Jump to content
IGNORED

New second generation Dirac Live


Recommended Posts

On 12/9/2019 at 9:22 PM, GoldenEar said:

Hi all,

the new version is out now! It plays up to 192 kHz in my system and sounds great. However, I am unable to claim my license at the Dirac portal. It keeps saying unknown key. Any of you has been successful?

 

Regards Kay

 

Hi, to which version of processor did you refer?

Link to comment
1 minute ago, GoldenEar said:

 

In my system (Mac Mini late 2012 with Mojave) I still can use Dirac Processor 1.2 as an endpoint for Roon, so probably anything else was wrong in your setup. However, I can confirm that the solution with Audio Hijack/Roon/Dirac Live Processor (new version 1.0) sounds worse than Audirvana with the same processor. With Dirac Live 1.2 it was the other way round.

 

Regards

Kay

 

Of course it will sound worse. Dirac should be also available as a plug-in for Roon DSP engine, or otherwise we will wait fot VST support there.

 

Another option on Windows not Mac I use is having output redirected to JRiver WDM driver, sound is good. But having multiple applications is awful

Link to comment
On 12/10/2019 at 6:34 PM, jcn3 said:

 

 

if you go into the dirac portal (portal.dirac.se) and then go to "user licenses and installation records", do you see v2 listed?  my understanding is that v1 licensees will be upgraded to v2, but haven't see any details at this point.

 

i'm licensed for v1, but haven't messed around with v2 -- was waiting for things to settle down and the kinks worked out.

 

Any news here?

If I go to old portal, I see my 1.2 Full license - 2 installations, but in the new portal it is empty,

 

Link to comment
27 minutes ago, michael123 said:

Hi

 

I am stuck in the middle of measurements with the new Dirac Live/Processor (Windows/JRiver) - where do I set an output device??

 

In previous betas, there was a choice of drivers, and I picked up one related to DAC.

 

Now, when I try to calibrate output volume, I don't hear any sound!

 

I figured out that I need to play a music in JRiver while measuring in Dirac Live.

 

But going to the last stage of filter export, I got an error I have no license for phase correction.

 

What should I do?

Link to comment
1 minute ago, flak said:

 

I'm sorry for the inconvenience, we are looking into that but on the other hand I'd like to outline that we upgrade free of charge even licenses that have been purchased in 2012 and the new version is expected to be valuable for many other years to come...

a good long term investment in my (biased) opinion :)

 

Flavio, could you please check why do I get missing phase correction license during filter export step?

 

I replied to the case

 

Thanks

Link to comment
6 minutes ago, shinyc said:

My playback chain has a very large latency. This creates an issue when measuring with Dirac. Test tone comes out some 15 seconds after clicking the measure button and by that time the test tone comes out, the progress meter on the Dirac Live screen hits 98% already and Dirac gives up listening to the rest of the test tone which does play in its entirety to the end though. My playback chain involves foobar, upnp renderer, jplay femto upnp server, jplay streamer. I recall HQplayer exhibited a very large latency as well. Does anyone else have similar problem? I already asked Dirac if they can elongation the test tone listening window. 

 

You can probably measure in JRiver or foobar, and then use the filter in your playback chain

Link to comment
7 minutes ago, shinyc said:

@michael123 Thanks for your kind comment.

 

Yes, I will probably resort to that idea for now but the whole idea behind vst chain measurement is that we can use the very same playback chain to measure the sound characteristics. I find jplay femto sound quite different than my other playback chains and I'd love to be able to use my preferred choice of playback combo to do the measurement.

 

But the latency you can and probably should discuss with Marcin IMHO

Link to comment
On 12/16/2019 at 2:06 AM, whoareyou said:

@flak

I have same issue as michael123.  I'm PM ticket to you

 

The missing Phase Correction license issue was fixed on Dirac's servers, if you saved a measurement project previously, just reload it and try to re-export the filter. For me it worked.

I would also logout and re-login before.

 

 

Link to comment
2 hours ago, shinyc said:

I finally got the production version working and am finding it much more stable. However, I am experiencing image shifting to the left about 20% (if the center-to-left is100%). I recall someone complaining about this phenomenon some time ago and also recall there was a correction made in the code. Anyone else experiencing image shifting? If I bypass the Dirac filter, image returns to the center. I also noticed that degree of shift varies between various measurements.

 

PS. Image becomes vague after applying Dirac. hmmm

 

Yes, that was me

 

Link to comment
5 minutes ago, shinyc said:

I asked the same question to flak and only explanation give to me was that they affect the 1st measurement position (the critical one) only in the set of 9(chair) or 17(sofa). Therefore, make sure the 1st measurement position is exactly in the sweet spot.

 

Sweet Spot is sweet spot, we know where we sit.

 

Link to comment
19 hours ago, shinyc said:

Home Audio Fidelity uses "two microphones in your ears (a.k.a. HRTF)" way to find sweet spot. Different algorithm. But you have to pay per measurement + turnaround time. In those regards, I prefer Dirac.

