Emotiva XMC-1 7.2, Dirac Live and bass management

[flak]

I'm writing this post because my statement that Dirac Live does not handle the bass management has sometimes being incorrectly interpreted as a limitation of the XMC-1 7.2 & Dirac Live solution.

Bass management is the responsibility of the manufacturer (Emotiva in our case) who has provided the ability to choose the crossover frequencies at which signals should be passed to the subwoofers.

Now what (reasonably) worries some forumers is the reasoning that individually correcting two subwoofers in different positions for an ideal frequency response could nonetheless generate an incorrect combined response... this has been outlined for example by URWI who said the following:

"Just because each response is more or less flat doesn't ensure that the combined response is also flat (this is caused by different phase responses)"
From this thread: emotivalounge.proboards.com/thread/37908/xmc-bass-management?page=9&scrollTo=664272

What URWI says is definitely correct as long as we use conventional parametric minimum-phase corrections, so if we use REW and EQ we will have to play with trims to try to get the best possible frequency response... but please do not focus on frequency response alone and do not forget that Dirac Live is different as it corrects the impulse response also.

If we go back to what URWI said we see that he explained that the issue is caused by different phase responses... and that is also correct, but it is where Dirac Live comes in!
Since Dirac Live corrects for the impulse response (i.e. changes the phase response) for each of the individual speakers, the summing/superposition will hold true, as far as physically possible, in the measurement area.

So from my admittedly biased point of view I expect that the Dirac Live solution will provide the best frequency AND impulse responses since it improves the impulse response of the speakers while minphase parametric EQs are not designed to do it... and as a side benefit it will be much easier to use.

But you are the final judges so I wait for your comments when Dirac will be in your hands
Ciao, Flavio

[hifiaudio2]

Thanks Flavio.

Two things... one.. did I miss an announcement of the release?

Two... so what you are saying is that I am likely to get a great result by leaving my 4 subs (which have been time aligned to each other using REW and a Xilica DSP) connected to one output of the XMC and let Dirac just hear them all at once and set what it thinks is appropriate?

[ansat]

Nov 10, 2014 9:35:01 GMT -5 flak said:

I'm writing this post because my statement that Dirac Live does not handle the bass management has sometimes being incorrectly interpreted as a limitation of the XMC-1 7.2 & Dirac Live solution.

Bass management is the responsibility of the manufacturer (Emotiva in our case) who has provided the ability to choose the crossover frequencies at which signals should be passed to the subwoofers.

Now what (reasonably) worries some forumers is the reasoning that individually correcting two subwoofers in different positions for an ideal frequency response could nonetheless generate an incorrect combined response... this has been outlined for example by URWI who said the following:

"Just because each response is more or less flat doesn't ensure that the combined response is also flat (this is caused by different phase responses)"
From this thread: emotivalounge.proboards.com/thread/37908/xmc-bass-management?page=9&scrollTo=664272

What URWI says is definitely correct as long as we use conventional parametric minimum-phase corrections, so if we use REW and EQ we will have to play with trims to try to get the best possible frequency response... but please do not focus on frequency response alone and do not forget that Dirac Live is different as it corrects the impulse response also.

If we go back to what URWI said we see that he explained that the issue is caused by different phase responses... and that is also correct, but it is where Dirac Live comes in!
Since Dirac Live corrects for the impulse response (i.e. changes the phase response) for each of the individual speakers, the summing/superposition will hold true, as far as physically possible, in the measurement area.

So from my admittedly biased point of view I expect that the Dirac Live solution will provide the best frequency AND impulse responses since it improves the impulse response of the speakers while minphase parametric EQs are not designed to do it... and as a side benefit it will be much easier to use.

But you are the final judges so I wait for your comments when Dirac will be in your hands
Ciao, Flavio

Thank you for that information Flavio. There is 2 additional concerns that I would like you to address if possible.

1. Phase alignment (continued) - At a crossover point. Can Dirac address this at all? without knowing the crossover point, I do not see a way in which to give priority to that point to help with giving a generally flat response.

2. Professional calibration - What is Dirac's recommended steps for calibrators out there? I would like to use Dirac as one of the tools in the calibration of the room and in its integration into past systems, it has been treated as such. Is the XMC different? or did that level of control in other systems come from their developers?


