The Official Dirac Live Thread: Start w/1,2,3 🙋‍♂️

[klinemj]

Jun 28, 2021 7:00:37 GMT -5 marcl said:

Jun 27, 2021 21:23:28 GMT -5 adam631 said:

With no more green zone in the latest Dirac Live, how loud do i need to make the levels?

Speaker efficiency varies so much that it's hard to say.  And the level you hear with the noise and tones during volume calibration is not the level you hear during actual measurements.

Here's what works in my 7.1.4 system where the side surrounds are least efficient and top rears are most efficient:

• Leave mic at 100%
• Play subwoofer tones and raise gain control until sub sweeps peak at -20db.  Subwoofer slider stays all the way up
• Pull sliders for top rears all the way down; leave sliders for side surrounds all the way up; verify levels for both are around -27db
• Using sliders for all other speakers, adjust their levels to -27db

In my system ... this results in a subwoofer measurement whose peak is at the same level as bass peaks for my large front speakers.  The level of the sweep tones during actual measurement is no louder than 75db and there are no clipping warnings or rattles/distortion from speakers.

It's wise to be conservative with levels the first time to be sure they are not too loud.  Dirac works over a wide range and it will warn you if levels are too high or low.

So - no more green zones in Dirac? Not even the faintly shaded zones they moved to after then took the green zones off? I'm hoping to run it this week or next once a consulting job is done.

Mark

[KeithL]

Nope... no more green zones... and no more blue stripes either... now it's just numbers...

Jun 28, 2021 13:45:32 GMT -5 klinemj said:

Jun 28, 2021 7:00:37 GMT -5 marcl said:

Speaker efficiency varies so much that it's hard to say.  And the level you hear with the noise and tones during volume calibration is not the level you hear during actual measurements.

Here's what works in my 7.1.4 system where the side surrounds are least efficient and top rears are most efficient:

• Leave mic at 100%
• Play subwoofer tones and raise gain control until sub sweeps peak at -20db.  Subwoofer slider stays all the way up
• Pull sliders for top rears all the way down; leave sliders for side surrounds all the way up; verify levels for both are around -27db
• Using sliders for all other speakers, adjust their levels to -27db

In my system ... this results in a subwoofer measurement whose peak is at the same level as bass peaks for my large front speakers.  The level of the sweep tones during actual measurement is no louder than 75db and there are no clipping warnings or rattles/distortion from speakers.

It's wise to be conservative with levels the first time to be sure they are not too loud.  Dirac works over a wide range and it will warn you if levels are too high or low.

So - no more green zones in Dirac? Not even the faintly shaded zones they moved to after then took the green zones off? I'm hoping to run it this week or next once a consulting job is done.

Mark

[adam631]

Jun 28, 2021 7:00:37 GMT -5 marcl said:

Jun 27, 2021 21:23:28 GMT -5 adam631 said:

With no more green zone in the latest Dirac Live, how loud do i need to make the levels?

Speaker efficiency varies so much that it's hard to say.  And the level you hear with the noise and tones during volume calibration is not the level you hear during actual measurements.

Here's what works in my 7.1.4 system where the side surrounds are least efficient and top rears are most efficient:

• Leave mic at 100%
• Play subwoofer tones and raise gain control until sub sweeps peak at -20db.  Subwoofer slider stays all the way up
• Pull sliders for top rears all the way down; leave sliders for side surrounds all the way up; verify levels for both are around -27db
• Using sliders for all other speakers, adjust their levels to -27db

In my system ... this results in a subwoofer measurement whose peak is at the same level as bass peaks for my large front speakers.  The level of the sweep tones during actual measurement is no louder than 75db and there are no clipping warnings or rattles/distortion from speakers.

It's wise to be conservative with levels the first time to be sure they are not too loud.  Dirac works over a wide range and it will warn you if levels are too high or low.

I did what you said and it worked perfectly, thank you!
I just bought custom made speakers and i uploaded my own curve and it sounds awesome

I'm in an apartment right now, so it's really cramped for space but next year i will hopefully buy a house.

