TOPIC: DDRC-88A - 7.2.4 with 2 units

I've read that 2 of the DDRC-88A's can be used together to provide correction for a system with more than 8 channels.

Will there be any issues using 2 units in the same system?

@ Cal900,

While we haven't done extensive test, we do believe that it should work fine providing you do as follow. Each DDRC-88 will do its own time and level alignment per unit. However this level and time may not be the same for UNIT A and unit B. So you'll need to use your AVR (doing the Atmos decoding) to tweak the "overall" level and gain to match up on "one" of the units. Let's try to explain.
A) Setup your first DDRC unit. It will time and level align all 8 speakers.
Connect your USB port to the second unit and perform the calibration there. It will time and level align all remaining speakers.

AT this point, you may need to use REW to know what is the latency and levelling that Dirac decided to apply on unit B and unit A. You then apply that difference through the AVR level/gain settings. e.g. DDRC1 @ 13mS total delay vs DDRC2 @ 15mS. Just apply 2ms offset in your AVR for all outputs going to DDRC1.

Does this make sense?

DevTeam

That does make sense which is what I was concerned about if I did try to use 2 units together. I have never used REW so I do not know how to check if 2 separate units are in time with each other.

Depending on how the speakers can be split up between the 2 units I did think of a logical way to try and ensure that 2 units are time aligned.

Unit A - Left, Centre, Right, Left Surround, Left Back, Top Front Left, Top Rear Left and Left Subwoofer

Unit B - Left, Centre, Right, Right Surround, Right Back, Top Front Right, Top Rear Right and Right Subwoofer

As the subwoofers usually have the most delay and in the situation where a pair of subwoofers are equally placed in the room and each at the same distance from the listening position I would hope that this process could ensure that each MiniDSP unit ends up with practically the same level and timing.

I do not know if having the front 3 speakers connected to each unit for each calibration would make any difference but I do not think that it would do any harm. The Front Left can be plugges back in to Unit A after calibration. The subwoofers would also need to be reduced by 2dB or 3dB each in order to allow for the gain added by using a pair that are not calibrated together.

It may also be worth trying both subwoofers connected to each unit for both calibrations instead of the Front Left on Unit B and the Front Right on Unit A. This could possibly be more accurate when calibrating the output of 2 subwoofers togerther as 1.

@ cal900,

mmmh, not sure if this setup is the best configuration..
If you ever have some delay in this configuration, it will be very obvious (i.e. imaging will become messed up). It's much better that you group the channel based configuration (i.e. L&R, center.. )on the same unit. If for example L&R (on unit1) are offset in time with rear L&R surrounds (on unit2), it's not so much of a big deal..

Another option would be to keep one speaker of your configuration (e.g. one common sub) as a reference and connect it to both units when doing the tuning. This way it will be used as a reference by Dirac Live algorithm.

DevTeam

Thank you.

I was guessing what may help to acheive the closest possible delay on each unit.

Your suggesion about the common subwoofer seems to be similar to the last part of my previous post. I will try that first and see how it works out.

I have seen it mentioned that you are working on a possible way to somehow link (or sync) 2 units to ensure the time and levels match, is this correct?

@ Cal900,

Yes, it's indeed something that one day we'd be looking at but no timeline.. Running 16ch = complete different new Dirac License, complete new DLL (i.e. new way the unit are discovered) so it's not as "plug&play" as it seems.

DevTeam

Devteam,

I would be interested in the ability for two DDRC-88A to be synced together and get 16-channels. In the meantime, would the suggestion of using the the sub input for each DDRC-88A be the correct way to go for syncing two independent DDRC-88A units? The reason being is that the sub generally has the longest delay associated of all the speakers and Dirac Live would take this into consideration when creating the filters for both independent units.

What about using two laptops with 88A number 1 on one laptop and 88A number 2 on the second laptop. Run mic position 1 for speakers and sub that are connected to 88A (1). Then switch cables from farthest speaker to second 88A (2) and run the calibration. This way the mic is in exactly same position. Then move mic to position 2 and run calibration on 88A (2), switch cables, placing farthest speaker back on 88A (1) and run second mic position. Repeat for remaining mic positions. This should give the same delay for both units, since the mic will be in the exact same position and since you are using the farthest speaker to set delay in both system, shouldn't delay be the same?

I could be wrong about this, but if the goal is to have Dirac Live calculate the delays and levels for two DDRC-88A units, then you will want to have one speaker connected to both units that is a. closer than all the others and b. louder than all the others.

