TOPIC:

This may be here somewhere, but I don't see it in any obvious place and I've become a bit confused.  A bit of a noob with DIRAC although I have used older MiniDSP 2x4's.  Is my understanding correct that one would install the DDRC-24, run a calibration in REW to figure out crossover points/timing/phase then analyze the results to build and apply parametric filters (assuming REW here) load them into a preset and THEN run DIRAC on that preset?  Or is DIRAC meant to completely replace the plugin?  Is the order of operation then: Crossover>Timings>PEQ>DIRAC
Or are there options such as:
Preset 1-nothing
Preset 2-REW/Plugin only
Preset 3 DIRAC only
Preset 4-REW/Plugin + DIRAC

I am also assuming that the DIRAC profile loaded between 3 and 4 must be independently determined.

Thanks for any help and I hope this isn't somewhere obvious that I haven't found.

 

What is the system context here? Main speakers plus subwoofer or active crossover in a 2-way loudspeaker. Of something else?

Generally there is no point applying any PEQ if Dirac will be used, unless potentially it's based on quasi-anechoic data to correct the speaker response. What you've described also doesn't need to be saved across multiple presets but rather you can just go from one stage to the next using the same preset.

Thanks for the reply.  Sorry about lack of context.  Still very much a novice at this.  I am using it in a 2.1 setup to integrate a sub with full range floor standing speakers (Polk Monitor 70's and a PSW 10 sub).  I have a 5.1 setup (with Polk CS2 and Monitor 50's in addition), but I like music in 2.1.  I am using an AV receiver (Yamaha RX-V685).  For home theater use, I use the integrated Yamaha YPAO room correction incorporating all speakers.  For music, I bypass all that stuff and use the 2.1 in "pure direct" mode.  Music runs from my PC via USB to the DDRC then to the receiver and I use the DDRC to control volume (leaving the AV and sub at gains I set and tested in REW, not sure this is really correct with respect to gain structure). What isn't clear to me is if DIRAC is complimentary to PEQ or supersedes it.
So, what I understand is one should set crossovers/timing/balance/phasing and maybe PEQ for the sub with the plugin, then just go directly to DIRAC.  I imagine that the corrections DIRAC comes up with would be different based on the presence of underlying PEQ from REW so deriving them independently and layering would be a bad idea.

I appreciate the help and input.

Sorry but I'm a bit confused by your last post. Are you asking about the DDRC-24 for use in your music system, your movie system, or both?

Whatever system, think or Dirac as a more advanced form of PEQ, so you'd use one or the other but not both. You do want to set any crossover filters and time delays to integrate a subwoofer with main speakers ahead of running Dirac Live though.

Ultrasonic wrote:

What is the system context here? Main speakers plus subwoofer or active crossover in a 2-way loudspeaker. Of something else?

Generally there is no point applying any PEQ if Dirac will be used, unless potentially it's based on quasi-anechoic data to correct the speaker response. What you've described also doesn't need to be saved across multiple presets but rather you can just go from one stage to the next using the same preset.

I've always taken the view that it is worth using the PEQ capability to get the overall performance of the speakers to the point where they are working acceptably to the ear (and to REW) before I run Dirac. I figure that Dirac has a limited amount of space that it can use for its filters, so making use of PEQ to correct gross anomalies & level matching issues before running Dirac will allow Dirac to get the most out of the memory space available to it. I must admit I haven't done much in the way of A-B comparison to check whether what I do is better than just handing the problem to Dirac but mu gut tells me it ought to work better that way.

Tony_J wrote:

Ultrasonic wrote:

What is the system context here? Main speakers plus subwoofer or active crossover in a 2-way loudspeaker. Of something else?

Generally there is no point applying any PEQ if Dirac will be used, unless potentially it's based on quasi-anechoic data to correct the speaker response. What you've described also doesn't need to be saved across multiple presets but rather you can just go from one stage to the next using the same preset.

I've always taken the view that it is worth using the PEQ capability to get the overall performance of the speakers to the point where they are working acceptably to the ear (and to REW) before I run Dirac. I figure that Dirac has a limited amount of space that it can use for its filters, so making use of PEQ to correct gross anomalies & level matching issues before running Dirac will allow Dirac to get the most out of the memory space available to it. I must admit I haven't done much in the way of A-B comparison to check whether what I do is better than just handing the problem to Dirac but mu gut tells me it ought to work better that way.

