New Neumann MA 1 - Automatic Monitor Alignment

[audafreak]

Interesting. Your target curve seems to be a bit more sloped than mine - i guess because your LP is 2,5 m while mine is 1.6 m.
Also nice to see how much better a bigger room is (mine is only 3 x 4.7 meters).

Yes, I see the same, My LP is 1.9m, room a bit bigger 3.5x5.5m but monitors are along the longer side 1m from front wall, wall with 7cm dumping panel behind me - very bad position especially for bass response but nothing else is possible.
@v3kt0r Do you have also screenshot with not activated calibration? Default system response. Aligned seems really perfect on bass side comparing to mine.
I had to raise around 20 Hz a bit.
Here is mine:
before

after

I set lower SPL in monitor because I can have higher signal out of DAC (lower noise, higher resolution) and higher signal in KH750 AD conversion (again higher resolution).
And it is also safer if I send full signal to monitors by accident.
Are there any reasons not to do it so?

 

[DJBonoBobo]

I set lower SPL in monitor because I can have higher signal out of DAC (lower noise, higher resolution) and higher signal in KH750 AD conversion (again higher resolution).
And it is also safer if I send full signal to monitors by accident.
Are there any reasons not to do it so?

No, it is actually a good idea. I plan to set it lower next time i recalibrate (when my second KH750 arrives). I use a digital connection now and 100 dB is way too loud when i accidentally switch the interface to 0 dB. Did it once while playing around with new RME toys and hopefully never again...

Interesting related article by @MC_RME about input- and output-levels: https://forum.rme-audio.de/viewtopic.php?id=25399

 

[thewas]

What makes we wonder when I see the screenshots posted here is some significant EQing also above transition frequency, which should deform the neutral direct sound of the monitors...

 

[ernestcarl]

What makes we wonder when I see the screenshots posted here is some significant EQing also above transition frequency, which should deform the neutral direct sound of the monitors...

The one above looks pretty wide Q... so other than a broad shift in level above 1.5 kHz, the linearity of the direct sound should largely remain intact.

 

[DJBonoBobo]

To me it sounds more neutral and "correct" after the calibration. When i tried Dirac Live for the whole spectrum i never had that feeling. So i think they know what they are doing.
Would be interesting to learn more about the algorithms they use and how they "decide" which corrections are made.

 

[ernestcarl]

To me it sounds more neutral and "correct" after the calibration. When i tried Dirac Live for the whole spectrum i never had that feeling. So i think they know what they are doing.
Would be interesting to learn more about the algorithms they use and how they "decide" which corrections are made.

The problem with Dirac is that the software has no idea about the directivity of the particular speakers being "corrected". The dedicated Neumann software probably knows this (based on the speaker model -- and the speakers are all very consistent) and so likely it takes that information into account, for sure.

 

[audafreak]

What makes we wonder when I see the screenshots posted here is some significant EQing also above transition frequency, which should deform the neutral direct sound of the monitors...

I also think it is just tweeter level matching. No EQ.

 

[thewas]

The one above looks pretty wide Q... so other than a broad shift in level above 1.5 kHz, the linearity of the direct sound should largely remain intact.

I also think it is just tweeter level matching. No EQ.

Still its a deformation of the direct sound when boosting just the tweeter/highs compared to the mids.

 

[abdo123]

Still its a deformation of the direct sound when boosting just the tweeter/highs compared to the mids.

the highs were flat before correction and remain almost unchanged after correction, or am I missing something?

 

[thewas]

the highs were flat before correction and remain almost unchanged after correction, or am I missing something?

As I wrote above the highs get boosted compared to the mids, you can't just view them independently.

 

[abdo123]

The problem with Dirac is that the software has no idea about the directivity of the particular speakers being "corrected". The dedicated Neumann software probably knows this (based on the speaker model -- and the speakers are all very consistent) and so likely it takes that information into account, for sure.

there isn't really any information to take into account when the speaker is super linear like most of Neumann's.

I got the best results with Dirac when I limited the correction range to the frequency where the directivity cease to be linear across my listening window. Which was 4KHz at the time.

 

[abdo123]

As I wrote above the highs get boosted compared to the mids, you can't just view them independently.

if it makes you feel better you can picture it as bringing the mids and lows down because there are no corrections on-going above 1KHz.

 

[ernestcarl]

As I wrote above the highs get boosted compared to the mids, you can't just view them independently.

While I agree that the Neumann probably does not truly need a boost above 1.5 kHz, as far as speaker corrections go, the effect (I suspect) likely is more that of a mild broad HF tone control. I think applying a -1 dB negative high shelf PEQ filter after-the-fact would mitigate this"deformation" of the direct sound by a good bit.

 

[thewas]

if it makes you feel better you can picture it as bringing the mids and lows down because there are no corrections on-going above 1KHz.

1,5 kHz is significantly above transition frequency where the room dominates, so it remains nothing different than a high shelving filter.
In the end loudspeakers, rooms and room "corrections" are all compromises as we do not perceive only the direct sound or only the PIR/SP but a mix of both, so I can even understand such an approach, still find it interesting to discuss as there are many different views on it.

 

[ernestcarl]

there isn't really any information to take into account when the speaker is super linear like most of Neumann's.

Some people have complained about harshness in the highest frequencies, say, above 10 kHz; and the dip in the power response. Using the anechoic curves, one could very well make some very minor (non-destructive) EQ -- that's well-informed by the available spinorama data.

*Q of 4 is a bit high there, but the boost is low enough in level that it's also hardly an improvement (if you can call it that). Mainly it's there only to satisfy my audiophilia nervosa at this point.

From Neumann:

 

[audafreak]

Still its a deformation of the direct sound when boosting just the tweeter/highs compared to the mids.

I don't think it is boost. 0dB line is relative value so in fact highs are unchanged, middle is mostly reduced by EQ, bass phase and EQ corrected. This is how I would do it. Does it make sense?

 

[abdo123]

Mainly it's there only to satisfy my audiophilia nervosa at this point.

I'm glad you called it the way it is because i hardly think the differences would be noticeable.

1,5 kHz is significantly above transition frequency where the room dominates, so it remains nothing different than a high shelving filter.
In the end loudspeakers, rooms and room "corrections" are all compromises as we do not perceive only the direct sound or only the PIR/SP but a mix of both, so I can even understand such an approach, still find it interesting to discuss as there are many different views on it.

it depends honestly on your school of thought, some think these microphones are only representative of what our ears hear when the room is minimum-phase (the sound power portion of the graph above), some think if the Direct sound is consistent with early reflections then you can correct the entire response with no drawbacks.

 

[abdo123]

I don't think it is boost. 0dB line is relative value so in fact highs are unchanged, middle is mostly reduced by EQ, bass phase and EQ corrected. This is how I would do it. Does it make sense?

0dB is usually what 1KHz is, in my experience.

 

[audafreak]

So it is relative. If you decrease SPL on 1kHz because of EQ like in the pictures above then you can have feeling the rest is boosted ...

 

[thewas]

I don't think it is boost. 0dB line is relative value so in fact highs are unchanged, middle is mostly reduced by EQ, bass phase and EQ corrected. This is how I would do it. Does it make sense?

In the end both lead to the same thing, although in this case I rather think the internal treble shelving filter was used as such broadband filter does not exist for the mids.