Sidebar 1: A Fly in the Ointment?
I decided to use the same music files to compare the sound of the Arcam FMJ SR250 with Dirac Live room correction to the Dirac Live filtering in my server. It was a fair fight, as long as I set the server to downsample to 24/96 and used the same target curves for both implementations. In this face-off, the Dirac Suite on my server seemed marginally but consistently more successful than did the Arcam implementation. Yeah, the microphone positions couldn't possibly have been exactly the same, but the purpose of the multiple positions is to create a statistically representative sample. The two sets of Dirac graphs looked quite similar, except for the region below 100Hz. Even there, I could find no meaningful differences.
Via my server, Dirac Live's filters are applied to the main speakers and subs after application of bass management—but Arcam puts Dirac Live measurements and correction into the signal-processing path before the application of bass management. I came across a similar arrangement in miniDSP's original nanoAVR 8x8, where, as I said in this column in the November 2014 issue, it "meant that the bass being rerouted from each of the main channels was equalized as if it were being reproduced by that speaker. In fact, after the subsequent bass management, it was actually being reproduced by a subwoofer—a quite different speaker in a different position in the room. As a result, the wrong filters were imposed on all the rerouted bass, and that compromised the success of the entire EQ effort."
Arcam told me that their arrangement can work as well as or better than the standard arrangement if all measurements are made with proper bass management in place. I agree, in theory, but there's nothing in the FMJ SR250's manual or in their notes for the Dirac Live app that tells you how to do this. In fact, when I tried to do it on my own, I found that, even when I turned on bass management before taking measurements, the test signals for the main channels did not affect the subs or change the results.
For me, this was a teeny-tiny issue. Had I not had the opportunity to compare it with my other implementation of Dirac Live, and had I not stumbled on a discussion of this matter online, I would not have criticized the sound I got from the FMJ SR250. Arcam is aware of all this, and is preparing a software update that is now in Beta testing (footnote 1).—Kalman Rubinson
Footnote 1: Now that I've finished this review, packed up the FMJ SR250, and sent it on to be measured by John Atkinson, I see that, to resolve these issues, Arcam has released new software for both the SR250 and Dirac Live (v.4.17). Apparently, Dirac Live processing still precedes bass management, but the measurement configuration should account for that.
Footnote 1: Now that I've finished this review, packed up the FMJ SR250, and sent it on to be measured by John Atkinson, I see that, to resolve these issues, Arcam has released new software for both the SR250 and Dirac Live (v.4.17). Apparently, Dirac Live processing still precedes bass management, but the measurement configuration should account for that.
