TOPIC:

naiku wrote:

Figured I should start posting my thoughts here, hopefully we can keep everyone's tips, tricks etc in a single place.

In my case I did the upgraded firmware from the 8x12 to the 8x12DL, the upgrade process was very quick and easy, it took me longer to find something that I could use to copy the firmware files onto a MicroSD card than it did to do the actual upgrade. Getting the activation key was also a piece of cake, sent in the request on a Friday evening (EST) and had my activation key by Sunday around 10pm (EST, so about 10am HK time).

The software did crash on me 3 times while switching between presets, I would start on Pre4, go to Pre1, then going back to Pre4 it would crash. One of the options was to close the software, the other a reset. After the 3rd time I did the reset option and so far it has been good since, although I have not switched between the presets as much as am just concentrating on Preset 1 for now.

At the moment, I am trying to figure out the best options for routing, I am running
3 way front
rear fill
single sub

I attached what I have currently set up, but I am not convinced it is correct.

My main concerns are around rear fill and the subwoofer, I had been thinking of using Dirac 1 through 8 for all speakers individually except the subwoofer, then to manually tune the sub with the help of REW. On the Routing tab if I turn off Dirac 7 and Dirac 8 for Bass Mgt then I lose the sub. On the Mixer Tab, should I leave everything off under the sub?

For rear fill, in the 8x12 it was simple enough, so I tried to mirror that with Dirac 7 and Dirac 8 by inverting one of the inputs. I am using toslink, so only using Input7d and 8d

If anyone can help confirm what the subwoofer and rear fill settings should be on those that would be great. I have read the manual a few times, but cannot quite grasp it for some reason.

You should be able to right click on sub and lower the % it’s mixing from each Dirac sweep.

How you have it looks good, if rears are full range I would keep it like is. That should give you more flexibility in the front when you unlink channels ,

As far as unlinking channels , I would first keep all the fronts linked to a single target , once you have a good sounding target and your gain structure is made so the “all before” tab shows a smooth response as possible, than unlink and play with each separate speaker in the 3ways crossover region/slope using the target leaving the target shape on the inband stuff alone ,

I’ve had good luck with using Dirac target as the beginning of my crossover, doing that will force an acoustical slope for crossovers and keep it coherent to the other side. Just remember when doing that to keep the “pairs” of speakers linked in 3 way but unlinking others otherwise you’ll make a de-correlation filter because your separation from the target on one side. Whatever unlinking you do just make sure the other side has the same thing.

The manual says you can unlink and follow a dip on one speaker to help with over eq however I have found that it messes things up even more as far as sound staging goes.

If you have a massive dip somewhere try pulling your target shape to go below the deepest dip and should sound better.
It may not be that way in a room but in a car the dips are usually comb filters so trying to boost into them even with Dirac can fatigue the speakers, if you keep your target under the dips it won’t try and use boost, it may seem like your cutting a lot of the signal but it’s really not that much and it’s well worth it so long as your gain structure is made right . You have plenty of voltage coming out of the cdspdl it Can easily be made up in amp gain without adding any noise to system or turning gain up past unity shouldn’t ever have to happen.


Good luck let us know how you like it

Andrew

naiku wrote:

My main concerns are around rear fill and the subwoofer, I had been thinking of using Dirac 1 through 8 for all speakers individually except the subwoofer, then to manually tune the sub with the help of REW.

You should use one Dirac channel per each speaker group. And you should use Dirac also for the subwoofer.

I have attached pics of my config, still untested... I use inputs 1 and 3. I noticed that input 2 has higher noise level, at least in the meter, so I switched to using input 3.

Tomorrow I will finally do the install, I have opposite problem with the weather: too cold and raining. I made some room inside my warmed garage, now I can fit my car in there...

btw. I have owned DDRC-88A just over 4 years and use it with the BM plugin. It looks very similar to C-DSP 8x12DL:n plugin. So mixer/routing is not (or shouldn't be) a a big problem.

