New Neumann MA 1 - Automatic Monitor Alignment

[Mesh]

The multiple subwoofer procedure is the same in 3 and 4(I asked support). There is just no full GLM4 manual yet because it's still in beta, only a quick start guide.

Yeah I know GLM4 is in beta. You didn't mention that you had spoken with support to validate that this is the same in GLM4 as in GLM3, so thanks for adding that now.

 

[boniek]

If you do daisy chain the subs to play one-channel bass with multiple subs using GLM you actually have to adjust the gain manually, it doesn't even do that automatically. You can find that info on pg74 of the GLM3 manual "Using Multiple Subwoofers."

Yeah, that is trivial one time thing. Still it could be automated because GLM knows how many subs there are.

And no, EQ on each sub individually isn't the same. The fundamental answer to all your questions is "it produces a better result."

Subs are corrected by GLM indivdually to correct for their placement in the room among other things. DLBC needs to verified by independent test if it is anything but snake oil.

The idea is that, if you want as close to a perfectly flat bass response across multiple listening positions as possible, you need 2-4 bass sources and with 4+ you can actually cancel standing waves directly.

This alone is nothing new. You can have pretty much as much subs as you want with GLM too. What is new in Dirac is this allpass AI filter which is magic black box. This needs to be tested.

 

[boniek]

@boniek , Zbigniew?

Nie

 

[Sancus]

This alone is nothing new. You can have pretty much as much subs as you want with GLM too. What is new in Dirac is this allpass AI filter which is magic black box. This needs to be tested.

You don't seem to be really understanding the distinction between EQing each sub individually and EQing for a sum, so I don't think there's any point to further discussion. Bottom line is that GLM's sub correction method is relatively primitive when a single channel of bass and multiple subwoofers are involved. It doesn't account at all for interactions in the combined output. That's all there is to it.

I'd love to see independent tests of DLBC and a comparison with GLM and Audyssey, but the extremely high cost of Genelec subs makes it a painful test to run unless you can get free loaner subs. In any case, calling it snake oil given Dirac's pedigree is a bit ridiculous.

 

[boniek]

You don't seem to be really understanding the distinction between EQing each sub individually and EQing for a sum, so I don't think there's any point to further discussion. Bottom line is that GLM's sub correction method is relatively primitive when a single channel of bass and multiple subwoofers are involved. It doesn't account at all for interactions in the combined output. That's all there is to it.

I'd love to see independent tests of DLBC and a comparison with GLM and Audyssey, but the extremely high cost of Genelec subs makes it a painful test to run unless you can get free loaner subs. In any case, calling it snake oil given Dirac's pedigree is a bit ridiculous.

Fair point. I don't get it and I don't know how where to find more easily understandable info about it. I'm just sceptical until proven otherwise that's all. You should understand it, because after all, we are talking about audio world here on the forum that was founded on strong empirical premises.

 

[CtheArgie]

Nie

I had a Polish father. He sometimes called me Karolek.

 

[jhaider]

The main thing I get out of this thread is Neumann audio measurement microphone available. The software is interesting enough for someone in that ecosystem. For someone who already uses something like a miniDSP SHD, less so.

But I wonder if the microphone calibration is also usable outside the system.

 

[DJBonoBobo]

But I wonder if the microphone calibration is also usable outside the system.

Unfortunately not. At least that's what the dealer says who I ordered it from and who probably talked to the developer about it. You enter the serial number and an id in the auto alignment software, then the software gets the calibration from the internet, I guess. I don't know why you don't just get the cal file like with UMIK. I hope they change that. Looks to me like an unnecessary chicanery, at least I can't see any reason for it. However, the microphone is probably good enough for most measurements even without calibration.

 

[thewas]

However, the microphone is probably good enough for most measurements even without calibration.

Probably only for 90° as almost all such mics have an inherent peak around 10 kHz at 0°:
https://www.hifi-selbstbau.de/grund...0-1000-mikrofonkalibrierungen-eine-uebersicht

 

[ernestcarl]

Unfortunately not. At least that's what the dealer says who I ordered it from and who probably talked to the developer about it. You enter the serial number and an id in the auto alignment software, then the software gets the calibration from the internet, I guess. I don't know why you don't just get the cal file like with UMIK. I hope they change that. Looks to me like an unnecessary chicanery, at least I can't see any reason for it. However, the microphone is probably good enough for most measurements even without calibration.

