Firmware 8.1.2 released

[Jean-Marie]

Now I have a decision to make [...] The purist in me says SAM off until resolved, but on a quick listen last night, the Ref 5 SAM profile actually sounded really good, and given that this is an identical driver to the Blade this might make sense from a technical perspective.  

@Confused, the purist in you has a very valid point  Just let SAM go. I understand those what would like squeeze more bass out of their monitors. But what do you need SAM for, with speakers of such refinement and power as KEF Blades? They can stand for themselves... Also, the bass drivers in KEF Blades II and the Reference series are different (smaller) than in KEF Blades.

The purist in me did not stick with "alternative" SAM profiles for very long!  I listened to one or two tracks on Sunday night that didn't sound quite right on Ref 5 SAM, so this this temporary fix was abandoned.   I'm back on ye olde 7.1.3 now, with classic V1 SAM that works.  And yes you are correct about the driver size, it was kind of in my head that they were the same size and I was just trying different things at the weekend, as much out of frustration that yet another firmware update had gone wrong.

As to SAM, it is true that the Blades have more than enough bass without SAM, but that is only part of the story, the phase and timing corrections do improve the quality of the bass.  With some music SAM can be quite an improvement, which is remarkable because the accuracy of the bass is one of the blades strong points.  When you get these kind of improvements, it's very hard to go backwards sometimes.

As an update, Devialet support have been more than helpful so far, they have confirmed that they can recreate the issue (although I'm not sure how?) and are working on a fix.  Time will tell ......

This is pure speculation from my part, but given what SAM is which is a very thorough and accurate modeling of speakers, I would not be surprised if they can run a simulation of their DSP processing plus ADH plus the model of the speaker, and if what you experience is a flaw in either the SAM profile, the SAM algorithm they should be able to reproduce it with the model and figure out what went wrong.