Thanks both, will hopefully find some time later today to work on things a little.

TimoJ wrote:

You should use one Dirac channel per each speaker group. And you should use Dirac also for the subwoofer.

I have attached pics of my config, still untested...

This method makes sense as it lines up with application guide in the manual (Section 4.10.4) for an active setup.

Installed the C-DSP already today. Seems to work fine. Tomorrow I'll check amp gains and measure levels etc. with REW. And re-measure front speaker distances and set correct delays per element so that they are time aligned (Dirac does the L/R/Sub listening position delaying). And then finally some Dirac magic, I hope

Running into problems here, opened the DLCT, it recognizes my C-DSP is plugged in, then recognizes the UMIK-1, but when I get to the Output and Levels tab, I get nothing at all from any speakers even with the Output Volume slider all the way up. Checked and double checked connections, can't see anything wrong.

Any suggestions?

No idea why DLCT doesn't work, if mute (also channel outputs, not just master mute) is not active and enough volume, it should output test tones.
Btw. download the latest 1.4a plugin, it fixes few issues that I noticed and reported earlier.

Yeah, nothing is muted, either Master or individual channels muted, really strange. Installed 1.4a and will see if that made any difference, although I doubt it, but will at least try again shortly. Otherwise, might have to shoot an email over to support. I found a thread where someone was having the same issue with a DDRC unit, but they never replied with what the solution was.

As expected, 1.4a made no difference... Sent an email to support.

Also looked at the troubleshooting section of the manual... :

a. Ensure that the processor is connected correctly into the audio
system. - Signal chain is: - Tablet -> USB DAC -> Optical -> C-DSP -> Amplifiers. Unless by this step the DSP should be connected to the laptop for the audio signal? It is not very clear though. I'm tempted to unplug the USB-DAC from the tablet and plug it into the laptop to see if that does work
b. Check that the downstream amplification is powered on. - Check
c. Check that the downstream amplification is not muted and doesn’t
have gain/trim controls set to zero. - Check
d. Quit DLCT, open the C-DSP 8x12 DL plugin and click Connect.
Connect an analog source to the inputs and confirm that signal
levels are seen on input and output meters. - I have not done this, but music is playing through the unit and I can see levels.
e. Check that master mute is not enabled. Check


Last Update for now... got it sort of working. In the DLCT I was thinking in terms of the Output channels being 5 (Left, Right, Sub, Rear Left and Rear Right). But, since I am using TOSLINK I have 2 outputs (7 and 8), so changed one of the outputs on the Output and Level tab to 7 and sure enough got some pink noise.

Went through all 9 measurements, clicked Optimize (the DLCT help says there is a button that says "Save Filters" which I am not seeing anywhere). There are still some pretty large dips, but overall the target and impulse look better. I need to get better microphone placement to be honest, was trying to balance the microphone on a tripod and just get an idea of things.

After clicking Optimize, I then went to the next tab and dragged the Dirac project into preset 2 on the DSP.

Fired up the DSP, started playing some music. Click Dirac ON and.............

Sounds. Exactly. The. Same.

If I look on the Dirac tab, the values don't change with Dirac on vs Dirac off. I'll be honest, it could well be user error, but as far as I can tell I have followed all the steps correctly. Even if the Dirac filters were no good due to microphone positioning, there should be a noticeable difference in sound and those values should change.

At least it is progress. Will likely try from the beginning again later.

rccotroneo wrote:

TimoJ wrote:

You should use one Dirac channel per each speaker group. And you should use Dirac also for the subwoofer.

I have attached pics of my config, still untested...

This method makes sense as it lines up with application guide in the manual (Section 4.10.4) for an active setup.

As long as Dirac channels are linked what difference does it make.. although yes I would agree each group should have its own target linked. How you get there is up to you, if I had extra Dirac channels to use dam skippy I would use them to control crossover regions on independent speaker pairs in a multi-way , and control them from the target and not using peq.
Although in some instances peq would be same thing and maybe work almost better, however using the target to shape crossovers will keep things correlated and coherent on non minimum phase problem areas.