Somewhat related: I was wondering if the Sonarworks mic calibration is also exclusively locked to their software... does anyone know?

 

[thewas]

Somewhat related: I was wondering if the Sonarworks mic calibration is also exclusively locked to their software... does anyone know?

You can freely download the 0° calibration to use it in programs like REW but the 30° calibration which is used from the Sonarworks DRC software is a binary format which can only be read by it.

 

[DJBonoBobo]

I received my MA 1 today and tried it quickly. I post my first impressions, but have to digest them a bit and try it out some more in the next days.

This is what the calculated target for my room (14 sqm, very dry, 1.6m distance, KH310 + 750 DSP) looks like. I didn´t modify it today. As you can see it is not "flat" and it is possible to modify it.

And this is what i got from REW:
Green: Estimated in room response KH310 only for reference, measured by Amir.
Red: The best i could achive by manually optimizing over the last months. Average of MMM-measurements (L + sub) + (R + Sub) with room EQ and settings described here.
Blue: Average of MMM-measurements (L + sub) + (R + Sub) after Auto Alignment with the MA 1 (for this REW measurement i used the UMIK for comparability)

Obviously i have to investigate the grand canyon between 80 und 100 Hz. In the past this was influenced by the delay of the sub and i had to chose between closing this gap and a smoother response between 100 und 200 Hz. So there is still optimizing to do for me...

Edit: 1.6m distance, not 16m - thanks Mesh.

 

[Mesh]

I received my MA 1 today and tried it quickly. I post my first impressions, but have to digest them a bit and try it out some more in the next days.

This is what the calculated target for my room (14 sqm, very dry, 16m distance, KH310 + 750 DSP) looks like. I didn´t modify it today. As you can see it is not "flat" and it is possible to modify it.

View attachment 97835

And this is what i got from REW:
Green: Estimated in room response KH310 only for reference, measured by Amir.
Red: The best i could achive by manually optimizing over the last months. Average of MMM-measurements (L + sub) + (R + Sub) with room EQ and settings described here.
Blue: Average of MMM-measurements (L + sub) + (R + Sub) after Auto Alignment with the MA 1 (for this REW measurement i used the UMIK for comparability)

View attachment 97836

Obviously i have to investigate the grand canyon between 80 und 100 Hz. In the past this was influenced by the delay of the sub and i had to chose between closing this gap and a smoother response between 100 und 200 Hz. So there is still optimizing to do for me...

Floor to ceiling reflection from the first axial mode is that canyon. Let me guess about 8 ft (2.40m) ceiling height? Do you have a cloud?

 

[DJBonoBobo]

Floor to ceiling reflection from the first axial mode is that canyon. Let me guess about 8 ft (2.40m) ceiling height? Do you have a cloud?

Yes, 100 x 100 x 12 cm Basotect (plus 10 cm gap). Room is 475 x 300 x 255 cm, and i think it is a whole pack of modes there, not only first axial. But i do not want to capture the thread with discussing my room here. For me it is just interesting to see that the software obviously chose a different delay and/or phase setting than me.

 

[Mesh]

Yes, 100 x 100 x 12 cm Basotect (plus 10 cm gap). Room is 475 x 300 x 255 cm, and i think it is a whole pack of modes there, not only first axial. But i do not want to capture the thread with discussing my room here. For me it is just interesting to see that the software obviously chose a different delay and/or phase setting than me.

Oh totally agree, and it is the same issue I have. That makes what you can achieve extra interesting to me (and probably many more!). I am undecided about Neumann or Genelec so please keep us updated.

 

[DJBonoBobo]

Oh no! They should have written a warning on it somewhere: Not compatible with the iPad-App! I started the app and now the auto alignment is lost. So i have to make all the measurements again. (I had already suspected this, but would have preferred a positive surprise).

Also, the last alignment of the app has not been restored correctly, but something went wrong. Seems like the sub is on a wrong level...

Red: Last setting of the app
Turquoise: What remained after i started the app again after the Auto Alignment. Looks like a bug...

So it seems the software has some first day issues. I hope they will sort them out soon and i will get my system back to work somehow...

