TOPIC:

I just got an email from minidsp announcing the CDSP 8x12 DL, which is an automotive Dirac Live interface. Just what I've been dreaming about!

Available for pre-order with plans to ship mid april!
www.minidsp.com/products/car-audio-dsp/c...8x12-dl?acm=56216_70

The one thing that gives me somewhat of a pause is that they mention that the old non-Dirac CDSP 8x12 can be infield upgraded to the DL version once they get the flashing system ready. I thought that Dirac Live required quite significant processing power, so it seems surprising that no new chips would be needed (although I suppose that "infield" doesn't necessarily mean no new chips). Is the CDSP 8x12 already so overpowered that it can handle DL without breaking a sweat?

Anyway, this seems like a gamechanger for car audio, and I am really excited.

The fir bank in the C-DSP is not used (until now)

Remember that they plan things years in advance. When this unit was made , DL was probably already in mind.



I would like to know what version of DL the C-DSP will be using . Standard or automotive?

I'm excited about this. Hope to upgrade my current 8x12 V2 to the DL version.

I'm don't fully understand Dirac Live yet so I have a question:

Does it use FIR filters for the crossover filters so they don't induce phase rotation?.. Or is the Dirac Live function only for eq?

rccotroneo wrote:

I'm excited about this. Hope to upgrade my current 8x12 V2 to the DL version.

I'm don't fully understand Dirac Live yet so I have a question:

Does it use FIR filters for the crossover filters so they don't induce phase rotation?.. Or is the Dirac Live function only for eq?

No it dosent the crossovers are minimum phase

Dirac itself uses mixed phase filters , linear phase and minimum phase,
It uses your measurements and creates a transfer function that matches for both sides

Mixed phase meaning IIR and FIR.
It does an amazing job, it does more than linearize a crossover. In a car , small reflections and small phase issues aren’t as big of a deal and stage height and depth are better use of your traditional crossovers, getting both sides to match is very tricky and this Dirac does a fantastic job at this.

I don’t know what version of DL it will run but out of experience with DL in my car it works wonders

denver8me4dinner wrote:

rccotroneo wrote:

I'm excited about this. Hope to upgrade my current 8x12 V2 to the DL version.

I'm don't fully understand Dirac Live yet so I have a question:

Does it use FIR filters for the crossover filters so they don't induce phase rotation?.. Or is the Dirac Live function only for eq?

No it dosent the crossovers are minimum phase

Dirac itself uses mixed phase filters , linear phase and minimum phase,
It uses your measurements and creates a transfer function that matches for both sides

Mixed phase meaning IIR and FIR.
It does an amazing job, it does more than linearize a crossover. In a car , small reflections and small phase issues aren’t as big of a deal and stage height and depth are better use of your traditional crossovers, getting both sides to match is very tricky and this Dirac does a fantastic job at this.

I don’t know what version of DL it will run but out of experience with DL in my car it works wonders

First off - You must be oabeieo on Diyma... I recognize your avatar I'm Truthunter over there.

Thanks for the info - but it leads me to other questions Really want to get my mind wrapped around how this will be utilized properly.

So, if the Dirac Live function is on the 8 outputs - are target curves chosen for the Dirac Live function to do it's corrections based on the drivers (tweet, mid, mbass, sub) the outputs will be used for? And if so - how with the Dirac Live function correct for phase coherancy in the crossover regions between the drivers?

My main issue currently with the standard 8x12 v2 is working to get optimized phase coherency between the drivers (Sub-Mbass, Mbass-mid...). I am hoping this Dirac Live function will make with easier/quicker.

:EDIT:. Nevermind... I found the answer to this elsewhere... Dirac function will occurs before mixer & xovers.

To get a better feel for the process of setting this thing up I read over the manual for the DDRC-88A as it was stated elsewhere that the 8x12DL will operate similar.

I noted in section 5.2.2 (Preparing for acoustic measurements / Subwoofer settings) it states "We recommend that the low pass filter of the subwoofer be disabled if possible, or set to its highest frequency if it cannot be disabled."

My question if the LP filter is disabled for measurments - how will the phase rotation that filter contributes be accounted for in correction filter?

rccotroneo wrote:

denver8me4dinner wrote:

rccotroneo wrote:

I'm excited about this. Hope to upgrade my current 8x12 V2 to the DL version.

I'm don't fully understand Dirac Live yet so I have a question:

Does it use FIR filters for the crossover filters so they don't induce phase rotation?.. Or is the Dirac Live function only for eq?

No it dosent the crossovers are minimum phase

Dirac itself uses mixed phase filters , linear phase and minimum phase,
It uses your measurements and creates a transfer function that matches for both sides

Mixed phase meaning IIR and FIR.
It does an amazing job, it does more than linearize a crossover. In a car , small reflections and small phase issues aren’t as big of a deal and stage height and depth are better use of your traditional crossovers, getting both sides to match is very tricky and this Dirac does a fantastic job at this.

I don’t know what version of DL it will run but out of experience with DL in my car it works wonders

First off - You must be oabeieo on Diyma... I recognize your avatar I'm Truthunter over there.

Thanks for the info - but it leads me to other questions Really want to get my mind wrapped around how this will be utilized properly.

So, if the Dirac Live function is on the 8 outputs - are target curves chosen for the Dirac Live function to do it's corrections based on the drivers (tweet, mid, mbass, sub) the outputs will be used for? And if so - how with the Dirac Live function correct for phase coherancy in the crossover regions between the drivers?

My main issue currently with the standard 8x12 v2 is working to get optimized phase coherency between the drivers (Sub-Mbass, Mbass-mid...). I am hoping this Dirac Live function will make with easier/quicker.

:EDIT:. Nevermind... I found the answer to this elsewhere... Dirac function will occurs before mixer & xovers.

