Home Theater Forum and Systems banner

REW or Dirac Live: which tells the truth?

20K views 33 replies 6 participants last post by  FargateOne 
#1 ·
Hello,
I am worried for my front right speaker or front right test signal. I'd like to know if it is nothing or if I have a real issue.

I performed a Dirac Live stereo calibration with NanoAVR-DL when I remarked that the curve of the right channel was very different in the low fr than the curves I was used to get with REW since many months now.
For REW measurements, the hdmi output from the laptop is connected to video 3 input in the receiver and I think that the test signal comes from the laptop via rew software.

For Dirac, hdmi output from nano is connected to video 1 in the receiver. All others settings are the same (2 channels PCM, no sub, speakers to large, no EQ in the receiver, no xover) As you know, for Dirac, the source bluray must send a video active signal and is setted to PCM With Dirac is the test signal coming from the laptop via usb cable or is it already encoded somewere in nano, I don't know.

Of course, the mic remains at is first position for rew measurements.

Dirac gives average 9 positions but it seems irrelevant to me for my purpose.

Here is front right Dirac:
Text Blue Line Technology Font


And front right from REW:

Text Line Plot Diagram Pattern


To see my problem here is the left from dirac:

Text Blue Line Technology Font


And the left from REW:

Text Line Font Plot Pattern


In REW I see what I think that it is the roll-off limits of my left B&W (approx 45 hz) BUT I do not see it in Dirac for the right speaker.

I was afraid that I had burned a speaker. (see my journey to install Dirac !! I am still a little bit scared about that)) So I reverse left speaker to right position and right to left and remake measurements.
I got the same (more or less) curves.

Do I have reasons to be worried? Is it a problem with Dirac software? Or is it nothing to worries about.
Thank again for your help and patience.
View attachment few g ddirac.mdat
 
See less See more
5
#3 ·
Rew curves were setted to var smooth which, as you know doesn ot smooth low fr where I have my interrogation; also I joined the mdat file for Rew measurements.
I repeated the experiment 2 or 3 times with the same result in the right channel.
Also the room is asymetric but it is the same fact for rew and dirac.
So, I would like to undeerstand the difference.
 
#4 ·
I think you're right to be concerned FargateOne and I would definitely not use the target curve on your right speaker Dirac post to create a filter. It will add a lot of spl to the low end which may, to your point, already be there.
Contacting MiniDsp again might be your best bet. Dirac tests the left speaker first so if it was an issue with your settings, like a crossover being set and the sub in standby mode, it would show as a problem with the left speaker and possibly the right also. But to have the left good and the right not good seems to point to something else.
 
#5 · (Edited)
Your answer confort me in the following. When I applied the filters set by Dirac to listen music, suddenly I heard on the right speaker (when tymbales were playing in the music) a low frequency noise not a sound. First I thought that it was a good result of Dirac making to hear something that was in the music but that I never heared before. Quickly I deactivated dirac filters because I suspected something wrong.
In fact, when I try Dirac for my 5.1 setting , those problems are not there because, the bass management in the receiver is set with a xover at 100hz and the left and right fronts have the same curves for the first and only sweep for the first measure with Dirac than in Rew with the xover to 100hz and no EQ.

Thank, I will contact Minidsp...

it is very very very difficult...I am near to give up Nano and Dirac
 
#7 ·
Referring to your first post, I agree that the difference you see for the right speaker measurements does not make sense. The REW and Dirac measurements should be a lot closer than that.

I recognize that you have had a frustrating path getting Dirac to work and that this thread represents a simplification to get Dirac working.

Just to be clear - all four measurements were taken without moving the mic and without changing system settings, right? Can you take the 4 measurements using the same HDMI input and no change of settings?
 
#8 ·
Yes indeed the mic was at the same place for all measurements. I deliberatly play only one sweep with Dirac, save the project, and do the same with Rew all the reasons why left channel curves are similar but not the right one.

About HDMI input, there was a difference. Sweeps for Dirac arrives via VIDEO 1 input in Rotel but sweep from Rew went through VIDEO 3 input.