[marcl]

On the topic of how to set levels in Volume Calibration ... now that there is no green zone, blue line (or even plaid) ...

What I have found is I get the best results - and most consistent - when all channels are measured at the same level. But what I also found is that setting all channels to the same db level in Volume Calibration does not result in equal levels in Measurement.  Specifically, if they are equal in Volume Calibration, my sub ends up 8-10db lower in Measurement.

My room has a BIG resonance at 40Hz and my subs and my front L/R play in this range. My goal in setting levels is to get that peak to measure the same for the subs and fronts. Here's a screen shot of how they should measure:

In order to achieve this I found (through multiple experiments) that the subwoofer level in Volume Calibration had to be 8-10db higher than all the other speakers.  So, in Volume Calibration I start with the subwoofer and - leaving its slider at the top 0.0db - I increase the Master Output until the sub tones are at a comfortable level.  Then I set all the other speakers 8-10db lower than that level.  What this does is three things:

• Ensures that none of the measurements will be too loud and clip (since the sub peak will always be the highest peak in the room)
• Ensures that the sub is measured at the same level as all the other speakers
• Ensures that the sub measurement does not pull down the levels of all the other speakers when Dirac sets the levels; and that the sub target curve does not end up with excessive boost

Your system may behave differently and maybe you don't see this disparity between the subs and other channels and maybe the difference is not 8-10db.  But if there is a difference, this is a good way to fix it.

Note also that the levels are also dependent on the efficiency of other speakers.  In my system my side surrounds are least efficient and the rear tops are most efficient.  Their sliders have to be all the way up at 0.0db and all the way down at -12db respectively, and the rest of the speakers somewhere in between.  I end up with the sub level around -18db and the rest of the speakers around -28 (within a db or two).

p.s.  I probably have not shared this in earlier posts on this topic, but I noticed recently that others are seeing the same issue and compensating in similar ways (i.e. AVS Dirac forum).

[ttocs]

marcl very nice and detailed explanation, thanks!!

What you point out is probably why some folks have difficulty with the clipping errors when trying Dirac for the first time and wondering why they are seeing that error.

[marcl]

Jul 3, 2021 5:54:52 GMT -5 ttocs said:

marcl very nice and detailed explanation, thanks!!

What you point out is probably why some folks have difficulty with the clipping errors when trying Dirac for the first time and wondering why they are seeing that error.

I hesitated because I wouldn't want anyone to take the numbers literally, and speaker efficiencies vary so much.  But at least this process ensures that everything measures lower than the biggest peak.

The bit about getting the subs to measure the same as other speakers ... it's puzzling why we need to figure this out.  I can't imagine why Dirac would measure the subs so much lower on purpose.  And the result of all this is it addresses what has been called "headroom" by some of our friends over the past year.  You do a Dirac calibration and all of a sudden you have to turn the volume knob up 5-10db higher on the processor for the same listening levels.  This process has resulted in much more consistent levels for me.

[p4t]

I have a question. If I set my XMC2 setting, center sub out set as LFE and going to dedicated subwoofer, then my left subwoofer out set as mono and going to another dedicated subwoofer (different subwoofer from LFE). My question is, during the movie, does the signal going to LFE is the same signal going to BM? Or is it different signal? If it is different what is the lowest Hz BM freq produce? Thanks.

[geebo]

Jul 7, 2021 8:38:33 GMT -5 p4t said:

I have a question. If I my XMC2 setting, center sub out set as LFE and going to dedicated subwoofer, then my left subwoofer out set as mono and going to another dedicated subwoofer (different subwoofer from LFE). My question is, during the movie, does the signal going to LFE is the same signal going to BM? Or is it different signal? If it is different what is the lowest Hz BM freq produce? Thanks.

