TOPIC:

Just checked and it shows up as two HID devices, not sound cards.

Yup

I’m thinking of a completely diffrent dsp

Sorry about that

Heh

I can’t think what the levels thing could be than

Here is what Support said about the level error:
This error is actually normal when you’re making a measurement that is too low. Make sure that your level is within -12dB as per the user manual otherwise you’ll get that error (in a car or of a hifi, that’s the standard error).
You need to make sure that your microphone gain is high enough. You can raise it from the first fader in the "outputs and levels)

And second message, after I argued that the levels were correct:
It’s a warning It’s just telling you the level is too low to make a proper measurement. It will work but you won’t get the right results.
In a car environment, the measurement is challenging but the issue is the same as if we were to make a measurement in a room with a level that is too low.
Why not have a trial at what we suggested before dissmissing it right away. Maybe it will work

Dear All,

Quite an active thread.. Great to see all the feedback summarized here!

- Please note that a new update was released yesterday to solve few minor bugs. Make sure to grab it as our team is doing our best to get that complex software free of any issues. Feel free to start a tech support tix if you have any questions!

- Error message is indeed very odd as we're now aware of 2customers (Timo/Naiku) with that errors. We also know of quite a lot of other customers running the 8x12DL perfectly fine.. As explained above, that error would typically pop up under a standard Dirac Live build when the level is too low (E.g DDRC-22/24/SHD). Somehow, we're not sure if that's the case here as we're waiting for Timo to run a test as suggested above, that error pops up even if level is fine. Please try to play with the OUTPUTS and level fader to make sure that is the case. Each car is different. drivers are sometimes very far from the microphone and "obstructed". That might cause some error in the algorithm. We can't reproduce that error here but we're already on the case to try few things here to reproduce it (i.e. obstacles like a car seat between the path of the speaker and the microphone... We're also talk to Dirac to see if they would know what could be the cause..
Also important note, please make sure to clarify your setup as the algorithm might be looking for something specific in the sweep recording (E.g. maybe take a screenshot of your recording FFT + impulse response of the measurement when you're getting the error). That will help!

As it's always the case, we're always happy to help...

denver8me4dinner wrote:

Yup

I’m thinking of a completely diffrent dsp

Sorry about that

Heh

I can’t think what the levels thing could be than

@ denver8me4dinner

Just to confirm for others, Unlike the DDRC-24 which is indeed a USB audio interface (DAC + 4ch recording interface) and a USB HID device (control path). The CDSP 8x12DL doesn't have an XMOS so can't do USB audio streaming. It's a USB HID interface only.

DevTeam

DevTeam makes a very good point actually,

I’m a car audio installer professionally and hear a lot of cars, rear speakers especially have a hard time getting loud. Seats often are the obstacle.


As far as the levels error , could you just try process of elimination

Turn off rears in mixer and run DLCT with just fronts and see if persists. Than turn off fronts and see if rears are problem. Try moving the mic right in the path directly in between each set of speakers while doing it to ensure no obstacles and a good direct line of sight to mic for measurement.

Do process of elimination and find out exactly which channel (s) are causing the error message.

In DLCT Make sure the mic gain is at 0 and not attenuated and the output levels are up all the way on each channel (not the master output in the levels section in DLCT and not the outputs on each channel in the plugin) the master output should be set so each channel sounds fairly loud

Depending on the gains on your amps somewhere in the -40s or -30s.

See if you can get the message to at least go away and rule out what exactly is causing it.

What is the correct/best place to do system equalization for low level listening? Use Dirac target? Use PEQ for outputs after a Dirac run? Use PEQ for outputs before a Dirac run and make the Dirac target curve equal to the PEQ curve? Logical/easiest place would be to do this before Dirac, but there are no PEQs for this in the DL version. Using biquads to do EQ in input LPF seems to work, when tested with the plugin, but not sure how things go in reality... My current headunit has no loudness.

TimoJ wrote:

What is the correct/best place to do system equalization for low level listening? Use Dirac target? Use PEQ for outputs after a Dirac run? Use PEQ for outputs before a Dirac run and make the Dirac target curve equal to the PEQ curve? Logical/easiest place would be to do this before Dirac, but there are no PEQs for this in the DL version. Using biquads to do EQ in input LPF seems to work, when tested with the plugin, but not sure how things go in reality... My current headunit has no loudness.