The other way (better, IMHO) to do it would be simply to calibrate both units separately. Assuming that you have a spare channel on the second unit, connect one of the speakers for the first unit to the second unit. Once calibration is complete, use REW (or similar) to measure delay and level of that "shared" speaker when driven through unit #1 and unit #2. That will tell you what adjustment to level/delay to make in the source AVR. The adjustments will either be to all channels of unit#1 or all channels of unit #2.

The sub would not be (I don't think) a good choice for calculating relative time delays, because it's hard to see exactly where the impulse response is.

Hi john,

Thanks for the feedback and great suggestion of using a common speaker to calibrate two independent DDRC-88A units... Is the reluctance of not using the sub channel due to the overall group delay associated with the sub or that filters set by Dirac for the sub is different compared to the standard speakers?

Was thinking of using the center speaker as this is generally a single channel by itself in a multi-channel soundtrack and generally the most important for movies.

Would like to discuss some specifics on the trying to sync two DDRC-88A using the center speaker. Would it be possible to have two DLCT apps running concurrently in the Windows environment and connected to two independent DDRC-88A? Is it okay for two DLCT apps to share a common UMIK-1 mic? Would there be an issue if the output volumes between the two DDRC-88A are not exactly the same?

Once the delta is determined using REW, the unit with the shorter delay would need to be increased so that it matches the unit with the longer delay - correct? This can be accomplished on the receiver/pre/pro.

Hi Stephen, I think you may be overthinking it a bit What I would do is just run the calibration for each unit separately. Afterwards use REW (or whatever measurement program) to measure the level and delay of the "common" speaker when connected through unit A and unit B. That will tell you the relative level and delay that you can use as your offset values in the AVR.

The only mic position that is critical is the first one, as that is the one Dirac uses to calculate levels and delays.

Once calibrated, the "common" speaker is of course connected to only one of the units. You "lose" a channel. I can think of some possible ways around it but I would "cross that bridge if I came to it"

I wouldn't use the sub as the common speaker because it is hard(er) to accurately measure the delay of a sub.

Hope that helps

It seems that I too have been overthinking this. Going back to the basics of my limited knoweldge of room EQ technologies confirms that it is usually the subwoofers which already have the most delaying and Dirac will be applying delay to the speakers to be in time with the subwoofers. Therefore I will attempt to use the centre speaker as the common speaker and check the results with REW.

Another thing that I was wondering about is the best way to calibrate a pair of subwoofers when using 2 DDRC-88A's as there are 3 possibilities.

1) both subwoofers split off a single output from one DDRC-88A
2) both subwoofers connected to separate outputs of one DDRC-88A
3) one subwoofer connected separately to each DDRC-88A

cal900,

Jerry Austin has written a guide on how to do (1).

I would think (1) would be the better option as:

(2) Timing differences between two independent subs acting in isolation may give strange results. One would need to time align both subs to act as one unit.
(3) More complex than (2) due to two separate DDRC-88A units and the slight delays between the two.

If you are refering to the guide in AusinJerry's sig over on AVS then that is using a MiniDSP 2x4. I was hoping not to have to add another box or any more cables than already planned.

My subwoofers are placed an equal distance from the MLP so I would not expect to have to manually time align them if connected to separate outputs on the same DDRC-88A. I would only expect to have to reduce the combined level of them after running the calibration.

I will try all 3 options to find out if I can tell any difference between them but ideally I will have to measure with REW just to be sure. I recently read that the UMIK-1 mic is no good for comparing time alignment in REW as it is a USB mic. So now have to research other options and add to the cost.

devteam wrote:
Hi DevTeam,

I already have one 88A unit and am thilled with it, it is fabulous, so much that I am going to add a 2nd unit to EQ the rest of my Atmos 4 height speakers (and 2 front wides also).

My question obviously regarding delay between the two DDRC 88A untis; assuming I cannot accurately measure/calculate the offset by running REW to time align, what is the "margin" of delay offset (in milliseconds) I'm most likely to have between the units. In your comments above you give an example of 13-15 ms. is this likely to be typical? Indeed, what is the "worst case scenario" for the biggest internal latent offset between two manufactured DDRC 88A's in milliseconds? If it were only a 2-3 milliseconds max between them I think I could live with that in my room without having to delve into REW to exactly time align.

many Thanks in advance for your support.

Cheers

Sean