Tony

[SticknStones]

Flavio,

No ask from me but a sincere thank you for participating on our forum from time to time as it really helps our personal investment confidence in our audio / video gear. When this is released I am sure we will let the software do all the talking! We are jazzed and oh so slightly anxious with this anticipated release!

In God we trust and everyone else is on a case by case basis!

[urwi]

The response of two subwoofers would need to be minimum phase at multiple points within the listening area in order to sum correctly but room responses aren't completely minimum phase. So even if Dirac Live improves non-minimum phase frequency regions for each single subwoofer, their combined frequency response is still not guaranteed to be smooth.

Nov 10, 2014 9:35:01 GMT -5 flak said:

I'm writing this post because my statement that Dirac Live does not handle the bass management has sometimes being incorrectly interpreted as a limitation of the XMC-1 7.2 & Dirac Live solution.

Bass management is the responsibility of the manufacturer (Emotiva in our case) who has provided the ability to choose the crossover frequencies at which signals should be passed to the subwoofers.

Now what (reasonably) worries some forumers is the reasoning that individually correcting two subwoofers in different positions for an ideal frequency response could nonetheless generate an incorrect combined response... this has been outlined for example by URWI who said the following:

"Just because each response is more or less flat doesn't ensure that the combined response is also flat (this is caused by different phase responses)"
From this thread: emotivalounge.proboards.com/thread/37908/xmc-bass-management?page=9&scrollTo=664272

What URWI says is definitely correct as long as we use conventional parametric minimum-phase corrections, so if we use REW and EQ we will have to play with trims to try to get the best possible frequency response... but please do not focus on frequency response alone and do not forget that Dirac Live is different as it corrects the impulse response also.

If we go back to what URWI said we see that he explained that the issue is caused by different phase responses... and that is also correct, but it is where Dirac Live comes in!
Since Dirac Live corrects for the impulse response (i.e. changes the phase response) for each of the individual speakers, the summing/superposition will hold true, as far as physically possible, in the measurement area.

So from my admittedly biased point of view I expect that the Dirac Live solution will provide the best frequency AND impulse responses since it improves the impulse response of the speakers while minphase parametric EQs are not designed to do it... and as a side benefit it will be much easier to use.

But you are the final judges so I wait for your comments when Dirac will be in your hands
Ciao, Flavio

[flak]

@ hifiaudio2

no, you did not miss any announcement
and no, I mentioned two subwoofers in different positions... there is an inherent problem in trying to compensate four subwoofers playing the same signal, but being in different physical locations.
This will inhibit Dirac Live's ability to do impulse correction (due to it trying to compensate within an area), it just is not pysically possible.

If you want to use all of your subs may be you can place them physically grouped 2 by 2, each group connected to one sub output of the XMC-1 so that Dirac Live will have the possibility of doing the impulse correction... or alternatively you could use just 2 subs.

@ Tony (ansat)

I don't know whether the crossovers are active during Dirac Live measurements, in which case Dirac Live would be able to compensate for them as well... Emotiva can answer that.

@ urwi

Room response correction is really what Dirac Live is about, so yes... the combined response will be as smooth as physically possible within the measurement area.

Flavio

[urwi]

"the combined response will" NOT "be as smooth as physically possible" because Dirac Live doesn't optimize the crossover region and it doesn't optimize multiple subs individually. I don't think Dirac Live can optimize something it has no knowledge of.

Nov 11, 2014 8:20:00 GMT -5 flak said:

Room response correction is really what Dirac Live is about, so yes... the combined response will be as smooth as physically possible within the measurement area.

Flavio

[urwi]

Isn't the goal of a room correction system to optimize the in-room response? Of yes then I would think that it is fundamental for any digital room correction system to know crossover frequency and filter slope.

Nov 11, 2014 8:20:00 GMT -5 flak said:

@ Tony (ansat)

I don't know whether the crossovers are active during Dirac Live measurements, in which case Dirac Live would be able to compensate for them as well... Emotiva can answer that.

[Jim]

Nov 11, 2014 15:33:34 GMT -5 urwi said:

Isn't the goal of a room correction system to optimize the in-room response? Of yes then I would think that it is fundamental for any digital room correction system to know crossover frequency and filter slope.