I would make 2 separate Dirac targets and use the remote to switch between

But in all honesty many ways that could be done

So I got my unit installed

Having the no Dirac on channels 1-6 using toslink input

Tried every possible configuration possible to no avail.


I also had too low level , and quickly realized my gains on amps are too high for the output voltage of CDSP. I lowered my amp gains to just below unity which is 2.83vrms and message now gone as I can play the outputs much louder on DLCT , message gone however gains still seem too high. I’m guessing I’m getting more than 5vrms from this unit. I will oscilloscope test tomorrow and set levels want to get Dirac channels working first

Only difference between mine and timoj is I had almost no waveform in measurements box as it was all gain and very little signal.

Another thing I've been wondering: how I should setup the output filtering before a Dirac run? 3-way fronts and filtering done with the output xovers with one Dirac channel for left, one for right.

Should I have all filtering enabled? Or should I have a little wider bands so that Dirac can "see" and correct also a little below/above the xovers? But when I then set the correct values, system phase and response changes...

@denver8me4dinner

Make sure that you're running the latest plugin and click Restore to default all "while connected". It will clear that bug that was fixed few days ago.

Thanks for the feedback on the low level. Very interesting!

DevTeam

TimoJ wrote:

Another thing I've been wondering: how I should setup the output filtering before a Dirac run? 3-way fronts and filtering done with the output xovers with one Dirac channel for left, one for right.

Should I have all filtering enabled? Or should I have a little wider bands so that Dirac can "see" and correct also a little below/above the xovers? But when I then set the correct values, system phase and response changes...

I would build a proper crossover using the filters set, and use REW and close mic measurements to speakers. Use PEQ and crossovers to fix problems at the speakers , than let Dirac do the rest (eq , timing ,) you definitely want a proper crossover. Meaning, make sure your acoustical slopes mostly match the electrical slope.

Do close mic measurements and examine the crossover area in rew. Look for large peaks in stop band , knock those down with peq a little just so it resembles the shape of the crossover you choose. Don’t over do it tho, just make them smooth so they sum correctly in the first octave. Most direct radiators (with the exception of tweeters) don’t need a lot of any eq in stop band unless they are of high efficiency type. That excludes a large % of all high quality car audio speakers IMO, but still worth a peek to see how the behavior is looking. I’ve found a good chunk of car audio speakers are heavily damped and want to shut down quickly.

But that’s the short version, there’s a lot of reading online about a proper crossover and you could take a look into some of that, especially if playing a speaker through resonance. There’s a lot of variables to look at. However a quick look with rew can usually make a crossover performance better with just basic tuning.

What I would be more concerned is with driver interaction in the crossover regions, that’s where separate Dirac channels will come in very handy.

If you stay with LR4 type crossovers it’s fairly fail safe in a car, it’s just the right amount of attenuation vs blending to reject interference and keep off axis behavior under control etc. and they sum well together. And no confusing polarity flips

And still more things I've been wondering... I currently use 2 channel analog inputs. 3-way fronts and a sub (I'll leave rear fills out of this)
Should I use the input HP-filtering and a separate sub channel with LP. Or is using the filtering there in my setup actually useless, since there still is output filtering? Is the input filtering causing phase/delay/response changes to the input signal? (that Dirac of course can't see and correct) And can't you do the same sub channel split to a Dirac channel also in the Routing matrix?

Got 8ch to work

Get the levels too low message before it would even take that measurement, something seemed weird .....but let’s wait. I’m going to do my gains the right way tomorrow. I know there a mess right now I can hear it and see it.


OTOH , oh my !!!!!! A entire new level of clarity
It’s unbelievably amazing. Much better than a plethora of devices.... and this is the first go round and it’s amazing.

There’s definitely something special about this Unit. I’ve been diracing for a minute and this is entirely different. Like , way way way to the extreme on point and very good.

(I should’ve bought a 88BM Year’s ago! )
But this is better!

Wow

About that level too low message: what Windows are you using? I'm using Win 7 Pro 32bit.

I'm going to do more Dirac measurements later today and put levels almost to the max just to see if the low message still appears and also if there is any audio signal output before the message.