 

I have installed Dirac beta previously on two systems. The one in the larger room with ~4.5 meters between the speakers is that one where the left-shift is most exposed. The good part is at least I am not the only one and there is a workaround (question: how much do we lose because of this workaround??). The bad part is that I reported to QA this issue 2 months ago and they did nothing. The issue did not exist in Version 1. 

 

And btw we have Bosch laser meter, sweet spot is exactly in the center.

But should it be??

It is supposedly to be where we usually sit.

And what that compensation is for?

 

 

Link to comment
  • 2 weeks later...
On 12/22/2019 at 3:44 AM, shinyc said:

I finally got the production version working and am finding it much more stable. However, I am experiencing image shifting to the left about 20% (if the center-to-left is100%). I recall someone complaining about this phenomenon some time ago and also recall there was a correction made in the code. Anyone else experiencing image shifting? If I bypass the Dirac filter, image returns to the center. I also noticed that degree of shift varies between various measurements.

 

PS. Image becomes vague after applying Dirac. hmmm

 

Did you try the new processor from Dec 19 (1.0.2)?

For me it looks like fixed, I used Sofa/Wide measurements

Link to comment
1 minute ago, shinyc said:

Yes, I am on 1.0.2. I still need to turn off the "Delay compensation" setting. Only then, the centering issue is gone. Do you not have that issue even if the "Delay compensation" setting is on? I still do not know what exactly the setting does, though.

 

Yes, it is ON.

Maybe that's just the points I picked up this time, .. although it shouldn't matter. And then Sofa/Wide. Other methods - chair and Sofa/Focused - sound quite bad for me

 

Anyway, so I was lucky this time.

 

Another annoying bug, which exists more or less for a year, is that sometimes during the measurements Dirac Live does not record, and I don't see the spectrum graph displayed. I do hear sweep sounds though. I then just move on to other points and afterwards go back to re-measure.

Link to comment
  • 1 month later...
14 hours ago, flak said:

If previous filters don't work one should be able to load a previous project and create new filters without having to retake measurements

 

Hi Flavio,

could you please be more precise about 'don't work' ? For me they work.. Should I skip the update then? Also, I thought that higher sampling rates were already supported before? In JRiver I have already sample rates configured up to 192/24

 

Where this Enhanced Phase Correction algorithm for improved stereo reproduction takes place?

 

How can I preserve my current filters during upgrade?

 

 

Link to comment
  • 3 months later...

I wanted to remeasure the room with the new Dirac Live (I already had V2, but from December), to my surprise I could not accomplish this 😂

 

Eventually, I filed a bug
https://jira.dirac.services/servicedesk/customer/portal/3/DLS-1511

There are few things:
1) When I go through the measurement points, there is always one, for which I can't measure. What happens is that I do hear the sweep signal, but the "EKG" on the screen is flat line, no measurements. Kind of microphone is dead.

 

This behavior is not new to me, and I already opened a bug, but there was always a workaround - when I skipped the problematic point (this is random point, not some specific place in the room), move to the next one, measure it, and move back to the one I could not before -- usually it succeeded.

 

With this new version, however, dead is dead. Measurement session was over.

2) Knowing the measurement problems from the past, I saved the project all the time. I restarted the computer, and thought I will be able to continue the measurement session in Dirac Live, but could not load the previously saved project. 

3) When I logout and login, I can't continue the session, is it by design? I see the files are stored on the disk, but Dirac Live is not using them
 

@flakFlavio, I would appreciate if you push this.. very frustrating

 

BTW, multi-channel V2 was promised by January 2020

 

Michael

 

 

Link to comment
  • 3 weeks later...
24 minutes ago, SteveS1 said:

 

Did you re-measure the a new project?  if so, I would have thought it is likely to sound different because even with the best efforts, putting the mic in precisely the same spots as a previous project would be impossible.  Have you tried re-loading an existing project and creating a second filter with the new version for a  like for like comparison?

 

I didn't re-measure as I wanted to conduct the full procedure to verify the new UI. i don't know precisely where were the fixes, was it just during the last step of filter prep or also during measurements? 

 

BUT, I want to say that that's the point of Dirac - even though the placement varies a bit, result should be more or less the same, isn't it?

 

I measure 'Wide' arrangement, I have quite a bit sofa and there are 17? points.

I remember Flavio told me once that does not really matter how exact you put the microphone, except for the sweet spot.

Anyway, my sofa has these sitting pillow so I put the mic kind of in between so it is more or less similar to what I did before.

 

Having said that, I see the new measurement produced same gain for the speakers, but bit different delay values. 

Why? Is it a fix?

 

 

Link to comment
  • 1 month later...
2 minutes ago, jcn3 said:

 

dirac is not a player -- it's a dsp engine.  to use it, you must run the calibration tool to generate the appropriate filters.  i'm still using v1, but you must select the file types -- 44.1, 88, 96, etc. -- to generate the appropriate filters.  i'm assuming you also do the same in v2.  you know what they say about assumptions . . .

 

I forgot already how it was in v1, but in v2 it is completely different. 

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...