New second generation Dirac Live

[wxq0223]

hi, any mac-version blu-ray player or vedio player can work with Dirac live 2.0 for multichannel playback?

[Racerxnet]

I am also wondering if this software will work with foobar2k on a Pc. Please provide info regarding this and what dsp plug-in to use. Can the filters be exported to the plug-in? My Genesis system has a stand alone amp with phase, low pass,  high pass, and volume control. How will it integrate with the bass amp and its adjustments?  There are Bryston 7BSST,s on the mid/high panels. 

 

Mak

[milva]

Hi Flavio,

 

test version Dirac 2.0 has ended. Is it going to be prolonged? I am not able to listen to the Dirac 1.0 anymore. When is the release date of the official Dirac 2.0 version?

[blue2]

On ‎9‎/‎10‎/‎2019 at 10:24 PM, Racerxnet said:

I am also wondering if this software will work with foobar2k on a Pc. Please provide info regarding this and what dsp plug-in to use. Can the filters be exported to the plug-in? My Genesis system has a stand alone amp with phase, low pass,  high pass, and volume control. How will it integrate with the bass amp and its adjustments?  There are Bryston 7BSST,s on the mid/high panels. 

 

Mak

Dirac support pointed me to this guide using the attached version of foo_vst.dll

foo_vst.dll

[flak]

9 hours ago, milva said:

Hi Flavio,

 

test version Dirac 2.0 has ended. Is it going to be prolonged? I am not able to listen to the Dirac 1.0 anymore. When is the release date of the official Dirac 2.0 version?

 

Sure, please contact us from here: http://helpdesk.dirac.se/

as usual, thanks

[KnockKnock]

Why is Dirac 1.0 not working anymorr? I am afraid of updating macosx bcs exatly that. I tried the Dirac Beta but it didn’t work at all. And due to very bad acoustics I NEED Dirac!!!!

[Racerxnet]

2 hours ago, flak said:

 

Sure, please contact us from here: http://helpdesk.dirac.se/

as usual, thanks

 

You did not answer the question on the release of Dirac2. Should we all contact the help desk?

[whoareyou]

Have people been receiving replies from the helpdesk?

[michelangelo]

Yes, usually within a few hours.

[whoareyou]

7 hours ago, michelangelo said:

Yes, usually within a few hours.

Ok.  I used to receive reply quickly, but now my Dirac helpdesk questions/comments are without response.  There seems to be some type of disconnect with the help system for me.

 

[Racerxnet]

1 hour ago, whoareyou said:

Ok.  I used to receive reply quickly, but now my Dirac helpdesk questions/comments are without response.  There seems to be some type of disconnect with the help system for me.

 

Same here. No response!

 

MAK

[whoareyou]

Perhaps they are very busy with 2.0 going live for some of their customers.  

It would just be nice to get an update to understand what's going on with my beta testing (which is expiring today).

[milva]

The help desk response time to my questions was very fast (120 minutes - at friday afternoon).

[whoareyou]

21 minutes ago, milva said:

The help desk response time to my questions was very fast (120 minutes - at friday afternoon).

Received response today.   Guess stuff happens.

[sound_cycle]

On 9/12/2019 at 11:29 PM, flak said:

 

Sure, please contact us from here: http://helpdesk.dirac.se/

as usual, thanks

Did that.

 

Switched back to v1, but there is no going back, really. Staging and bass esp. are much better in v2.

GR

 

 

 

[Racerxnet]

5 minutes ago, sound_cycle said:

Did that.

 

Switched back to v1, but there is no going back, really. Staging and bass esp. are much better in v2.

GR

 

 

 

Dirac did reply back to me and I appreciate that. GR, what are you using for playback with Dirac?

 

MAK

[sound_cycle]

22 hours ago, Racerxnet said:

Dirac did reply back to me and I appreciate that. GR, what are you using for playback with Dirac?

 

MAK

Roon, Schiit Gumby+saga*+Vidar/monobloc into Magnepan 1.7i, REL t9/i

(*I know, there was a time when I said xlr ? me?)

Dirac have written back to me. Hope to have it running again this evening

[DSD256]

On 9/15/2019 at 3:01 PM, sound_cycle said:

Did that.

 

Switched back to v1, but there is no going back, really. Staging and bass esp. are much better in v2.

GR

 

 

 

Dito...find 2.x much more realistic regarding low end rendering. With 1.x I have had hard times to fix that "too much bass" in the filters.

