AMP and ARC are independent devices. If you want to make valid comparisons between ARC and your B&O speakers, Group ARC with AMP. In this configuration you are starting with the same data input to both systems.

Note that humans are very poor at comparing audio devices because we usually rate the louder device as being “better”. Unless your ARC and B&O speakers are playing at exactly the same level, the listening jury will be swayed to favor the louder unit.

Thanks buzz.  AMP and ARC are grouped and all the levels are the same.

No they are not. They may have the same number on the volume scale, but that does not mean the volumes are the same on the two different devices.

Thanks John.  Does that also mean the EQ settings aren’t the same across devices?

It is impossible to know with the Amp.  You may have very efficient speakers which are louder at X output, or you may have power hungry speakers which are not as loud at X output.  

Mike Will,

You’ll need to use some sort of sound level meter as you set the Volumes and even this is treacherous if you use a single frequency for your measurement because you may accidentally pick a frequency that one speaker emphasizes. Even if the levels match at that frequency, they may not match very well over a broad range.

Another problem is the Volume adjustment step levels available to the operator are too large in this context.

You can install various acoustic measurement Apps on a phone/pad that will help to approximately match levels. Note that these are uncalibrated Apps. While you can certainly get a “number” for each system, don’t try to compare these numbers with anyone else because phone/App quality and measurement technique vary greatly.

In the end, only you know what sounds “best” -- to you.

As you make your comparisons, keep the equalizers set to “flat” and Loudness OFF.

Thanks again buzz.  Setting the Loudness to OFF made the comparisons between groupings and devices much easier to understand.  I didn’t appreciate that the setting might boost frequencies differently.