To John Stewart,
maybe I express myself poorly. What I wanted to point in my previous post is that I followed the NanoAvr owner manual instructions with the standard path : source to nano to receiver. Here EQ off but bass management active i.e. delay, SPL and xover at 100hz then speakers. Dirac measures speakers without filters but bass managed. So, for Rew check, I respect the same path for the sweep. To achieve that for Rew, I had to change some settings in the receiver to by sure that it did not add filters to the fronts channels but to keep the bass management. Doing so, in 5.1 setup, Rew reads the signal beginning in the sub and continuing in the speker as for Dirac. And, as expected the 2 curves look familiar.
 
#12 ·
It is a mystery. I'm afraid I am out of ideas. I hate to have someone abandon Dirac Live disappointed, but I have no more ideas for you.

From your measurements, it looks like the right speaker is getting affected by the 100 Hertz crossover while you are taking the Dirac measurement, and is not affected by the crossover when you take the REW measurement. Why that would be I could not say.
 
#14 · (Edited)
Dear Wayne,
I think that the following might of interest for every body.
The idea was gracefully suggested by a member of an other forum wherre I have posted the same.
Here's the suggestion:
1. Turn the subwoofer OFF to make sure that it's not somehow being used.
2. Connect your laptop for REEW measurement to the second input of the nanoAVR (not to a separate input of the receiver).
3. Check your signal-to-noise - on the Output and Levels screen, where is the meter sitting if you turn the output level all the way down?
4. If it's still doing the weird thing, swap the left and right channels at the speaker leads. Is the strange reading still on the right channel in DLCT, or did it follow the speaker?

For all measurements the following are the same:
mic position
one sweep with Dirac
one sweep with REW
Receiver settings: front speakers to LARGE-Center to none-Surrounds to none- Subwoofer to NONE and just in case XOVER to OFF
levels left and right to 0db
Laptop to NanoAVR-DL via hdmi 2 input and from nano to receiver connected to Video 1 in the receiver (Rotel RSX-1562)

Now the procedure: One round with standard connections and one round with the front left output cables connected to the RIGHT speaker and the front right output cables to the LEFT speaker.

Now the results.
1- Standard connection

Left speaker Dirac

Text Blue Line Plot Technology


Right speker Dirac

Text Line Font Technology Screenshot


There's the problem

2- Reversed connection (left channel speaker to right speaker and vis-versa
LEFT speaker dirac connected to the cables from the right output of the receiver)

Text Line Technology Oscilloscope Font


Here again ...

Right speaker Dirac connected to the cables from the right outputs of the receiver

Text Line Font Technology Diagram


...but not here.

Now REW via Nano
1- LEFT speaker standard connection and REVERSED

Text Line Plot Font Slope


2- RIGHT speaker standard connection and REVERSED

Text Line Plot Font Slope


So , no problem with rew measures via nanoavr, no?

What is the conclusion and what to do?
 
#15 ·
If I read correctly the problem is with Dirac right channel with Stereo project and no problem with AVR, wires or speakers.
Also no problem with Dirac front right channel with 5.1 project.

So the question might be what is different between Stereo and 5.1 projects?

Were the Dirac input and output levels the same for each project and do you recall what they were?

You set nano to 5.1 for 5.1 project, yes? Did you set nano to Stereo or Auto for Stereo project?

I think any suggestion is a guess at this point, but very happy to try to understand and help while you wait for MiniDsp.
 
#16 ·
If I read correctly the problem is with Dirac right channel with Stereo project and no problem with AVR, wires or speakers.
Also no problem with Dirac front right channel with 5.1 project.
With 5.1 project Dirac does not show the mistake because each sweep sees the sub and the speaker as 1 speaker because bass management in the receiver has xover at 100 and speakers are set to small. Then the Dirac curve is very like Rew curves in the same situation

So the question might be what is different between Stereo and 5.1 projects?
Indeed.

Were the Dirac input and output levels the same for each project and do you recall what they were?
Approximatly the same : input more or less +5-6 db and output more or less -19 db

You set nano to 5.1 for 5.1 project, yes? Did you set nano to Stereo or Auto for Stereo project?
That was Minidsp solution for one of many problems I got. I setted nano to stereo for stereo project and if I recall correctly maybe once it was set to auto.

I think any suggestion is a guess at this point, but very happy to try to understand and help while you wait for MiniDsp.
Thanks. My patience with Dirac is very very limited now.
 