With 1.x filter I usually went -1 --2db in the low end.

2.x is way more realistic regarding low end rendering.

 

Besides: I found something interesting:

 

Compared to mathaudio RoomEQ I always found that music in Audirvana/Mac always distorted a little or very much by Dirac 2.x.
Interestingly, the audio unit beta of 1.x does the same distortion, although not as big..
If a piece was distorted I always loaded mathaudio RoomEQ into Audirvana and noticed that nothing was distorted here.

 

To deepen the story I opened Reaper (a DAW) and loaded a contemporary compressed track,  into the master I've inserted Dirac Live, Dirac processor and mathaudio RoomEQ as vst or audio-unit. I always play a loop of a region.

 

All three plugins are bypassed at the beginning and therefore Reaper will level out a maximum of 0db at the loop.

mathaudio RoomEQ stays at 0db. Here nothing goes into the range above 0db and therefore there are no distortions to be heard.

 

With Dirac 1.x it distorts, Reaper levels out +3.4 db at the beginning of the region. Something is wrong. That's probably the reason why Audirvana always has distortions.

Also the dirac 2.x with processor 0.94 causes distortions in the loop of the region. Reaper even shows me +4.8 db with a target filter. No wonder that in Audirvana it is distorted like crazy.

Now I'm in Dirac live 2.x again and created a new filter of a project, but I changed the filter in its entirety over -5db down in volume.

When I load this filter into processor 0.94 in the Reaper Project, Reaper now shows me no distortion at all. Reaper levels a maximum of 0db left and right.
This filter is now loaded into Audirvana and now the distortions within Audirvana with dirac2.x processor 0.94 are a thing of the past.

No distortions anymore.

Depending on the music material it seems that obviously the vst/au plugins of 1.x and the processor 0.94 of Dirac 2.x are distorting the audio between +0db to +5db when using a typical target curve for a filter.

 

The measurements were made in Dirac 1.x with the RME Babyface pro and the Audient id14.  The measurement microphones Beyerdynamic MM1 and Umik-1 were used for the measurements.
In Dirac 2.x the measurements were made with the Beyerdynamic MM1/ Umik-1 and the Audient id14.

 

Since replay-gain was almost always activated in Audirvana, noticeable distortion was much less common.
Since I have deactivated replay-gain in the last weeks, since then the distortions became an issue.

 

Finally I suspect that in the processor of diraclive 2.x as well in the vst/au beta of Dirac 1.x
a technique like limiting or normalizing to prevent distortion is missing.

Vid that shows my observations

 

Have similar observations already been made?

 

[macuniverse]

9 hours ago, DSD256 said:

Have similar observations already been made?

 

Yes, I have played the same track with Audirvana and JRiver. JRiver has the option to monitor clipping. It shows some clipping with DL2 VST engaged.

 

My trial&error solution is to follow the speakers' natural room response as much as possible and use mostly cut filters.

[Racerxnet]

I tried Dirac Live with Foobar2k and could only connect with a virtual cable. The latency was so bad it stuttered. Without any plug-in to load this into software, it is not going to work. The calibration files are unable to load in Foobar. Until it is more user friendly, I'll watch from the sideline. Right now MathAudio has worked well.

 

Mark

[flak]

On 9/20/2019 at 6:36 PM, DSD256 said:

Dito...find 2.x much more realistic regarding low end rendering. With 1.x I have had hard times to fix that "too much bass" in the filters.

With 1.x filter I usually went -1 --2db in the low end.

2.x is way more realistic regarding low end rendering.

 

Besides: I found something interesting:

 

Compared to mathaudio RoomEQ I always found that music in Audirvana/Mac always distorted a little or very much by Dirac 2.x.
Interestingly, the audio unit beta of 1.x does the same distortion, although not as big..
If a piece was distorted I always loaded mathaudio RoomEQ into Audirvana and noticed that nothing was distorted here.

 

To deepen the story I opened Reaper (a DAW) and loaded a contemporary compressed track,  into the master I've inserted Dirac Live, Dirac processor and mathaudio RoomEQ as vst or audio-unit. I always play a loop of a region.

 

All three plugins are bypassed at the beginning and therefore Reaper will level out a maximum of 0db at the loop.

mathaudio RoomEQ stays at 0db. Here nothing goes into the range above 0db and therefore there are no distortions to be heard.

 

With Dirac 1.x it distorts, Reaper levels out +3.4 db at the beginning of the region. Something is wrong. That's probably the reason why Audirvana always has distortions.