Nov 11, 2014 8:20:00 GMT -5 flak said:

@ Tony (ansat)

I don't know whether the crossovers are active during Dirac Live measurements, in which case Dirac Live would be able to compensate for them as well... Emotiva can answer that.

I've never heard of any room correction system knowing crossover frequency and slope and doing something with that data.

I generally think of room correction systems handling the room. Not the speaker and sub integration.

[urwi]

Nov 11, 2014 16:31:08 GMT -5 Jim said:

Nov 11, 2014 15:33:34 GMT -5 urwi said:

Isn't the goal of a room correction system to optimize the in-room response? Of yes then I would think that it is fundamental for any digital room correction system to know crossover frequency and filter slope.

I've never heard of any room correction system knowing crossover frequency and slope and doing something with that data.

Well, it better should because that is what you hear. You don't listen to your sub or your left main or your right main or any other speaker in isolation. You listen to them as one source depending on the signal in each channel.
By the way, there are room correction systems that do "something with that data".

Nov 11, 2014 16:31:08 GMT -5 Jim said:

I generally think of room correction systems handling the room. Not the speaker and sub integration.

Any room correction system can only change the signal leaving the speaker. So it will always have an effect on speaker interaction and room alike.

[bluescale]

Nov 11, 2014 16:31:08 GMT -5 Jim said:

I've never heard of any room correction system knowing crossover frequency and slope and doing something with that data.

I generally think of room correction systems handling the room. Not the speaker and sub integration.

When you EQ the room, you use the xover and slope to determine what the filters are for your configuration. Dirac, as wonderful as it may be, isn't magical.

[Jim]

Nov 11, 2014 19:43:14 GMT -5 bluescale said:

Nov 11, 2014 16:31:08 GMT -5 Jim said:

I've never heard of any room correction system knowing crossover frequency and slope and doing something with that data.

I generally think of room correction systems handling the room. Not the speaker and sub integration.

When you EQ the room, you use the xover and slope to determine what the filters are for your configuration. Dirac, as wonderful as it may be, isn't magical.

That was my train of thought...

[htinparadise]

Flak:

The bigger question is: will Dirac work properly with active speakers, like say, for example the Emo Airmotive units?

Regards,
HTinP

[Gary Cook]

Nov 18, 2014 22:44:58 GMT -5 htinparadise said:

Flak:
The bigger question is: will Dirac work properly with active speakers, like say, for example the Emo Airmotive units?
Regards,
HTinP

Why would they be any different to any other power amplifier and any other speaker? They just happen to be in the same box, so I can't see that there would be any difference. What am I missing?

Cheers
Gary

[hifiaudio2]

Nov 11, 2014 8:20:00 GMT -5 flak said:

@ hifiaudio2

no, you did not miss any announcement
and no, I mentioned two subwoofers in different positions... there is an inherent problem in trying to compensate four subwoofers playing the same signal, but being in different physical locations.
This will inhibit Dirac Live's ability to do impulse correction (due to it trying to compensate within an area), it just is not pysically possible.

If you want to use all of your subs may be you can place them physically grouped 2 by 2, each group connected to one sub output of the XMC-1 so that Dirac Live will have the possibility of doing the impulse correction... or alternatively you could use just 2 subs.



Flavio

Hi Flavio, I am just now getting back around to reading this and thinking about it. So I can hook this up in two ways:

1) Use my current setup where I have time aligned the 4 subs with an outboard Xilica processor and the help of some fine folk at the Home Theater Shack and REW. I would connect the 4 subs to just one output of the XMC as I have now and let Dirac do its thing while "hearing" the combined response of all 4 subs on that one output. The response of the 4 subs together was measured and distance settings were set in the XMC so that I got the best response and phase alignment near the crossover region and the smoothest overall frequency response.

OR

2) Connect the front and rear subs to one output and left and right subs to the second output, turning off all attempts to time align the subs in the Xilica. The front and rear are pretty close to the same distance from the MLP (maybe a foot difference at most I think) and the left and right are also similar. Then just let the XMC Dirac setup hear each set of subs separately and do its calculations. This seems to be what you are suggesting that I do above?