I suspect you're coming at this from the perspective of using a miniDSP as a crossover in a loudspeaker rather than to integrate a subwoofer? Above I referred to making PEQ corrections on quasi-anechoic data which would fit with the former, but I didn't repeat this in my last post as it doesn't apply to the OP. 

Here's why I don't personally think it's worth bothering with PEQ before Dirac Live if it's just based on in-room measurements:

www.minidsp.com/forum/shd-series/18380-peq-before-dirac-live

The one other exception I can think of is actually something I may end up doing myself: using PEQ functionality to apply a high shelf to elevate treble but then only use Dirac Live for low bass frequencies.

Ultrasonic wrote:

Tony_J wrote:

Ultrasonic wrote:

What is the system context here? Main speakers plus subwoofer or active crossover in a 2-way loudspeaker. Of something else?

Generally there is no point applying any PEQ if Dirac will be used, unless potentially it's based on quasi-anechoic data to correct the speaker response. What you've described also doesn't need to be saved across multiple presets but rather you can just go from one stage to the next using the same preset.

I've always taken the view that it is worth using the PEQ capability to get the overall performance of the speakers to the point where they are working acceptably to the ear (and to REW) before I run Dirac. I figure that Dirac has a limited amount of space that it can use for its filters, so making use of PEQ to correct gross anomalies & level matching issues before running Dirac will allow Dirac to get the most out of the memory space available to it. I must admit I haven't done much in the way of A-B comparison to check whether what I do is better than just handing the problem to Dirac but mu gut tells me it ought to work better that way.

I suspect you're coming at this from the perspective of using a miniDSP as a crossover in a loudspeaker rather than to integrate a subwoofer? Above I referred to making PEQ corrections on quasi-anechoic data which would fit with the former, but I didn't repeat this in my last post as it doesn't apply to the OP. 

Here's why I don't personally think it's worth bothering with PEQ before Dirac Live if it's just based on in-room measurements:

www.minidsp.com/forum/shd-series/18380-peq-before-dirac-live

The one other exception I can think of is actually something I may end up doing myself: using PEQ functionality to apply a high shelf to elevate treble but then only use Dirac Live for low bass frequencies.

Yes, I am looking at it from the perspective of creating a crossover for a N-way active speaker, rather than for subwoofer integration. Your measurements in the thread you quote are interesting - I may try doing a similar set for my 3-way setup to see what that reveals!

Tony_J wrote:

Your measurements in the thread you quote are interesting - I may try doing a similar set for my 3-way setup to see what that reveals!

It will be interesting to see the results if you do  . 

System is 5.1.
TV/cable box plays through AV receiver via HDMI to 5.1 speakers with room EQ done by integrated YPAO.  Not a big fan of the sound for music.  Movies seem fine.
Computer/music with my own or Qobuz source plays via USB to DDRC-24 to AV Receiver /Sub using 1-3 as L/R/Sub outs (Sub is Y connected to DDRC and AV Receiver).
With AVR in non-processing direct mode, music plays to front and sub (2.1) and the DDRC-24 is basically a pre-amp.  I control volume with the DDRC remote/master out.
All the same system. all the same room.
For the music 2.1 mode:
1) Used REW with L/R/Center averaged sweeps for both fronts and sub to set crossover between sub and fronts/timing/PEQ.  Installed on preset 2. Sound goes from good to awesome.
2) With only crossover/timing settings in preset 3, I used DIRAC with 9 or so measurements to generate a profile.  Sound goes from good to awesome but different.
3) Activating DL while on preset 2, sound is good to awesome, but different still.

Before breaking out the equipment and redoing everything, I was wondering what the intended process is, as it is not clear if it is either or both.  From your referred post it looks like Dirac results are similar with or without PEQ.  But to me the sound is different in each case.

I am guessing that the recommendation here is to skip the PEQ and just use DIRAC.

Thanks for the interest and responses.  I may need to spend an afternoon going through all the scenarios to document what is going on.


 

"I am guessing that the recommendation here is to skip the PEQ and just use DIRAC."

I would put it differently. Choose the approach that gives you the results you like best. If you find that just using Dirac gives very similar results to using Dirac + EQ, then there's no reason to go that extra step. If you prefer the result from Dirac + EQ, then do that.

What I have found when setting up 2/3 way actives is that if I spend a lot of time tuning the crossover settings and EQ to match levels, get rid of nasty peaks, etc., then the additional difference made by running a Dirac sweep is small, but nonetheless very worthwhile. As Ultrasonic has said, that is different (incomplexity if nothing else) from integrating a sub with a pair of passives, but the principles are still the same - you need to match levels, determine the right crossover frequency, decide whether the sub polarity should be reversed to get the best performance ar crossover frequency, etc., all of which you could simply leave to Dirac to sort out but you may find that a bit of pre-tuning using the PEQ and crossover settings will bring benefits. 