It’s a matter of knowing your cars acoustics and behaviors. But I still wouldn’t leave Dirac channels off the table in multi-way , again, if all linked to a single target what difference does it make except give you more potential control as you learn to navigate your system with Dirac

naiku wrote:

Yeah, nothing is muted, either Master or individual channels muted, really strange. Installed 1.4a and will see if that made any difference, although I doubt it, but will at least try again shortly. Otherwise, might have to shoot an email over to support. I found a thread where someone was having the same issue with a DDRC unit, but they never replied with what the solution was.

As expected, 1.4a made no difference... Sent an email to support.

Also looked at the troubleshooting section of the manual... :

a. Ensure that the processor is connected correctly into the audio
system. - Signal chain is: - Tablet -> USB DAC -> Optical -> C-DSP -> Amplifiers. Unless by this step the DSP should be connected to the laptop for the audio signal? It is not very clear though. I'm tempted to unplug the USB-DAC from the tablet and plug it into the laptop to see if that does work
b. Check that the downstream amplification is powered on. - Check
c. Check that the downstream amplification is not muted and doesn’t
have gain/trim controls set to zero. - Check
d. Quit DLCT, open the C-DSP 8x12 DL plugin and click Connect.
Connect an analog source to the inputs and confirm that signal
levels are seen on input and output meters. - I have not done this, but music is playing through the unit and I can see levels.
e. Check that master mute is not enabled. Check


Last Update for now... got it sort of working. In the DLCT I was thinking in terms of the Output channels being 5 (Left, Right, Sub, Rear Left and Rear Right). But, since I am using TOSLINK I have 2 outputs (7 and 8), so changed one of the outputs on the Output and Level tab to 7 and sure enough got some pink noise.

Went through all 9 measurements, clicked Optimize (the DLCT help says there is a button that says "Save Filters" which I am not seeing anywhere). There are still some pretty large dips, but overall the target and impulse look better. I need to get better microphone placement to be honest, was trying to balance the microphone on a tripod and just get an idea of things.

After clicking Optimize, I then went to the next tab and dragged the Dirac project into preset 2 on the DSP.

Fired up the DSP, started playing some music. Click Dirac ON and.............

Sounds. Exactly. The. Same.

If I look on the Dirac tab, the values don't change with Dirac on vs Dirac off. I'll be honest, it could well be user error, but as far as I can tell I have followed all the steps correctly. Even if the Dirac filters were no good due to microphone positioning, there should be a noticeable difference in sound and those values should change.

At least it is progress. Will likely try from the beginning again later.

Somethings wrong.
Keep playin with it, you’ll get it. It should sound killer , it seems your Dirac may not be assigned to the right outputs maybe, and also remember each Dirac bank (1-4) are also associated with the current configuration (1-4). If you drag Dirac into box 1 you have to listen to configuration 1 to hear it

You’ll get it ,

Also and I can’t stress this enough
(Going on 2years with Dirac in car) keep you target below the deepest null or dip in response
Even if it’s -10db deep just draw your target to the shape you desire but keep it below the dips in measured response, it will sound 100x better

naiku wrote:

If I look on the Dirac tab, the values don't change with Dirac on vs Dirac off. I'll be honest, it could well be user error, but as far as I can tell I have followed all the steps correctly. Even if the Dirac filters were no good due to microphone positioning, there should be a noticeable difference in sound and those values should change.

Dirac values should not change. Those are the measured delay and level values that are now applied to that C-DSP preset all the time, no matter if Dirac is enabled or not. To actually fully compare, you should create identical preset with the C-DSP plugin but with an empty Dirac slot allocated to that preset. But then delays (unless set with output settings) are all over the place, so it's not a fair comparision.

