polarity correction

jrobbins50

Member
Thread Starter
Joined
Oct 4, 2018
Messages
54
Hello, all. In my main system, I use two different 2-channel DACs which run with 2 subwoofers. My Marantz AV-8802A preamp does the bass crossovers and thus Audiolense sees two full range speakers. I do my 2-channel measurements using the Project PreBox S2 Digital DAC, saving them in mono format for HQP convolution and separately in Roon format for use with my Bel Canto streamer (which only has ethernet input and thus can't be directly measured by Audiolense).

The measurements sound great with the PreBox. However, the same filters in Roon only sound great if I reverse the polarity of one of my main speakers.

I have always had Auto Polarity Correction running in Audiolense. How would I go about reversing the polarity of the filters in Audiolense (as I don't want to physically have to get up and switch the speaker inputs each time I switch from the PreBox to the Bel Canto DAC)? Do I have to do a separate set of measurements with the PreBox with APC shut off, or can I take the existing measurement and run Correction Procedure Designer in a manner that will reverse the polarity without having to do a whole new measurement set? And, of course, if the polarity for the PreBox is correct all along, deselecting APC won't help me anyway, as I see it -- without APC in this case, the polarity result will be the same as with the APC engaged and I would still need to reverse the polarity in Audiolense post-measurement, but before generating the filters.

Bernt, if such a thing can't be done at the moment in Audiolense, here's a good thing to add in for the next iteration of the software that you put forth.

Crew, I appreciate your advice. Thanks. JCR
 

juicehifi

Audiolense
Staff member
Joined
Feb 5, 2018
Messages
240
Hi jrobbins,

The two corrections should be identical if I understand your case correctly.

I can take a look at the two corrections and check if they are identical. Or you can do it yourself from the Analysis menu.

If they in fact are identical, one of the channels must be inversed in Roon for some reason.

If you want me to take a look, please send me the two corrections together with the subject measurement file.
 

jrobbins50

Member
Thread Starter
Joined
Oct 4, 2018
Messages
54
Thanks, Bernt. I will send via private email, as measurement and other files are not supported types for uploading here in the forum, at least for me. But, I am gratified to see that in Roon, the Procedural EQ filters allow for polarity reversal by channel, so this should be fixable! JCR
 

juicehifi

Audiolense
Staff member
Joined
Feb 5, 2018
Messages
240
Hi jrobbins,

The corrections seem to have the same polarity ... all of them.

I am a bit puzzled about your measurement though.

Here is the left impulse:



42896


And here's the demo measurement:
42897


The combimation of what I see here and the issue you are experiencing are making me unsure of the measurement.

The demo speaker is a conventional 3-way with regular tweeter on flat baffle.

If you have horn-loaded tweeters, it could explain the difference.

But I would like for you to take two consequtive measurements of different duration, e.g. 3 and 10 seconds. Then we can compare. The impulses should be practical identical in the two cases.
 

jrobbins50

Member
Thread Starter
Joined
Oct 4, 2018
Messages
54
Hello, Bernt. I've sent the new 3 and 10 second measurements of my 2.2 system to you by email, along with a photo of my main speakers, a pair of Triangle Celius 202s. Let us all know what you see. Thanks. JCR
 

juicehifi

Audiolense
Staff member
Joined
Feb 5, 2018
Messages
240
Got it. I checked both measurements JCR. And as far as I can see the two are practically identical. The rugged pulse is most likely due to the tweeters' wave guide. And I see nothing wrong with these measurements.

Edit:
Ref your initial issue, with inverted left channel in Roon: I believe an inversion must be going on somewhere, most likely in Roon. So the best fix would be to sort that out.

I hesitate to implement a manual invert feature in Audiolense. It adds complexity and will provide more opportunity for user errors. But it is easy to invert in Audacity: Split, invert, merge and save. Audacity is open source audio editor, btw. Quite easy to getting started with.

(I was confusing your case with the other one.)
 
Last edited:

jrobbins50

Member
Thread Starter
Joined
Oct 4, 2018
Messages
54
Got it. I checked both measurements JCR. And as far as I can see the two are practically identical. The rugged pulse is most likely due to the tweeters' wave guide. And I see nothing wrong with these measurements.

Edit:
Ref your initial issue, with inverted left channel in Roon: I believe an inversion must be going on somewhere, most likely in Roon. So the best fix would be to sort that out.

I hesitate to implement a manual invert feature in Audiolense. It adds complexity and will provide more opportunity for user errors. But it is easy to invert in Audacity: Split, invert, merge and save. Audacity is open source audio editor, btw. Quite easy to getting started with.

(I was confusing your case with the other one.)
Thanks, Bernt. As I noted somewhere above, Roon has a procedural EQ filter to invert polarity by channel. I did that and it indeed fixes the problem. So, it is an easy workaround.

The really great result is that after a remodel of this listening room — replacing a shag carpeting with hardwood flooring, enclosing an alcove, scraping the “popcorn” texturing off the ceilings and installing new furniture — the Audiolense filters have corrected all of the differences and it sounds as good as before the remodel! Hats off to your software! Cheers. JCR
 
Top Bottom