TOPIC:

Hi,

I have just started using a minidsp ddrc-22d and the stereo image is biased to the left. Although, the sound definition is much better with the minidsp, the stereo image problem is clearly a deal breaker. I hope some of you have a solution. Here some information about the problem.

1) The stereo image is centered without the minidsp.
2) The stereo image is centered when I set the filter to "off" on one of the four filter on the minidsp.
3) Before buying the minidsp, I used the trial version of the Dirac live calibration tool with the UMIK-1 microphone. I noticed that the Dirac software put a delay on one of the two speaker with the result of an uncentered stereo image. However, the Dirac processor has the option to modify this delay. By setting the delay to zero I could have a centered stereo image.
4) During the mic configuration step, I noticed that the right speaker sounded a bit louder (very little difference).

I took the measurements a few times to be sure that everything was fine but with the same problem every time. Is there a way to control to delay one of the speaker with the minidsp? Could the point 4) above be the cause of the problem?

Anyone has a solution?

Thanks

When you do the first measurement, is the microphone the exact same distance from both speakers?

If so, then I don't know what could be causing it. Perhaps resetting to defaults might clear something... or you may need to send your project file to miniDSP support to have a look at. Unfortunately there is no way in the software to see what levels and delays Dirac has set.

Thanks for the answer. The microphone was at the same distance from both speakers. I did not know that I could send the project file to miniDSP support, I'll do that at last resort.

I investigated a bit more and I might have found where does the problem come from.

As I said before, during the mic configuration step, the right speaker sounds louder than the left one. In the Dirac software, the left speaker is linked to channel 1 on the computer while the right speaker is linked to channel 2. I switched the channels so that the left speaker was linked channel 2 and the right speaker was linked to channel 1. In this configuration, it was the left speaker that was louder instead of the right one. It seems that both channels on my laptop don't have the same output level. However, I could not find an option in windows to adjust the output level of each channel.

I will try the Dirac software on a different laptop (macbook) this weekend, but other suggestions are welcome.

I'm not sure I'm following.... but:

* The laptop settings will have no effect when doing a calibration
* If you suspect that the laptop settings are not correct when you are playing music after calibration, perhaps you can try playing a CD through it from a CD /DVD/Bluray player?

I'm not a specialist, but here how I understand it.

-The sound measurements are done with the laptop, so, the right speaker is louder than the left one during the measurements. Because of that, the Dirac software creates a filter which takes into account this difference. (configutation: windows laptop with the Dirac software, UMIK-1 plugged to the laptop, laptop goes to the miniDSP-DDRC -22d which goes to a DAC which goes to the amplifier.)

-I already use a different source for playing music. If the output levels are the same for this source (which should be the case), then the filter in the miniDSP tries to correct a difference in the output levels that is not there, which should create the stereo image problem.

I really appreciate your help. Don't hesitate to tell me if I'm not clear, English is not my first language.

Hi again - the Dirac Live test signals are generated in the DDRC. Therefore, the laptop soundcard settings will not affect the test signals.

Did you reset to defaults with the DDRC utility? (Also, do refresh DSP program.)

You should also check that you have the latest firmware and software versions. It seems like you should, but it won't hurt to check.

After that, the best way to establish whether there are level differences is to measure it.

* On the Output&Levels screen, you can see the meter. Is there a difference there?
* Do a calibration with a single measurement point. Then run a sweep on each channel (using e.g. REW). Overlay the two plots. That will tell you clearly if there is a left-right difference after calibration.
* I read that REW now has a way to measure time delays when using a USB microphone. I haven't used it, but you could try it to measure time delay on both channels.

To open a support ticket -> minidsp.desk.com

HTH

Thanks for the suggestions, it will help to better understand what is going on. I'll be busy for a couple of weeks, so it may take a while before I come back with the results.

@ Genitruc

We got your tech support ticket. The engineers will look at your project and get back to you there.
Thanks @ John for your help!

Devteam

@ Genitruc

Hello,

Did you solve the problem?

I have now the same dilemma: I made a few measurements with a Dirac Live trial license on my PC and the filters always add a delay and gain for one speaker, although the sweet spot is perfectly symmetrical relative to the speakers and the 9 measurements were well executed. I have a 2.0 speaker configuration.

Also, at setup, the two output level indicators were showing perfectly equal signal levels when tested.

When using the filters, if I leave the gain and delay settings off, the sound is great, but with them turned on everything is compromised.

I intend to order a DDRC-24, but if I cannot disable the specific speaker delay/gain settings (like i am able to do with the PC software version), then this is a deal breaker.

Thanks!

How much is the delay and gain difference?

Hi, John,

I'm not very sure about the exact values now, because I'm not at home, but the left speaker had a 0.6 db gain and aprox 0.3 ms delay.

I have made 3 measurements with the default optimisation curve and one of them had the left speaker with a 0.7 db gain. Although the speakers and the listening position are in a perfect symmetrical configuration, the amount of objects/furniture in the right side of the room is significant higher. They are outside the listening triangle, anyway.

Maybe this uneven landscape in the room is the cause of the delay/gain adjustments, but in terms of stereo imaging everything is much better with these settings off than when using them.

My speakers are two Elac 243BEs on stands, driven by a Cyrus 6a amplifier. The source is the PC with the ODAC usb dac.

Hope this helps,
Thanks

Hi catalinmarin,

I solved my problem by manually creating different target curves for each left and right speakers. In my case, I had to put a difference of approximately 3.5db to get a centered stereo image. As for the delay, you have no manual control on that with the software version for the minidsp. However, there is no indication if the software put a delay or not. Since, the minidsp correct for time response I would guess that a delay is used to improve time response but it is only a guess.

Some advice from my experience.
1) I used mono recording to test the stereo image. You can use Audacity to convert stereo recordings to mono.
2) Before using the minidsp I tried to improve as much as possible the stereo image by moving the speakers. If your listening room is not symmetric, it is possible that the optimal speaker position will not be symmetrical.
3) When manually creating the target curves, I first created target curves with the same profile (same shape) but with different gain. After I've got a generally centered stereo image, I noticed that different frequencies could be slightly more to the right or to the left. With those observations I could fine tune the target curves to improve further the stereo image.

I hope this is helpful
Genitruc

catalinmarin wrote:

the left speaker had a 0.6 db gain and aprox 0.3 ms delay.

The first thing I would do is check with a tape measure that the first mic position is exactly the same distance from the speakers. That kind of difference is easy to enough to introduce if you just "eyeball" it. (How do I know this...) Perhaps you did that, but I thought I should mention it. I'm not sure if it would make a difference to the imaging but if you're wanting to get to the bottom of it it would be something to do. You should only need a single point measurement to find out what the delays/gains are set to by Dirac Live.

The DDRC-24 (unlike the DDRC-22) does have the ability to adjust gains and delays in the output channels. You can't change what Dirac Live does, but you can change those things later. See User Manual. You also can't directly view what Dirac Live sets so you would have to do measurements with REW to figure it out. (Or just experiment)

If the DDRC-24 can adjust gain and delay per channel, regardless of what the Dirac plugin does, then this is the great news for me!

Thank you for your answer!