• 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). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Dirac Live 3.1.1 bugs?

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Hi there, I updated Dirac Live to the latest 3.1.1 version on my Mac last night. There appears to be two bugs which I wanted to see if others are experiencing:

1. The bright high contrast colour scheme makes the text in the filter generation page where you can check/uncheck items to see on the graphs appear the same colour as the background. This makes the text illegible.

2. When loading previous version measurements, it appears 3.1.1 does some extensive updating and the program takes some time to rebuild the filters. During the rebuild, some measurements came up with an error that phase correction could not be applied. Subsequent to that, a button appeared at the bottom allowing me to try phase correction again. When clicking the button, it seemed to do something but I can’t confirm if it did indeed complete the action.

The second item may be the root cause of the main issue I’m facing with 3.1.1, the applied filters sound different than the previous version, with one set of measurements creating a filter so bad that it sounds like the speakers are wired out of phase. The sound stage for that on set is diffuse and sounds very strange. I’ve opened a ticket with Dirac in regards to the above and will let this forum know if I hear back.

At this point, I’d recommend backing up your measurements before opening them with 3.1.1 in case you don’t like the outcome. I unfortunately didn’t do that, which means I’ll have to take all the measurements again with 3.1.1 to see if I can get the sound back to how it was before. Let me know if you come across these issues as well. The amplifier I’m using with Dirac is the NAD M33.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Dirac Support have confirmed a bug for #1 and a fix will be issued in the next release. The other colour themes within the program don’t have the issue so they can be selected as a temporary fix.

Support has also confirmed that filter calculation has been updated in this release but it should not have a dramatic change like I’ve experienced. I’m sending them my log files and project files to examine.
 

Loron

Member
Forum Donor
Joined
May 17, 2021
Messages
41
Likes
33
Same here. Much worse than 3.0.14.

I used REW Instead. Not been impressed with Dirac since I have it. Wasn’t impressed with Audyssey before either. I have extensive room treatment. I am trying to optimize response by acoustic treatment, speakers and listening position location. I was using Dirac only up to 400hz and even there it was no real improvement. Now with 3.1.1 it is just really bad. Clearly much worse than my last 3.0.14.
it totally destroyed the imaging specially center image.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Still waiting on Dirac to analyze the log files. In the interim, I’ve redone all the mic measurements which helped. The first time, there was a significant loss of bass. REW showed phase issues. Redid the main listening position measurement again and the phase issue is gone.

With the new measurements, there‘s a very large difference between tightly focused and the other settings. Tightly focused narrows the sound stage and makes the centre image very defined. I’m still thinking there‘s something not right as it doesn’t sound as good as I remember it.

Will report back once Dirac has had some time to analyze the files I sent.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Dirac have confirmed they see something funky in the new filter calculations in my project related to phase and Impulse response. They’re now trying to identify the root cause, and if it’s a one off issue or something more.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Same here. Much worse than 3.0.14.

I used REW Instead. Not been impressed with Dirac since I have it. Wasn’t impressed with Audyssey before either. I have extensive room treatment. I am trying to optimize response by acoustic treatment, speakers and listening position location. I was using Dirac only up to 400hz and even there it was no real improvement. Now with 3.1.1 it is just really bad. Clearly much worse than my last 3.0.14.
it totally destroyed the imaging specially center image.
Please see my post #5 in this thread. You may want to open a help desk ticket snd send Dirac your project files. They can review them to see what’s going on. Might be the same issue I had with one set of measurements.
 

Loron

Member
Forum Donor
Joined
May 17, 2021
Messages
41
Likes
33
Thanks. I will as soon as I find the time.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Dirac is still investigating the issues, but it sounds like others are also having issues. As a result, they will be releasing a new version which will allow us to manually override some settings to bring back the correct imaging. Just FYI.
 

f1shb0n3

Active Member
Forum Donor
Joined
Aug 29, 2020
Messages
175
Likes
235
Location
Seattle Area
The second item may be the root cause of the main issue I’m facing with 3.1.1, the applied filters sound different than the previous version, with one set of measurements creating a filter so bad that it sounds like the speakers are wired out of phase. The sound stage for that on set is diffuse and sounds very strange. I’ve opened a ticket with Dirac in regards to the above and will let this forum know if I hear back.

I had exactly the same experience with Dirac Live 3.1.1 on my Mac with MiniDSP SHD - the speakers were actually out of phase - when I inverted one of them from MiniDSP it sounded just right. Rolled back to 3.0.14 and it's all good now. Appreciate you following up with them, I got lazy and just gave up as previous version is perfect :)
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Received confirmation from Dirac that they’ve identified the bug with 3.1.1, which only affects MacOS users. They recommend using 3.0.14 or Windows until a patch is released in the near future.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
The fixed version 3.1.2 is now out. Give it a try and see if it fixes the issue. I won't have time to re-measure my house until the weekend so do let us know if you try it and it works correctly.
 

