use this for free in its current incarnation, let alone shell out $500 to do some personal home theater EQ. I'll stick with REW until they figure out how to make good a good experience with results worth the investment.

Ever wonder if it’s just you?
 
Top Bottom