Glad it got answered , yeah it’s me lol

At least on the 88BM when I tune with that, it’s one common target across many passbands
It will move from driver to driver (Dirac sweeps) and let you adjust the target in the passband. It’s not 8 full 20-20k targets to manipulate (I wish it was) it’s one common target. But it works very well ,

denver8me4dinner wrote:

Glad it got answered , yeah it’s me lol

At least on the 88BM when I tune with that, it’s one common target across many passbands
It will move from driver to driver (Dirac sweeps) and let you adjust the target in the passband. It’s not 8 full 20-20k targets to manipulate (I wish it was) it’s one common target. But it works very well ,

Thanks! Another question:

Do you disable the LP filter on the sub before your measurements?... like is noted in the 88A manual in section 5.2.2 (Preparing for acoustic measurements / Subwoofer settings) it states "We recommend that the low pass filter of the subwoofer be disabled if possible, or set to its highest frequency if it cannot be disabled."

If so - if that LP filter is disabled for measurments - how will the phase rotation that filter contributes be accounted for in correction filter?

You can try it like that

I would , it probably wants to make the midbass
Which is hard to do

Try it both ways

rccotroneo wrote:

denver8me4dinner wrote:

Glad it got answered , yeah it’s me lol

At least on the 88BM when I tune with that, it’s one common target across many passbands
It will move from driver to driver (Dirac sweeps) and let you adjust the target in the passband. It’s not 8 full 20-20k targets to manipulate (I wish it was) it’s one common target. But it works very well ,

Thanks! Another question:

Do you disable the LP filter on the sub before your measurements?... like is noted in the 88A manual in section 5.2.2 (Preparing for acoustic measurements / Subwoofer settings) it states "We recommend that the low pass filter of the subwoofer be disabled if possible, or set to its highest frequency if it cannot be disabled."

If so - if that LP filter is disabled for measurments - how will the phase rotation that filter contributes be accounted for in correction filter?

Keep in mind this subwoofer is primarily 100% minimum phase in a car , if the sub is on the input of the Dirac does it need to be turned on acoustically? It would be better for it to make midbass smooth and calculate the sub on your own , And also with your bass knob you’re going to want more base than what’s within the target so your base knob you would want down Almost all the way if you were including it in measurements

So I’m not sure why, but I’ve run mine with no sub on the acoustic sweeps and had excellent results. The midbass in that config was dialed and adding sub later and doing just normal peq worked good.

It would be a 2 way plus sub instead of a 3 way


But I’m not sure , I’m just saying I’ve done it and it sounded good

It will be interesting to see the differences in the user manuals between the 88A & 8x12DL. The 88A user manual is primary targeting HT applications whereas the automotive application 8x12 should be quite different. For instance will it have different listening environment options other than chair/sofa/auditorium.... like perhaps driver's seat only / front seats / all seats. Also in HT applications - the speakers, whether mains or surrounds, are pretty much full range and then there is a sub, whereas aftermarket automotive is usually all discreet drivers for different freq ranges.

rccotroneo wrote:

It will be interesting to see the differences in the user manuals between the 88A & 8x12DL. The 88A user manual is primary targeting HT applications whereas the automotive application 8x12 should be quite different. For instance will it have different listening environment options other than chair/sofa/auditorium.... like perhaps driver's seat only / front seats / all seats. Also in HT applications - the speakers, whether mains or surrounds, are pretty much full range and then there is a sub, whereas aftermarket automotive is usually all discreet drivers for different freq ranges.

Yeah It will be interesting to see for sure

I'm all about having this unit replace my current 6x8cdsp ....My system is has tweets front and rear / mids/midbass front and rear > All biamped + 2 subs front and rear (opposed in between seats / front and rear facing) ... I also use the DDRC88BM for my HT setup so im very familiar with how that system operates with my 7.1 speakers - those speakers are a 3 way passive xover ...

How do you think the new 8x12 will be setup with speakers bi-amped in my truck ?

Now that I think about it
Phase correction under about 125hz in a car is sorta useless. Eq does everything you need

A fir linear phase or linearized crossover on a sub is nice , because we know what the behavior or the crossover is simply speaking. So we can predict it’s effects. But the car in the sub range shouldn’t have any phase issues in the sub areas much under 70hz at least.

It actually makes sense now to not have the sub as part of the room correction. The group delay added by a port or the crossover can be mitigated by other means , but not as a part of the room correction filter that is tied to the highs. The sub should be done independently with eq and use of a linear phase fir filter as a crossover.


The Dirac should get things in the ballpark that it may not even be a issue and eq would be the only thing.


The sub LPF phase rotation isn’t hardly noticeable unless the midbass is way off compared to it. If the midbass is corrected, it will lead the sub. But yes a linearized filter dedicated just on the sub would be the “best” on top of Dirac. But that’s not what it’s doing as part of “room correction “ not “crossover linearization “

You would want your crossover right at where the sub drops off from the 1st combfilter dip in the 70-125hz range in 99.9% of cars due to the room shape and all else should work just fine.

rts100x5 wrote:

I'm all about having this unit replace my current 6x8cdsp ....My system is has tweets front and rear / mids/midbass front and rear > All biamped + 2 subs front and rear (opposed in between seats / front and rear facing) ... I also use the DDRC88BM for my HT setup so im very familiar with how that system operates with my 7.1 speakers - those speakers are a 3 way passive xover ...

How do you think the new 8x12 will be setup with speakers bi-amped in my truck ?

It would blow your mind.
With some small tweaks and a sober approach to how to implement the rears it would be amazing

Might have to add subtract a little delay to rears post Dirac to get your center dialed and get the rears to reinforce the center image in middle of dashboard. But Dirac should do most of it with ease.