... I also suppose that if it mattered I could attempt to make the time alignment of the front/rear and left/right even closer by still using the delays of the Xilica even if I chose option #2 above instead of just letting them pass the signals through. Is it worth it to take this extra step if I go with option #2?

Which is "more correct" and should yield the best results from the XMC and Dirac?

[smgoff]

Flak,

In the XMC-1 does bass management occur before Dirac equalization or afterwards?

[bootman]

Dec 2, 2014 17:31:55 GMT -5 smgoff said:

Flak,

In the XMC-1 does bass management occur before Dirac equalization or afterwards?

Flak works for dirac not Emotiva.

Bass management happens on the XMC first. Dirac doesn't do bass management.

[ansat]

Dec 2, 2014 17:31:55 GMT -5 smgoff said:

Flak,

In the XMC-1 does bass management occur before Dirac equalization or afterwards?

From Flavio

Nov 11, 2014 8:20:00 GMT -5 flak said:

@ hifiaudio2

no, you did not miss any announcement
and no, I mentioned two subwoofers in different positions... there is an inherent problem in trying to compensate four subwoofers playing the same signal, but being in different physical locations.
This will inhibit Dirac Live's ability to do impulse correction (due to it trying to compensate within an area), it just is not pysically possible.

If you want to use all of your subs may be you can place them physically grouped 2 by 2, each group connected to one sub output of the XMC-1 so that Dirac Live will have the possibility of doing the impulse correction... or alternatively you could use just 2 subs.

@ Tony (ansat)

I don't know whether the crossovers are active during Dirac Live measurements, in which case Dirac Live would be able to compensate for them as well... Emotiva can answer that.

@ urwi

Room response correction is really what Dirac Live is about, so yes... the combined response will be as smooth as physically possible within the measurement area.

Flavio

[douglasssf]

Flavio-

Thanks for your input in this forum and in the miniDSP nanoAVR DL forum.  I'm trying to get a basic grip on the starting-off case of "small" LCR/surrounds and a single subwoofer that handles both the LFE channel and the bass portion of the LCR/surround channels redirected to it by the receiver's bass management.  I.e., a garden-variety 5.1 setup.

Is the following a fair summary of how Dirac Live works on the XMC-1 for this configuration?

1.  Dirac will properly correct the subwoofer's output (combined LFE plus BM-redirected signal) in the frequency domain, assuming BM was active during measurement and isn't changed afterwards.
2.  Unless the subwoofer is physically close to the speakers whose low frequencies are being redirected to it by BM, Dirac will have a hard time correcting in the time domain, because each of the LCR/surround channels will be coming from two different locations--the higher frequencies from a "small" speaker and the lower frequencies from the subwoofer.

If this fairly describes how Dirac Live works on the XMC-1, can you suggest a work-around for #2?

Are there other Dirac Live implementations that time-correct the BM signal (or the combined LFE-plus-BM signal) as if it were a separate channel?  The literature on the Datasat RS20i makes it sound like it might do that.

Thanks.

[ansat]

Dec 17, 2014 1:59:55 GMT -5 douglasssf said:

Flavio-

Thanks for your input in this forum and in the miniDSP nanoAVR DL forum.  I'm trying to get a basic grip on the starting-off case of "small" LCR/surrounds and a single subwoofer that handles both the LFE channel and the bass portion of the LCR/surround channels redirected to it by the receiver's bass management.  I.e., a garden-variety 5.1 setup.

Is the following a fair summary of how Dirac Live works on the XMC-1 for this configuration?

1.  Dirac will properly correct the subwoofer's output (combined LFE plus BM-redirected signal) in the frequency domain, assuming BM was active during measurement and isn't changed afterwards.
2.  Unless the subwoofer is physically close to the speakers whose low frequencies are being redirected to it by BM, Dirac will have a hard time correcting in the time domain, because each of the LCR/surround channels will be coming from two different locations--the higher frequencies from a "small" speaker and the lower frequencies from the subwoofer.

If this fairly describes how Dirac Live works on the XMC-1, can you suggest a work-around for #2?

Are there other Dirac Live implementations that time-correct the BM signal (or the combined LFE-plus-BM signal) as if it were a separate channel?  The literature on the Datasat RS20i makes it sound like it might do that.

Thanks.

Dirac treats the speaker as full range. BM is applied to the corrected signals.

Tony