 

[hyperplanar]

I received my MA 1 today and tried it quickly. I post my first impressions, but have to digest them a bit and try it out some more in the next days.

This is what the calculated target for my room (14 sqm, very dry, 16m distance, KH310 + 750 DSP) looks like. I didn´t modify it today. As you can see it is not "flat" and it is possible to modify it.

View attachment 97835

And this is what i got from REW:
Green: Estimated in room response KH310 only for reference, measured by Amir.
Red: The best i could achive by manually optimizing over the last months. Average of MMM-measurements (L + sub) + (R + Sub) with room EQ and settings described here.
Blue: Average of MMM-measurements (L + sub) + (R + Sub) after Auto Alignment with the MA 1 (for this REW measurement i used the UMIK for comparability)

View attachment 97836

Obviously i have to investigate the grand canyon between 80 und 100 Hz. In the past this was influenced by the delay of the sub and i had to chose between closing this gap and a smoother response between 100 und 200 Hz. So there is still optimizing to do for me...

Thanks for being the first one to try it out and report back! I'm surprised and a bit disappointed the software didn't choose a better delay setting for that 80-100 Hz suckout. Also, am I understanding the graph correctly in that the software decided to boost the treble by around 1.5 dB? And what are your subjective impressions of the result versus your manual alignment?

 

[DJBonoBobo]

You are welcome! But this is really only a first impression and not a review. Please be careful to draw any conclusions. Maybe I didn't do everything right on the first try or the result might be very specific for my room. I don´t know about "better" delay - the part between 100 and 200 Hz is clearly better with the auto alignment, and i am not sure what causes the dip.

See for example how the R (+sub) + L (+ same sub) measurements look like (auto alignment above, my own settings below), each MMM. This looks very strange, and maybe something just went wrong. I have to investigate in the next days and maybe i can get rid of the dip without sacrificing the better 100-200hz area.

Too early for subjective impressions, not only because i destroyed my results by starting the app after a few minutes and have to do it again anyway. I even forgot to measure both channels together. But i understand the treble is boosted, too, indeed.

 

[Mesh]

You are welcome! But this is really only a first impression and not a review. Please be careful to draw any conclusions. Maybe I didn't do everything right on the first try or the result might be very specific for my room. I don´t know about "better" delay - the part between 100 and 200 Hz is clearly better with the auto alignment, and i am not sure what causes the dip.

See for example how the R (+sub) + L (+ same sub) measurements look like (auto alignment above, my own settings below), each MMM. This looks very strange, and maybe something just went wrong. I have to investigate in the next days and maybe i can get rid of the dip without sacrificing the better 100-200hz area.

Too early for subjective impressions, not only because i destroyed my results by starting the app after a few minutes and have to do it again anyway. I even forgot to measure both channels together. But i understand the treble is boosted, too, indeed.

View attachment 97985

What distance are you measuring at? It looks like the top range is more linear with the auto alignment and if you are listening at near or mid field, I would think that is what they assume is desired. EDIT: Ah, you meant 1.6m distance earlier. I looked at that 16m you mentioned first and wondered.

Good to see you have one channel trying to work to cover that gap. Even in this regard, I have the exact same situation as my left channel is fighting to compensate for what both the sub alone and the right channel alone is struggling with. In my case, that is a diagonal standing wave caused by one of my back wall corners being diagonal with the door into the room rather than 90 degrees. With channel separated EQ, I can get the sum to even out more but I am not sure I actually perceive it as better. I really need to do some blind tests to know for sure my brain isn't making stuff up that the ears aren't actually hearing...

 

[onion]

Fair point. I don't get it and I don't know how where to find more easily understandable info about it. I'm just sceptical until proven otherwise that's all. You should understand it, because after all, we are talking about audio world here on the forum that was founded on strong empirical premises.

Eqing individual subs separately ignores any interference pattern between the sound waves generated by them; and will miss any peaks and nulls resulting from this interference. When playing music normally, both subs play simultaneously and the interference that was missed by GLM rears its ugly head.

I still use GLM. The subs are happily out of phase with each other when phase-aligned with the mains, and the bass is fairly even over the listening area (sofa) - this is more by luck than anything GLM does. If I get a third sub, I'd play around with its settings manually and judge the results with my ears (and maybe REW).