A measured response: the KEF LS50 + Devialet’s SAM


Phono stage. DAC. ADC. USB-S/PDIF converter. Network streamer – five additional roles playable by the Devialet Expert 200, which first and foremost is a loudspeaker amplifier. But even here there’s a wrinkle. The French manufacturer have added a little (optional) special sauce to the mix.

The Expert 200 comes loaded with SAM (“Speaker Active Matching”) which harnesses the power of DSP to ensure the loudspeaker’s output signal more closely follows the input signal. To do this, the loudspeaker must first be measured; a process that is carried out either at Devialet’s Paris headquarters or on the road using the company’s portable SAM Labs kit (which we first saw at Munich High-End 2015).

This Devialet video pulls back the curtain on the measurement procedure:

Once measured, the loudspeaker is declared ‘SAM Ready’ and is listed on the Devialet website. To date, over 600 loudspeaker models, many from some of the world’s most prestigious manufacturers, have been measured and modelled.

For the home user with a SAM Ready loudspeaker connected to an Expert amplifier, SAM is activated via Devialet’s online configuration tool: the loudspeaker is selected from a drop-down list of SAM Ready models, the amended config file downloaded to an SD card and then inserted into the Expert’s onboard memory card reader.

An advanced options panel brings greater flexibility. Dialled in using the Expert’s rather fancy rotary remote, the amount of SAM applied to the loudspeaker can be specified according to taste.

Each loudspeaker’s SAM profile also includes an excursion protection mechanism that prevents any mid/bass driver from exceeding its (measured) limit. This is invoked even when SAM is activated but set to zero. For KEF’s LS50, it’s 5mm.


Quoting from my own 6moons review of the Expert 200:

“I’ll confess to not liking the full force of SAM when applied to the KEF [LS50] monitors. It’s as if too much is being asked of the speaker all at once, especially in the lower frequencies. Picture a fat man trying to squeeze through a very narrow door way. That’s the sound of a SAM’d-up LS50 rendered into visual form.”

“This commentator’s favoured sweet spot arrived at a very precise 57% [of SAM]. Even with the same speakers, yours might be different. With different speakers, it’s a different deal all over again. Devialet paves the way for each to find their own.”

Clearly this is not a summary dismissal of SAM’s efficacy but a matter of personal taste. Arrangements are already in the works to subjectively assess SAM as applied to another well known loudspeaker brand.

In the meantime, with DEQX PreMATE+ loudspeaker correction already tried and tested with the LS50, an opportunity presented: how does SAM correction – applied in full – alter the KEF standmount’s frequency response?


The LS50’s impulse response was measured using DEQX’s Calibration software and a microphone.. From there, its frequency response was derived. That’s the blue line in the graph below. This measurement process was then repeated with SAM applied and turned up to 100%. That gave us the red line below (which was bumped up by 2db to match blue’s levels).


In this piece our focus is on how the uncorrected LS50 compares to the SAM-corrected LS50. The DEQX hardware and software was used only to conduct the measurements and none of the PreMATE+’s own loudspeaker correction was applied. Confused? Catch up here.

Above 200Hz, the graph shows that SAM correction makes no changes to the LS50’s frequency response. Below 200Hz, SAM boost’s the LS50’s bass response and is entirely consistent with Devialet’s claim that SAM correction takes the LS50’s bass response from 44.8kHz down to 28.6kHz.

Additionally, SAM appears to elevate slightly the LS50’s response between 60Hz and 150Hz.


How does SAM influence the LS50’s group delay?


Above 200Hz [see footnote], the red line (SAM corrected LS50) and the blue line (uncorrected LS50) appear to follow a similar trajectory – group delay remains unchanged by SAM.

The 30Hz – 40Hz window excepted, below 200Hz our red and blue lines diverge. This suggests that SAM adds some group delay to the lower frequencies. On this, DEQX’s Alan Langford reckons that it’s probably the result of SAM’s bass boost. Whether or not the human ear is sensitive enough to pick up on small increases to group delay in the bass region remains debatable.

Back in the listening room, we should remind ourselves that SAM isn’t compulsory. The Expert owner who doesn’t find joy in the SAM-corrected sound can switch it off via the remote or decommission it entirely in the online Le Configurateur. The KEF LS50 sound terrific when powered by the Expert 200 but with no SAM applied – that’s my preference. Others might prefer SAM’s extra dollop of bass.

Much like a tone control, Devialet’s SAM gives us choice. We can enjoy the scenic route or the bypass. There is no right way or wrong way. There is no absolute sound.

Further information: Devialet | KEF | DEQX


Footnote: the 200Hz divider here has no connection to the 200Hz used to gate the DEQX PreMATE+’s own loudspeaker correction; purely a coincidence.

Written by John H. Darko

John lives in the NOW + HERE = NOWHERE. He derives an income from the ad revenues of DAR. John is also an occasional staff writer for Stereophile, 6moons and TONEAudio.

Twitter: DarkoAudio
Instagram: DarkoAudio
Facebook: DAR


Leave a Reply
  1. We cannot deny it anymore.. DSP is the future of high-end audio. FPGA dac, DEQX, DDFA, MQA, digital motional feedback, digital room correction, immersive audio, etc…

  2. Fantastic article, John. Your writing is amazing. How you can explain such complicated things so clearly is beyond me. It must take you a long time to write these articles. Many thanks for your efforts and for a very interesting and informative article.

  3. So reading between the lines, looks like you personally prefer DEQX speaker correction to SAM? Look forward to your thoughts on the DEQX room correction functions.

    • You read correctly. DEQX’s group delay correction really brings quite a bit to the table.

  4. no, there is no absolute sound. or sounds. we each have different priorities & different rooms, even if we all had the exact same speakers. (and i do have a set of ls50s, primarily because they present the human voice well, imo). please keep up the investigations of speaker correction & active speakers — these areas are virtually ignored by your ‘more-established’ peers. thanks! best, david