The LFE is a separate channel completely. It's the .1 in 5.1 or 7.1. Normally it is mixed with the BM but with the center channel set for LFE it gets the LFE and only the LFE. Your other sub will get redirected bass from the other channels and nothing from the LFE channel. So they are different signals. I don't know if there is a lower limit on the BM signal but I do know it goes to at least 16Hz and probably lower.

[ttocs]

Jul 7, 2021 8:38:33 GMT -5 p4t said:

I have a question. If I my XMC2 setting, center sub out set as LFE and going to dedicated subwoofer, then my left subwoofer out set as mono and going to another dedicated subwoofer (different subwoofer from LFE). My question is, during the movie, does the signal going to LFE is the same signal going to BM? Or is it different signal? If it is different what is the lowest Hz BM freq produce? Thanks.

Bass for all speaker channels per Dolby Spec is 40Hz (I looked it up last night), but in reality can go down to 28Hz, and maybe lower, I've measured as low 28Hz-ish. The speaker channels setup as Small will use a subwoofer channel setup as Mono, so just think of the subwoofer channel and speaker channel as one speaker with very capable bass producing woofer.

Bass for the .1 LFE channel per Dolby is from 31.5Hz to 120Hz, but in reality, there are lots of movies with much lower frequencies down to 20Hz or so, some lower but I've not measured any of those. I've only measured 22Hz in LFE from movies which could just be a limitation of my subs or the movies I measured.

The .1 LFE channel has no relationship to Bass Management using crossover settings for Small channels, it is totally separate and unique.

Here is some info from Dolby. They talk using language such as Screen Speakers and Screen Subwoofer. The Screen Subwoofer is the .1 LFE channel. Screen Speakers represent any channel other than the .1 channel, and "can" use subwoofers and bass management with crossover settings or be full range large speakers.

The really simple way to think about it is this:
There are Speaker Channels which need to produce "FrequencyRange:40Hzto16kHz,+3/–6dB"
And there is a subwoofer channel, "Screen Subwoofer: Sound Pressure Level: +10 dB (Compared to Center Loudspeaker)" and "Frequency Response: 31.5–120 Hz, ±3 dB".

What we use for Speakers is up to us. If we have speakers with small woofers, then we can use a subwoofer to reach the low frequencies, but this won't mean that LFE will be routed to these subwoofers, that will only happen if the Center Sub is not setup as LFE.

Here is the fun part. We have the ability to assign to the Center Sub Output: LFE. This takes LFE from being combined with the rest of Bass Management and putting it where it belongs, in a separate channel. This leaves us with the option of having all Large speakers, or using subwoofers for the bass for speakers setup as Small.

edit: You can read the Dolby info HERE.

[p4t]

Jul 7, 2021 9:14:54 GMT -5 ttocs said:

Jul 7, 2021 8:38:33 GMT -5 p4t said:

I have a question. If I my XMC2 setting, center sub out set as LFE and going to dedicated subwoofer, then my left subwoofer out set as mono and going to another dedicated subwoofer (different subwoofer from LFE). My question is, during the movie, does the signal going to LFE is the same signal going to BM? Or is it different signal? If it is different what is the lowest Hz BM freq produce? Thanks.

Bass for all speaker channels per Dolby Spec is 40Hz (I looked it up last night), but in reality can go down to 28Hz, and maybe lower, I've measured as low 28Hz-ish. The speaker channels setup as Small will use a subwoofer channel setup as Mono, so just think of the subwoofer channel and speaker channel as one speaker with very capable bass producing woofer.

Bass for the .1 LFE channel per Dolby is from 31.5Hz to 120Hz, but in reality, there are lots of movies with much lower frequencies down to 20Hz or so, some lower but I've not measured any of those. I've only measured 22Hz in LFE from movies which could just be a limitation of my subs.

The .1 LFE channel has no relationship to Bass Management using crossover settings for Small channels, it is totally separate and unique.

Here is some info from Dolby. They talk using language such as Screen Speakers and Screen Subwoofer. The Screen Subwoofer is the .1 LFE channel. Screen Speakers represent any channel other than the .1 channel, and "can" use subwoofers and bass management with crossover settings or be full range large speakers.

