Devialet Chat

Sam 1, Magico Q1
I've spent several hours today experimenting with SAM profiles on speakers that closely mimic my DIY spec speakers. When I built these speakers I intentionally went after the Magico Q1 specification. My design goal was to "at least" match it and top it wherever possible and I succeeded in every aspect. Matched frequency response exactly, beat it in sensitivity. Matched it in internal volume and beat it in mass/weight 60 vs 68lbs.. While the Q1 features a 7" Nanotec woofer/driver and a beryllium tweeter my NADA's feature a 7" ScanSpeak Illuminator underhung paper woofer with neodymium magnets and largish voice coil with 9mm excursion that more closely mimics a 10" woofer. Both deliver 32hz. Tweeters between the two speakers while different manufacturers are both ~1" beryllium mine again ScanSpeak Illuminator and Magico's is unknown to me. So of course the Magico Q1 was the first SAM profile I copied but not the first I tried. But before I get into results let me bring you up to speed on the 2nd SAM profile I tried first and why.

SAM 2: Crystal Cables Minissimo
The reason I opted to try this profile first will be immediately obvious, ie; it uses exactly the same woofer/tweeter I use, is a reflex design of approximately the same internal volume. Simply put, it is hands down the closest match to my speakers as that relates to actual components used and I felt safest to use it as my jumping off point.

Any guesses to how this is going to shake out? I think you're going to be surprised. I certainly was. Gape-jawed is a more apt description.

I'll save the best for last and start with SAM 2: Biggest perception was a smaller sound stage followed by an overall smaller sound. As if I had turned watts down from 200 to a buck fifty or so and every singer and instrument had shrunk by 25% or so. Still that highly detailed Devialet presentation sparkling highs and smooth, articulate mids and taut utterly controlled bass. To achieve normal listening levels I had to bump the volume 20% from -30/-32 to -23/-24. No accompanying weirdness whatsoever. No static, hiss, crackling or pops etc nothing threatening or ominous sounding. Totally black noise floor. In fact zero negative or extraneous noise. 100% as refined as when running no SAM... just smaller. When turned up to compensate for its diminutive SQ it got louder but still perceived as smaller with less power. It lacked richness and fullness.

Now here's the treat: The SAM profile for the Magico Q1 is a keeper! But I'll still continue to experiment with this as its just too friggin easy! But Q1 made my 2 week old 200 sound better than ever! I mean a whopping increase in everything positive. Lows, mids, highs, sound stage etc although noise floor simply cannot get any blacker. But its more lively, transparent, more refined tone/texture, delicacy with increased realism and transparent-to-source!

I only allowed 30 minutes audition for each profile tho I'm back on Q1 as I write this. Oh and all this using 7.1.3 beta FW. I'm gonna be real disappointed in myself if after trying a a couple dozen different profiles that I don't find a really awful one in the lot. I somehow thought if I chose a wrong one my XO's would short & melt or I'd shred a woofer or shatter a tweeter. None of that happens and so far nothing even sounds really bad... just impish in the case of my 1st attempt. Can't believe I waited so long to try this! But I'm going out for a ton-O-SD cards tomorrow! I'm gonna play with every speaker's SAM profile that's anywhere near the same zip code! Its just good clean fun! With the simplicity of doing this I can even see using this as a tweak for a specific desired result, ie my rock SAM, jazz SAM, orchestral SAM, bad recording SAM, great recording SAM etc. And even if that didn't pan out I'm having the time of my life just playing!!

Hi Manoet,

Excellent post!. I and a few others have tried SAM profiles on our non-SAM speakers, but I don't think anyone has put the process and results together as well as you have here. As you did, I tried profiles that were for speakers that have similar drivers and cabinet specifications; again, your Q1 and Minissimo profiles are probably closer to your speakers than for most of the rest of us. I believe that my personal best -match was using the ProAc Tablette Anniversary SAM profile on my original Tablettes. But I did find that an 'inappropriate' SAM application does help each of my speakers. I do, however, turn the percentage down to somewhere between 0 and about 30%. The bass clarity improves, I believe, or some of the bass spectrum and the midrange, in each case, definitely feels 'freer' and less congested. It is curious that even at 0% SAM, there is an audible difference. Again, in the midrange more so than in the bass.

If you sniff out the other threads or posts on this idea, you will find that we all understand that this sort of 'good clean fun' is frowned upon. And I understand that some damage could result, depending on the mismatch between the profile and the speaker, and the welly given the ole' volume knob. But giving various speaker profiles is a good diversion, and gets some interesting results. And all this fussing without buying any new equipment!

I think the only real risk of damage is if you try the profile of a hugely different transducers on your speaker. for instance if you have a 7'' 80W RMS woofer with 6mm of excursion and you try the profile of a speaker with a 12'' 150W RMS, you could destroy your speaker.

Similarly, be careful with very long throw woofers if yours is much lower excursion.

But if you stay between speakers having similar mechanical and power characteristics, I really think that the only risk is less accurate reproduction.