Χ Ξ Σ

Senior Member
Joined
Jul 16, 2020
Messages
328
Likes
1,259
Location
UTC-8
I installed the 3.1.2 and loaded a measurement taken by the 3.1.0 and the frequency response graphic looked the same. Generated filters and they sounded normal. The frequency response graphic on the 3.1.1 looked goofy so I didn't even dare to generate filters. I have no need to take new measurement so I didn't test that.
 

Loron

Member
Forum Donor
Joined
May 17, 2021
Messages
41
Likes
33
I also installed Dirac Live 3.1.2. Result we’re kind of similar to 3.0.14. I had revert to 3.0.14 after having really bad results with 3.1.1. I am on Mac OS. I ended up disabling delay land gain compensa . Hopefully it keeps the speaker distances that I entered at set up. there is no information from Dirac on this function.

I did solve the huge volume difference issue I always had with Dirac. Not sure it is changing much compared to my user settings.

I worked hard trying to improve my room acoustics and used REW to optimize speakers placement. I keep Dirac calibration from 300 to 400 Hz down. I am still unimpressed with room correction system be it Dirac Live or Audyssey. Even when I run REW after Dirac calibration the results aren’t that much differen. It makes sense when you deal with room nodes.
I will compare with music and see if there is a change vs Direct. So far the imaging issue seems to be fixed And with the delay and gain compensation disable the volume is affected as much as when enable.
 

Stefan

Member
Joined
May 29, 2021
Messages
51
Likes
6
Two hours ago I did the first measurement with the UMIK-1 I received yesterday.

Setup:

NAD M33 on newest BluOS 3.14.26
MacBook Pro with latest MacOS Big Sur 11.6
UMIK-1 ver.2 attached to the MacBook Pro via USB-C
Dirac Live 3.1.2

Result:

Speakers are out of phase :facepalm:

Any advice?
 

Stefan

Member
Joined
May 29, 2021
Messages
51
Likes
6
Now that I have changed polarity on my speakers (first only the left channel, then only the right channel) to check which channel is actually affected, it seems that the problem is more complicated than that. None of the two changes yielded a positive result regarding the overall polarity / soundstage...
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
Two hours ago I did the first measurement with the UMIK-1 I received yesterday.

Setup:

NAD M33 on newest BluOS 3.14.26
MacBook Pro with latest MacOS Big Sur 11.6
UMIK-1 ver.2 attached to the MacBook Pro via USB-C
Dirac Live 3.1.2

Result:

Speakers are out of phase :facepalm:

Any advice?
Have you done many measurements before with Dirac? How confident are you with the process? Based on your answer, we may be able to help more. With the 3.1.1 issue, flipping the speaker cables on one speaker would fix the phase issue.
 

Stefan

Member
Joined
May 29, 2021
Messages
51
Likes
6
With the supplied mic of the M33 I did several measurements in the last few months, never had any issue. I skipped DL 3.1.1 because of the reported issue and looked forward to 3.1.2 (now with also the UMIK-1).

Very disappointed at the moment.

What I can confirm is, that loading the last older measurement with the supplied mic of the M33 with DL 3.1.2 and let DL recalculate the filters, the sound improved noticeably (in a positive way) vs. DL 3.0.14.
 
OP
F

fcracer

Member
Forum Donor
Joined
Nov 10, 2020
Messages
93
Likes
129
With the supplied mic of the M33 I did several measurements in the last few months, never had any issue. I skipped DL 3.1.1 because of the reported issue and looked forward to 3.1.2 (now with also the UMIK-1).

Very disappointed at the moment.

What I can confirm is, that loading the last older measurement with the supplied mic of the M33 with DL 3.1.2 and let DL recalculate the filters, the sound improved noticeably (in a positive way) vs. DL 3.0.14.
I’ve tested the included microphone and it wasn’t meaningfully different than my UMIK1 so go with the NAD mic if it’s better for you. I find the NAD microphone more forgiving of the MLP placement. The UMIK1 seems to be more sensitive to placement.

I love Dirac, but the measuring process isn’t very intuitive and is a bit finicky. I would highly recommend following the Hartman or Storm Audio guides to setting up Dirac (https://www.harmanluxuryaudionews.com/2019/07/23/dirac-live-hints-and-tips/).

A couple of mistakes I made when learning:
1. Make sure to not take measurements too close to boundary walls (must be >45 cm away!)
2. The MLP measurement is by far the most important and must be perfectly positioned (I use a laser measure)
3. Each measurement should be >30 cm away from any other measurement
4. The microphone should point up
5. The microphone should have line of sight to the speakers
6. The room should be as quiet as possible
7. Treat volume and mic sensitivity as signal to noise. Increase volume to increase signal, reduce mic sensitivity to reduce noise
8. Try one of the Harman curves at first because they’ll give some bass that you’re likely accustomed to with all the room modes

There‘s an amazing Storm Audio seminar with the Dirac guys on YouTube that’s 1hr:45min that has many nuggets of info on measuring and Dirac, but it’s something you can watch in the future once you have one good set of measurements.
 
Top Bottom