The really simple way to think about it is this:
There are Speaker Channels which need to produce "FrequencyRange:40Hzto16kHz,+3/–6dB"
And there is a subwoofer channel, "Screen Subwoofer: Sound Pressure Level: +10 dB (Compared to Center Loudspeaker)" and "Frequency Response: 31.5–120 Hz, ±3 dB".

What we use for Speakers is up to us. If we have speakers with small woofers, then we can use a subwoofer to reach the low frequencies, but this won't mean that LFE will be routed to these subwoofers, that will only happen if the Center Sub is not setup as LFE.

Here is the fun part. We have the ability to assign to the Center Sub Output: LFE. This takes LFE from being combined with the rest of Bass Management and putting it where it belongs, in a separate channel. This leaves us with the option of having all Large speakers, or using subwoofers for the bass for speakers setup as Small.

edit: You can read the Dolby info HERE.

The purpose of my question do I need to sync (time align,etc) between dedicated BM subwoofers and dedicated LFE subwoofers. Since the signal is totally different than I do not need to sync between BM subwoofers and LFE subwoofers. Cause if the signal the same and you are not sync them, it will messed up the frequency response. So, if I set front speakers large (have their own dedicated sub) and Center speaker large (also have own dedicated sub), dedicated sub for LFE and dedicated sub for BM, it should be ok?

[ttocs]

Jul 7, 2021 10:46:42 GMT -5 p4t said:

The purpose of my question do I need to sync (time align,etc) between dedicated BM subwoofers and dedicated LFE subwoofers. Since the signal is totally different than I do not need to sync between BM subwoofers and LFE subwoofers. Cause if the signal the same and you are not sync them, it will messed up the frequency response.

So, if I set front speakers large (have their own dedicated sub) and Center speaker large (also have own dedicated sub), dedicated sub for LFE and dedicated sub for BM, it should be ok?

The BM channels do not need to sync with the LFE channel. They are two independent signal chains. The subs you use for BM only need to align with the speakers setup as Small.

Yes, with the caveat being that you need to check your LCR speakers to see if it's ok for them to receive low frequencies in the 20Hz range without damage. Martin Logan told my specifically that my Motif center channel speaker does not have any protection built-in for really low frequencies. So your might have a similar limitation and would need a filter to squelch the lowest frequencies. I'm not certain about your ESL's, but they may be ok.

[p4t]

Jul 7, 2021 10:56:20 GMT -5 ttocs said:

Jul 7, 2021 10:46:42 GMT -5 p4t said:

The purpose of my question do I need to sync (time align,etc) between dedicated BM subwoofers and dedicated LFE subwoofers. Since the signal is totally different than I do not need to sync between BM subwoofers and LFE subwoofers. Cause if the signal the same and you are not sync them, it will messed up the frequency response.

So, if I set front speakers large (have their own dedicated sub) and Center speaker large (also have own dedicated sub), dedicated sub for LFE and dedicated sub for BM, it should be ok?

The BM channels do not need to sync with the LFE channel. They are two independent signal chains. The subs you use for BM only need to align with the speakers setup as Small.

Yes, with the caveat being that you need to check your LCR speakers to see if it's ok for them to receive low frequencies in the 20Hz range without damage. Martin Logan told my specifically that my Motif center channel speaker does not have any protection built-in for really low frequencies. So your might have a similar limitation and would need a filter to squelch the lowest frequencies. I'm not certain about your ESL's, but they may be ok.

The subs for BM I just time align so all subs work as one dirac live will align them with all speakers set as small. For Martin logan ESL front I think should be ok without using any low cut off crossover. As for Martin Logan ESL-C center I will use my minidsp to align and set the crossover to sync with dedicated center channel subwoofer. Hmmm... I think I will experiment this by setting LCR as large and dedicated subs for BM and dedicated subs for LFE separately.