Thanks guys and I hope I've explained this effort in such a way that your and my concerns have been demonstrated in both word and deed. And ideally everyone here understands this is no attempt at making a bookshelf sound like the proverbial 800lb gorilla (floor-stander). That its merely exploratory in nature and in keeping within prudent and well thought-out practices. I totally agree with you guys on researching potential SAM profile speakers and comparing them to your particular speaker and staying within close proximity of each other spec-wise. Of course that's easier to do if your speaker manufacturer uses off-the-shelf drivers from the likes of Accuton, Fostex, Morel, Scanspeak, Seas and others as their specification list is always more extensive, comprehensive and widely available than manufacturers using proprietary drivers who limit their public disclosure to minimum required to market a speaker. Sure they'll list freq response, sensitivity, ohms etc... they have to! But they never list things like linear excursion and total mechanical excursion or other info that would be helpful to us in the unique position of finding a potential SAM match that works better than SAM "off" for unSAMable speakers. And that's the ONLY reason I can find to even consider this pursuit. Ya gotta know I'm not doing this to find less than I had originally. But the exceedingly good luck I had yesterday while appreciably rare drives me on. I realize, against all odds that I probably found Valhalla yesterday but until I try other profiles I remain optimistic that at least the promise of something even better remains. Only thing I know for sure; I'll never know unless I try!

Later this morning I'm off to snag a dozen or so 2GB SD cards as I'm woefully short to do this right and be able to attempt any kind of qualitative side-by-side A/B comparisons. I do love how configurator allows on-the-fly configuration changes in the Devialet with only a 15 second pause in music when inserting the new SD card without powering down. The way you can swap configurations mid-track and keep going... how cool is that!?!

The one thing I'd like to specifically address is Jean-Marie's parting comment; "I really think the only risk is less accurate reproduction" while I contend there's also the risk of MORE accurate reproduction! If "less" was the bona fide, guaranteed outcome in every SAM profile swap then this effort would be tantamount to Einstein's (perhaps mis-attributed) theory of insanity.

The likelihood of your drive units in your cabinets having the same bass frequency and phase response as any SAMmed speaker is pretty well zero, so whilst you may get marked effects and may enjoy them you will emphatically not be getting correct frequency or phase response from you DIYers, just some bass boost effects which may or may not be fun. You will not increase accuracy in any defined way. Jean-Marie is technically quite correct.

This is also the case using SAM for any speakers than those for which the SAM profile was created, you may note that Devialet even do two SAM profiles for KEF reference speakers with 2 alternative reflex port lengths so SAM done correctly must be pretty precise.

(11-Jun-2015, 13:31)Manoet Wrote: [ -> ][...]
The one thing I'd like to specifically address is Jean-Marie's parting comment; "I really think the only risk is less accurate reproduction" while I contend there's also the risk of MORE accurate reproduction! If "less" was the bona fide, guaranteed outcome in every SAM profile swap then this effort would be tantamount to Einstein's (perhaps mis-attributed) theory of insanity.

Call me pessimist or skeptical or scientist, but I leans towards F1Eng's side. My understanding of SAM, and I have spoken with Pierre-Emmanuel Calmel several times about SAM, is that every aspect of a speaker I can think of (from electrical to acoustic charging to mechanical to thermal, phasing, etc.) is being modeled. 

So I think there is no chance that an other profile will sound as accurate as a proper SAM profile for this speaker.

Now it is not impossible that although not ideal, a profile from an other speaker would move that speaker closer to the ideal accuracy than without, but that is still very hazardous to me, especially when you hear how seemingly similar speakers sound differently.

Filtering is a very important component of a speaker and is something that is modeled and compensated by SAM, and this is not easy compare between two speakers without extensive measurements.

Just my 2 cents....

(11-Jun-2015, 13:31)Manoet Wrote: [ -> ][...] Sure they'll list freq response, sensitivity, ohms etc... they have to! But they never list things like linear excursion and total mechanical excursion or other info that would be helpful to us in the unique position of finding a potential SAM match that works better than SAM "off" for unSAMable speakers. [...]

By the way, some do. for instance the woofer of my AperturA Armonia is a H1571-08 U18RNX/P from SEAS. if you follow the link you see that the liner coil travel is 6mm and maximum is 12mm. Not surprisingly, if you go the the AperturA Armonia SAM page, you see that they are protecting the woofer at 5mm excursion.

Then I guess I'll just have to remain tormented by an inappropriate SAM profile with dramatically "better" SQ , nowhere near just 'bass boost' than the previous non-SAM profile that is patently "correct" for any/every non-SAM'd speaker. Shouldn't be too hard ;-)

(11-Jun-2015, 15:53)Manoet Wrote: [ -> ]Then I guess I'll just have to remain tormented by an inappropriate SAM profile with dramatically "better" SQ , nowhere near just 'bass boost' than the previous non-SAM profile that is patently "correct" for any/every non-SAM'd speaker. Shouldn't be too hard ;-)

I'm not trying to discourage you and did the same before my speakers were SAMed... but it was not conclusive for me.

Jean-Marie

Thanks and no worries Jean-Marie, I'll keep going tho I did get a chuckle from f1's implication that I was simply getting some 'bass-boost' effect and nothing more. As a nigh on 70 year old man who's been deeply involved in audio for just shy of 5 decades and doesn't use so much as a sub with my bookshelf speakers, or run my Devialet eq settings higher than zero/zero I find it insulting he might think mere bass boost or some 'loudness button' effect would blow a lot of cool breezes up my skirt. Silly man!