TOPIC: Experiencing Bass Boost with Dirac Calibration

After running the Dirac Live calibration on my DDRC-88A, the low end of my frequency response curve is showing a 10dB boost (see attached REW measurement file). My configuration is not typical. I have a Marantz 8801 Pre-Pro, which is connected to the 88A using balanced connectors. The 88A is then connected to an Emotiva XPA-7 power amp, also using balanced cables. However, the sub signal path is UNBALANCED. From the Pre-Pro to the 88A, the cable is unbalanced. And the bass signal goes from the 88A into a MiniDSP 2x4, also using an unbalanced connection.

I have paid close attention to gain control. The sub output from the Pre-Pro measures 1.7Vrms, so the input on the 88A is configured for 2Vrms. The output from the 88A to the 2x4 is set to .9Vrms to match the .9Vrms input on the 2x4. I have measured all the voltages using a digital multmeter, and everything is as expected.

The sub is configured in the Dirac Live Calibration Tool custom configuration on output channel 3, and the subwoofer box is checked. Nothing sounds unusual during the calibration sweeps, yet when measuring the results using REW, the sub channel is boosted. I even measured the level of each channel with a test disk after the calibration, and the speakers were measuring 68dB, while the subwoofer measured 76dB

I am confused, and don't know what next steps to take. Kindly advise how to trouble-shoot this issue. I am ready to try just about anything.

If you don't see the attachment, here is a link: dl.dropboxusercontent.com/u/78476446/AVS...s%201-18/Post/lf.jpg

Hi Jerry, if you uncheck the Subwoofer checkbox when doing the measurement/calibration, does the sub go to the expected level? (If you have time to do this, might be best to start a fresh project)

I'll try to do the same thing later tonight or tomorrow.

john.reekie wrote:
Your suggestion should be easy to try, and I will do so today. I have received several others suggestions as well:

- Remove the single unbalanced connection that the sub channel uses, and make everything balanced. This would require either the purchase of. The balanced 2x4, or using an XLR-to-USB converter cable into the unbalance 2x4.

- Scrapping balanced connectors, replacing everything with unbalanced connectors.

- Eliminating the 2x4 completely.

Of the three suggestion, the second one would be the easiest to implement, but I really wanted to stay with balanced connections. My feeling is that the 88A should work as it is currently configure,and if there is a bug in the product, none of the suggestions may work.

By the way, is un-checking the sub box a long-term, solution? I was under the impression that Dirac handled sub channel filters differently.

John,

I have spent the entire day going through the various test scenarios. Summary:

- Un-checking the sub box didn't make any difference.
- Removing the MiniDSP 2x4 didn't make any difference.
- Changing the sub channel from unbalanced to balanced resolved the issue.

I think there is a bug, We should be able to mix balanced and unbalanced connections.

Hello Jerry,

mmmh, interesting issue and we can investigate here if we can reproduce as well. The balanced vs unbalanced does sound like the issue here as there is a difference in levels between the 2 but maybe something else? . Could you please do us a favor and share with us a project file that did have the issue? (i.e. not the latest file which seem to be all good). We'd want to investigate it. Easiest is to share it under a tech support ticket (minidsp.desk.com) this way we're sure to keep track of the issue.

Quick question to double confirm something:
- When you're making sweeps, it feels like signal is correct and pretty much leveled? i.e Bear in mind that the sweep is generated inside the DDRC-88A so trying to establish that issue is here already at that stage.)
- When doing the measurement with REW, where are you connecting it to? Your AVR input or the DDRC-88A input straight to one of the RCA?
I'm just trying to investigate if by any chance the error is being caused upstream. (i.e. inside the AVR, some specific setting that the DDRC-88A just isn't aware at all because it's downstream).

If you are connecting to your AVR input, would you mind to do a quick test with the REW signal directly to the DDRC-88A instead? (i.e. Measure the Left channel with REW vs right channel of the DDRC).

Last but not least, there is a misunderstanding with regards to what the SUB check box does and it's our fault with one earlier post somewhere... Let us try to clarify once for all hopefully:
- The Subwoofer check box does change the sweep signal and does change the way the filter is being calculated. It's recommended to make sure the signal is treated correctly.
- However, based on confirmation with the engineers and the type of command we're sending for the DDRC-88A, we do "not" add a 10dB boost for the Sub. So I "believe" (and we will know more from your project) that the issue here is elsewhere.. .

Thanks for your patience and sorry to hear that you've been troubleshooting all day. Please start a ticket next time so our team is aware when an issue arise. Sounds good?

DevTeam

I will be happy to provide the information you requested, and to answer the questions you have posed, but since it is late in the evening local time, I will do this tomorrow.

Sounds great. No rush in getting to the bottom of this I'd say since you just got your setup working maybe..
Best that you take few days to relax a bit, enjoy your Dirac setup before you go down the path of figuring out the issue here for the previous setup.

DevTeam

I have conducted additional testing on this issue, and the results are reported here: www.avsforum.com/forum/90-receivers-amps...42.html#post31031626

While I have resolved the issue by selecting the proper input voltage on the 2x4, I really don't understand why this resolves the issue, unless there is a bug in the 2x4's firmware. I would be interested in hearing the Support Team's comments.

Hi Jerry, with the 2x4 set to 0.9V input sensitivity, the 2x4 will output about 7 dB higher for the same input voltage (provided it is less than 0.9V). It's not clear to me from your description whether you re-ran the DLCT calibration with the two jumper settings - ?

Couple more questions: did you use the Channel volume sliders? I believe miniDSP is working on an update related to that.

Also, are you able to use the 2x4 plugin at the same time as DLCT? (This is not related to your issue/question here, but I am curious )

Let me see if I understand you correctly, John. Simply changing the input voltage setting on the 2x4 from 2Vrms to .9Vrms will increase the output by 7dB, assuming the 2x4 is receiving the same voltage signal in both scenarios? Is that documented somewhere? If I recall correctly, I was making the connection changes without re-running the DLCT each time, so if there really is a 7dB difference, re-running the DLCT should correct for it, right? And a 7dB boost would account for the bass boost in the REW measurements.

I was told never to use the channel volume sliders. Evidently there is a bug that needs to be worked out.

I recall getting this error when trying to run the 2x4 plug-in and one of the two 88A programs simultaneously, either the DDRC-88A Utility or the DLCT, can't remember which.

dl.dropboxusercontent.com/u/78476446/AVS...Not%20successful.JPG

EDIT: My bad. The 2x4 manual, in the gain structure section, clearly mentions the 7dB loss. I need to RTFM.

Hi Jerry, re-running DLCT should get the expected result, as Dirac Live sets levels internally so they are all equal. IOW, regardless of the 2x4 input sensitivity setting, DLCT should produce the same result. [Edit: with respect to the relative levels of the channels, the absolute levels may differ.]

I've just noticed this announcement from devteam, so best to hold off on doing any more calibrations if you've got it running satisfactory for now

www.minidsp.com/forum/announcements/1071...gs-being-fixed#17599

John is right for both statements. I think that what you're seeing here is a combination of a bug (level page). I'd also recommend that we wait early next week (Dirac's release date for their combined package of the DLL) so we can see if that issue is solved on your setup. Quite a lot of things have changed improved in that new rev so hopefully it will solve major issues. Thanks for your patience.

DevTeam

Good news, thanks!