Also the dirac 2.x with processor 0.94 causes distortions in the loop of the region. Reaper even shows me +4.8 db with a target filter. No wonder that in Audirvana it is distorted like crazy.

Now I'm in Dirac live 2.x again and created a new filter of a project, but I changed the filter in its entirety over -5db down in volume.

When I load this filter into processor 0.94 in the Reaper Project, Reaper now shows me no distortion at all. Reaper levels a maximum of 0db left and right.
This filter is now loaded into Audirvana and now the distortions within Audirvana with dirac2.x processor 0.94 are a thing of the past.

No distortions anymore.

Depending on the music material it seems that obviously the vst/au plugins of 1.x and the processor 0.94 of Dirac 2.x are distorting the audio between +0db to +5db when using a typical target curve for a filter.

 

The measurements were made in Dirac 1.x with the RME Babyface pro and the Audient id14.  The measurement microphones Beyerdynamic MM1 and Umik-1 were used for the measurements.
In Dirac 2.x the measurements were made with the Beyerdynamic MM1/ Umik-1 and the Audient id14.

 

Since replay-gain was almost always activated in Audirvana, noticeable distortion was much less common.
Since I have deactivated replay-gain in the last weeks, since then the distortions became an issue.

 

Finally I suspect that in the processor of diraclive 2.x as well in the vst/au beta of Dirac 1.x
a technique like limiting or normalizing to prevent distortion is missing.

Vid that shows my observations

 

Have similar observations already been made?

 

 

Hi DSD256,

 

thanks for reporting about the above,

we could reproduce here so it has been added as a bug to be fixed.

 

Flavio

 

[flak]

On 9/21/2019 at 4:23 AM, macuniverse said:

 

Yes, I have played the same track with Audirvana and JRiver. JRiver has the option to monitor clipping. It shows some clipping with DL2 VST engaged.

 

My trial&error solution is to follow the speakers' natural room response as much as possible and use mostly cut filters.

 

Thanks macuniverse,

 

we are looking into this as well..

[flak]

A beta tester sent the following valuable input:

Level is not automatically reduced by -8db as for Dirac Live 1.x, but when reducing the level by -8db in the VST plugin in JRiver (right hand site level bar in the DiracLiveProcessor) digital clipping should be avoided.

Also the above setting is not saved so that it has to be entered again.

 

I expect that those bugs will be fixed in the new version that we hope to make available within the expiry date of the current DLP builds

[whoareyou]

2 minutes ago, flak said:

"...

Level is not automatically reduced by -8db as for Dirac Live 1.x, but when reducing the level by -8db in the VST plugin in JRiver (right hand site level bar in the DiracLiveProcessor) digital clipping should be avoided.

Also the above setting is not saved so that it has to be entered again.

 

I expect that those bugs will be fixed in the new version that we hope to make available within the expiry date of the current DLP builds"

 

I'm not sure just reducing the level is the solution. 

Multichannel rarely clips to the degree, if at all,  that 2 channel is clipping.

 

Also, reducing the level by -8db will fix the 2 channel problem, but lowers the multichannel when it need not be.

 

Even before lowering Jriver control (to avoid clipping) I generally have to raise my amp's volume by 8 to 10db,  to achieve same listening level as with filter disengaged.  Now if I lower VST volume control by an additional 8DB, I'd have to raise level of my amp by a total of 16 - 20 db to achieve the same listening levels with Dirac engaged so as to avoid clipping.  That just doesn't seem right to me.

 

"Also the above setting is not saved so that it has to be entered again."

With JRiver, the volume *is* persisted but the UI slider bar does not properly reflect the setting (at least for me).  True, the display always reverts to full volume but clicking on the control demonstrates this is clearly not the case. 

 

For example if you click on the control when it indicates maximum volume, nothing should happen.  Control indicates we are at maximum volume.   For me volume jumps to much higher level.  This caught me very much by surprise initially.

 

 

 

 

 

 

[macuniverse]

I think this is, sort of, related to the overall volume/clipping issue discussed above posts. I need some guidance here with the shaping of the target curve.

 

I'm starting with Dirac default target curve. Before customising the target curve I wish to reduce the overall dB. Rather than manually doing it I was expecting some sort of vertical slider (similar to VST filter overall volume slider) available for this purpose. I have also played with some mouse click+keyboard mod keys and drag with no joy.

 

Is this possible?