[ttocs]

Jul 7, 2021 11:15:50 GMT -5 p4t said:

The subs for BM I just time align so all subs work as one dirac live will align them with all speakers set as small. For Martin logan ESL front I think should be ok without using any low cut off crossover. As for Martin Logan ESL-C center I will use my minidsp to align and set the crossover to sync with dedicated center channel subwoofer. Hmmm... I think I will experiment this by setting LCR as large and dedicated subs for BM and dedicated subs for LFE separately.

Dirac will time align all channels it sees in the processor.

When the miniDSP outputs to multiple speakers or subs which are fed from a single channel, then it's your job to align all those speakers/subs because Dirac will only see the one channel connected to the input of the miniDSP.

Dirac will see the XMC-2/Center-Sub-LFE-Output, but the miniDSP will feed multiple subwoofers.
Dirac will see the XMC-2/Left-Sub-Mono-Output, but the miniDSP will feed multiple subwoofers.
In each case, Dirac will only see and correct the output, not the multiple devices that the output feeds.

Another example would be my Front Speakers, I'll just use the Left one. Dirac only sees the Front Left output, but my Front Left is actually a speaker plus two subwoofers.

[marcl]

audiophilestyle.libsyn.com/dirac-research-interview

"Audiophile Style talks to Dr. Mathias Johansson, Dirac Research co-founder and chief product officer, about the beginnings of the company in Upsala Sweden, its work with BMW, the current Dirac offerings for audiophiles, the differences between Dirac software and Dirac built into hardware, and many general DSP & room correction topics."   

It's 1:15 long, but worth listening to especially if you have never heard Mathias speak.  A few interesting points I had not heard before ...

• IIR filters handle minimum phase correction and FIR filters handle non-minimum phase correction. (Note: PEQ filters in AV processors, miniDSP, etc are generally only IIR filters)
  
• Dirac used frequency dependent windowing in early versions, but they now have a "more robust" (proprietary) solution.
• One good reason to use Dirac full spectrum and not just on the bass, is that Dirac uses time delay (phase) correction to align frequency bands.  This alignment improves impulse response across the whole frequency band.
• Some customers - knowing their personal hearing loss characteristics - tailor their target curve to compensate.
• Some of the best imaging and overall response Mathias has heard using Dirac has been with electrostats and a subwoofer.

Finally ... based on listening to Mathias say "Dirac" dozens of times .... it's "Di-RAC" ... not "DEE-rac" or "DIE-rac" ... just sayin'  

[oleops]

Jul 7, 2021 8:38:33 GMT -5 p4t said:

I have a question. If I set my XMC2 setting, center sub out set as LFE and going to dedicated subwoofer, then my left subwoofer out set as mono and going to another dedicated subwoofer (different subwoofer from LFE). My question is, during the movie, does the signal going to LFE is the same signal going to BM? Or is it different signal? If it is different what is the lowest Hz BM freq produce? Thanks.

The Spec of the LFE is limited to 120Hz up

But there have been some miss mixes, so some encoders cutof lower than 120 and often it can be adjusted at your liking ... the discussion on the «best» limitation goes on forever and you must trust your own judgement or leave it up to the mixers.

I have not found any lower limit to the LFE from reliable source, but at forums it has been mentioned 3Hz

Whats the limit on BM I dunno, but think that also must be same as LFE


EDIT:  I see the discussion took another path, I missed that..
But I found the Dolby limitation in the Dolby-5.1-Channel-Production-Guidelines for those interested..
All channels actually are full range from 3Hz-20KHz and LFE 3Hz-120Hz

«5.1-channel audio typically consists of five discrete, full range main channels (Left,
Center, Right, Left Surround, and Right Surround) plus an optional band-limited Low
Frequency Effects (LFE) channel for added bass (the .1). Dolby Digital bitstreams
deliver full frequency bandwidth main channels, from 3 Hz to 20 kHz, and a limited
frequency bandwidth LFE channel, from 3 Hz to 120 Hz»

What ttocs linked to was minimum response of the speaker setup not the systems limitation
If there have been some changes to the system limitation with Atmos and other I dunno, but doubt it...