But you should still notice a big sound and imaging difference with Dirac on/off within the same preset, unless your system was already perfect

denver8me4dinner wrote:

As long as Dirac channels are linked what difference does it make.. although yes I would agree each group should have its own target linked. How you get there is up to you, if I had extra Dirac channels to use dam skippy I would use them to control crossover regions on independent speaker pairs in a multi-way , and control them from the target and not using peq.
Although in some instances peq would be same thing and maybe work almost better, however using the target to shape crossovers will keep things correlated and coherent on non minimum phase problem areas.

Not sure, it would be nice to have Dirac's opinion/technical details to this. I'm not going to use output PEQs, I'll see how REW measurements look, then see what Dirac does and then tweak crossovers etc. Maybe I'll later test with separate Dirac channels for front lows and use one Dirac pair for front mid/high. And since Dirac does very accurate delay measurements, I think I'll test with front each split to different Dirac channels, that way I can get the delay difference numbers for each element and then use those (instead of tape measure) with my current setup. Lots of things to test...

I did REW measurements etc. Things looked pretty nice without any extra modifications.

Then I tried Dirac... Looks like it's all messed up or I didn't get enough sleep... It doesn't output to the right channels, there is no logic. Signal that should come from the front, comes from the rear, subwoofer should use Dirac 5 but instead test tone comes via Dirac 7 etc. It's a total mess. This was with Dirac Custom setup. I'll have another look soon, reset the device and try to make notes and contact support about this.

TimoJ wrote:

Dirac values should not change. Those are the measured delay and level values that are now applied to that C-DSP preset all the time, no matter if Dirac is enabled or not. To actually fully compare, you should create identical preset with the C-DSP plugin but with an empty Dirac slot allocated to that preset. But then delays (unless set with output settings) are all over the place, so it's not a fair comparision.

But you should still notice a big sound and imaging difference with Dirac on/off within the same preset, unless your system was already perfect

In my case though, the Level and Delay columns are showing a zero and the Output is the same as the Input. And, yeah, my system is far from perfect, no noticeable difference with it on / off.

However, I had emailed support last night and they replied requesting copies of the DSP preset file as well as the Dirac project file. Sent both of those over and they had initially replied saying:

The issue is actually quite simple. You have misconfigured the setup I’m afraid.
You have a stereo source but you have 5 x speakers/zones to tune. Correct? See how your stereo source is being routed to 5 x “Dirac” channels.

The setup you should select from the drop down menu as a CUSTOM device in Dirac Live is therefore a 5ch setup, not a 2ch.. When you go to Dirac Live, make sure that you select Custom 5ch, not 2ch as you dropped now. The channels will also be DIrac 1-5, not Dirac 7/8.

I then responded back to them saying that I had initially been using CUSTOM and 5 speakers, but that was when I was getting zero output of tones from within DLCT, that I only got any tones when using 7 & 8. After that they then responded with:

Thank you! We filed it as bug and try to investigate for a fix in few days, sorry for that.

In the meantime, you can only route Left Right digital inputs to Dirac 1 and Dirac 2 (in “Routing” page) and make use of “Mixer” page to setup the cross-points.

So, at least for using digital inputs it looks like there is a bug. I may tinker with it on Dirac 1 and Dirac 2 only just to get used to microphone positioning, tweaking the targets etc. But am hoping the bug fix does not take too long.

naiku wrote:

So, at least for using digital inputs it looks like there is a bug. I may tinker with it on Dirac 1 and Dirac 2 only just to get used to microphone positioning, tweaking the targets etc. But am hoping the bug fix does not take too long.

Like I wrote earlier today, I had major problems also with analog inputs. Looks like they have somehow mixed input mixer settings to how Dirac sets it's output channels. I use inputs 1 and 3. Only Dirac 1 (front left) test signal works correctly, Dirac 2 (should be front right) has no output, Dirac 3 (should be rear left) outputs from front right and both rear channels all at the same time. Dirac 4 (should be rear right) and 5 (should be sub) have no output. And subwoofer test signal outputs via Dirac 7. ..

So no Dirac tests today.