Tony_J wrote:

As Ultrasonic has said, that is different (incomplexity if nothing else) from integrating a sub with a pair of passives, but the principles are still the same - you need to match levels, determine the right crossover frequency, decide whether the sub polarity should be reversed to get the best performance ar crossover frequency, etc., all of which you could simply leave to Dirac to sort out but you may find that a bit of pre-tuning using the PEQ and crossover settings will bring benefits. 

Actually I'd say the parts in bold can't really be sorted by Dirac.

For clarity, I'd level match using the subwoofer's gain control and then use REW measurements and the miniDSP plugin to define time delays and crossover filters. It is only PEQ that I wouldn't bother with if someone is going to be using Dirac Live anyway.

Higher up the frequency range I'd say the key distinction is whether PEQ is being used to compensate for known issues in the direct signal (i.e.via quasi anechoic data of some form, or introducing an intentional effect like a treble lift or a 'BBC dip'), or if it's just going to be an adjustment made on (ungated) in-room measurements. If it's the latter then intuitively I don't really see the point as this is what Dirac will be doing anyway? Not something I'll be experimenting with though as I don't like the result of using Dirac at higher frequencies. 

Ultrasonic wrote:

Tony_J wrote:

As Ultrasonic has said, that is different (incomplexity if nothing else) from integrating a sub with a pair of passives, but the principles are still the same - you need to match levels, determine the right crossover frequency, decide whether the sub polarity should be reversed to get the best performance ar crossover frequency, etc., all of which you could simply leave to Dirac to sort out but you may find that a bit of pre-tuning using the PEQ and crossover settings will bring benefits. 

Actually I'd say the parts in bold can't really be sorted by Dirac.

For clarity, I'd level match using the subwoofer's gain control and then use REW measurements and the miniDSP plugin to define time delays and crossover filters. It is only PEQ that I wouldn't bother with if someone is going to be using Dirac Live anyway.

Higher up the frequency range I'd say the key distinction is whether PEQ is being used to compensate for known issues in the direct signal (i.e.via quasi anechoic data of some form, or introducing an intentional effect like a treble lift or a 'BBC dip'), or if it's just going to be an adjustment made on (ungated) in-room measurements. If it's the latter then intuitively I don't really see the point as this is what Dirac will be doing anyway? Not something I'll be experimenting with though as I don't like the result of using Dirac at higher frequencies. 

You're correct of course - Dirac won't give you the right crossover frequency, but it would certainly make a stab at fixing things if the sub polarity was wrong. Maybe not a good stab, but a stab.

I guess my intuition tells me that it is best to fix as many "known" problems as possible before unleashing the automagic of Dirac on the problem.

 

Tony_J wrote:

I guess my intuition tells me that it is best to fix as many "known" problems as possible before unleashing the automagic of Dirac on the problem.
 

This comes back to what you mean by 'known problem' as I tried to cover above. If  you mean in terms of using PEQ to match to an in-room target curve then I see no distinction to just using Dirac Live, but if you mean something else then I do  . 

Ultrasonic wrote:

Higher up the frequency range I'd say the key distinction is whether PEQ is being used to compensate for known issues in the direct signal (i.e.via quasi anechoic data of some form, or introducing an intentional effect like a treble lift or a 'BBC dip'), 

Mulling this over I suspect that if Dirac Live is run over the frequency range affected it will actually tend to undo changes such as above.

This has all been very helpful so thank you both.

The sub integration with my front passives is definitely handled well with REW.  I only needed 1 filter to lower a bump on the sub that was also very long on decay.  That, plus the crossover and timings and it is greatly improved.  I would say going from muddied boom to actually musical.  So I keep that in all cases.  I didn't need anything drastic with the PEQ for the higher frequencies on the passives but the results between just Dirac and just REW/PEQ are plainly different, and I haven't decided which I like best.  But, preference aside, my personality compels me to understand just what it is I am doing and that I am following the protocol to ensure reproducibility.  I'd like to use the equipment in the way it was designed for.  Eventually I plan to add an amp and cut out the AVR for music listening at which point I will re-do everything anyway.  Hearing your opinions/preferences on this has been much more useful than simple answers.

Cheers!