[oleops]

Jul 10, 2021 14:04:22 GMT -5 marcl said:

Finally ... based on listening to Mathias say "Dirac" dozens of times .... it's "Di-RAC" ... not "DEE-rac" or "DIE-rac" ... just sayin'  

As the company was named in honor of Paul Dirac it should be "duh·rak" then ...     and Yes he is British so English pronouncment 
Those Swedes..  

pronouncment Paul Dirac

[marcl]

Jul 10, 2021 16:37:10 GMT -5 oleops said:

Jul 10, 2021 14:04:22 GMT -5 marcl said:

Finally ... based on listening to Mathias say "Dirac" dozens of times .... it's "Di-RAC" ... not "DEE-rac" or "DIE-rac" ... just sayin'  

As the company was named in honor of Paul Dirac it should be "duh·rak" then ...     and Yes he is British so English pronouncment 
Those Swedes..  

pronouncment Paul Dirac

I was actually poking at our southern US friends who can't help the DIE-rac and DEE-rac 

[LCSeminole]

Jul 10, 2021 16:50:33 GMT -5 marcl said:

Jul 10, 2021 16:37:10 GMT -5 oleops said:

As the company was named in honor of Paul Dirac it should be "duh·rak" then ...     and Yes he is British so English pronouncment 
Those Swedes..  

pronouncment Paul Dirac

I was actually poking at our southern US friends who can't help the DIE-rac and DEE-rac 

While I'm one of those southerners you reference(really a transplant from the north), here in Tallahassee, FL, we had the honor of having Paul M. Dirac for the last 10 years of his life as a professor of Quantum and Theoretical Physics at Florida State University. Even the FSU Science Library is named after him, which leads me to the purpose of this post, the librarians when I was at FSU would correct those students of the pronunciation of his name, which the southern drawl does not do his last name justice.

[ttocs]

[leonski]

Jul 10, 2021 14:04:22 GMT -5 marcl said:

audiophilestyle.libsyn.com/dirac-research-interview

"Audiophile Style talks to Dr. Mathias Johansson, Dirac Research co-founder and chief product officer, about the beginnings of the company in Upsala Sweden, its work with BMW, the current Dirac offerings for audiophiles, the differences between Dirac software and Dirac built into hardware, and many general DSP & room correction topics."   

It's 1:15 long, but worth listening to especially if you have never heard Mathias speak.  A few interesting points I had not heard before ...

• IIR filters handle minimum phase correction and FIR filters handle non-minimum phase correction. (Note: PEQ filters in AV processors, miniDSP, etc are generally only IIR filters)
  
• Dirac used frequency dependent windowing in early versions, but they now have a "more robust" (proprietary) solution.
• One good reason to use Dirac full spectrum and not just on the bass, is that Dirac uses time delay (phase) correction to align frequency bands.  This alignment improves impulse response across the whole frequency band.
• Some customers - knowing their personal hearing loss characteristics - tailor their target curve to compensate.
• Some of the best imaging and overall response Mathias has heard using Dirac has been with electrostats and a subwoofer.

Finally ... based on listening to Mathias say "Dirac" dozens of times .... it's "Di-RAC" ... not "DEE-rac" or "DIE-rac" ... just sayin'  

I've been a proponent of FIR filters for years.      You actually have to READ the stuff at the MiniDSP Website (GASP!) to find those products with the computing power for FIR filters.   Like the 2x4HD.....to start.
Than you have to figure out HOW to use 'em......a real learning curve.

What a concept!   Time Allign.    Never saw THAT one coming.

maybe 'stats (and by extension Maggies?) sound good is that the SINGLE diaphrapgm is a single planar 'launch' for a wave.    No driver setback to create time align issue.

Does anybody CARE how it is pronounced?