#19 ·
Hi FargateOne,
Tried to reproduce your results and could not. In each case left & right front had low frequency extension to 23Hz at 0db on Dirac's scale. No need to confirm with REW.

Front Set to Large. All other speakers off in AVR. AVR set to normal listening volume setting of -18db. Dirac input approx +6 Dirac output -19db.

Tried with Nano set for 5.1, Stereo and Auto. No difference with any.

For your reference also:
Software Version: 1.1.0.5939

Hardware:
nanoAVR-DL
Version v1.3
Firmware 1.22

I suppose you've already tried Refresh on the Nano Utility Program?

Sorry not more helpful.
 
#22 ·
Sir,
your message touched me. Thank you very much for asking. A thicket is open with Minidsp. They think that my results are odd. I have send them my projects files and waiting for their feedback. Someone elsewhere suspects very low frequencies unaudible.
Others think that the main suspect should be the receiver ( I was one of them at the beginning). But one way or an other, nothing explains why the sweep signal of dirac passing through the same path than rew doesn't give the same curve for the right speaker channel.
I do not use the nano for the moment. As someone said, it is prudent not to use the results neither for stereo nor for 5.1 setting because dirac could over compensate the low fr response of the right speaker.
Even if I have a 5.1 project who seems correct, I am not in the mood to try it.; I would not be objective in my listenning. As every one knows, our brain can trick us !! I prefer to wait and see if the problem can be solved or not.
Stay tuned...
 
#23 · (Edited)
An update.

It seems that the problem comes from the bluray ( I bought it in nov 2015 all firmwares updated). I disconnected it. Then I made Dirac meaurements with the following configuration: laptop-NanoAvrDL-receiver (no bluray no cable box). The front right channel curve was correct. I would have like to check with Rew but for an unknown reason I couldn't send Rew sweep signal in this system. It was like if windows 10 did not send any sound from the laptop to the nano. Strange, because Dirac software is installed in the same laptop and worked well.

Then I did a reset to factory default in bluray. I connected it to the nano and nano to the receiver and laptop to nano. I did a Dirac measure and the front speaker curve response looks fine. As if the problem would be solved.

I do not see the logic in all this. How is it possible that when bluray is not passing any audio signal (only image via hdmi) it can affect the bass in the right channel signal send by Dirac to the receiver via Nano. AFAIK, the signal from laptop or from nano does not go to the bluray before to return into nano and to continue to the receiver!!??:dontknow:.

In front of such mysteries, I am still debating if I continue my journey.:|

I did not take chance and I sent the bluray to repair if any is needed. It is on warranty.

Could this be an explanation knowing that my settop box was connected to cambridge (which is the same platform as Oppo BTW)?
http://aldoussystems.blogspot.ca/2010/07/why-edid-management-in-hdmi-is-so.html:dontknow:

BTW why Rew was not affected by all this ? An other mystery.
 
#24 ·
That is really strange. How did to ever think to disconnect the blue ray for a test?

Regarding REW not working through the same signal chain, was the usb disconnected from the nano at that time? Also I've noticed that sometimes after creating and installing filters for the nano a power cycle of the nano is needed for it to pass a signal. Always with the usb disconnected from the nano.
 
#25 ·
That is really strange.
You bet! Imagine for me!

How did to ever think to disconnect the blue ray for a test?
Technical support of MiniDsp.

Regarding REW not working through the same signal chain, was the usb disconnected from the nano at that time?
No, I will keep that in mind.


Also I've noticed that sometimes after creating and installing filters for the nano a power cycle of the nano is needed for it to pass a signal. Always with the usb disconnected from the nano.
OK thank for the tip. Please can you explain what you mean by " a power cycle of nano is needed" ? Do you mean disconnect and connect again the nano or to put on and off every devices of the chain (bdplayer-nano-receiver-display)?
 
This post has been deleted
#33 ·
This may be a old thread but I suspect Dirac uses a different type of sweep than rew and adds to the power responce during certain mic positions of the calibration.

I am not saying it is fact but it is a strong hunch.

Seems when I run sweeps with Dirac some of the mic positions have more energy applied than others .

I would try doing a comparison with two mics one using rew and one w Dirac and let Dirac generate the sweeps
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top