Posts: 1,151
Threads: 14
Joined: May 2014
Reputation:
26
03-Aug-2016, 16:16
(This post was last modified: 03-Aug-2016, 16:16 by Rufus McDufus.)
I used to have Linn equipment (Klimax DS) and exactly the same thing used to happen - new firmware would come out with (say) only functional changes or a bug fix or two, nothing to do with sound quality, yet many people would report SQ differences. My opinion (looking at it from a software point of view which was my trade a while back, though usually C/C++ as opposed to the FPGA bytecode or whatever runs on the Dev) is that with a different version even minor changes will have a slightly different run profile on the processors in the Devialet, and perhaps slightly different power requirements. It would seem unlikely that these would be significant enough to make a difference but it's certainly a possibility. Another factor could be how the firmware is compiled down. We don't even know how Devialet manage their firmware build environment but I'd hope they keep it extremely consistent between builds. Different compilation flags between builds could affect how the firmware runs on the processor.
Posts: 310
Threads: 0
Joined: Jun 2015
Reputation:
15
(03-Aug-2016, 16:16)Rufus McDufus Wrote: I used to have Linn equipment (Klimax DS) and exactly the same thing used to happen - new firmware would come out with (say) only functional changes or a bug fix or two, nothing to do with sound quality, yet many people would report SQ differences. My opinion (looking at it from a software point of view which was my trade a while back, though usually C/C++ as opposed to the FPGA bytecode or whatever runs on the Dev) is that with a different version even minor changes will have a slightly different run profile on the processors in the Devialet, and perhaps slightly different power requirements. It would seem unlikely that these would be significant enough to make a difference but it's certainly a possibility. Another factor could be how the firmware is compiled down. We don't even know how Devialet manage their firmware build environment but I'd hope they keep it extremely consistent between builds. Different compilation flags between builds could affect how the firmware runs on the processor.
I will give credit to Devialet for the fact that they let us move forwards and backwards at will. Certain other companies make it difficult or impossible to go back to an earlier version of firmware once a newer version has been installed.
NUC i5/ROCK/Roon Lifetime => Original d'Atelier CoreInfinity #066 => Transparent MusicWave+ =>Wilson Sophia 3
Posts: 581
Threads: 4
Joined: May 2014
Reputation:
20
(03-Aug-2016, 16:16)Rufus McDufus Wrote: I used to have Linn equipment (Klimax DS) and exactly the same thing used to happen - new firmware would come out with (say) only functional changes or a bug fix or two, nothing to do with sound quality, yet many people would report SQ differences. My opinion (looking at it from a software point of view which was my trade a while back, though usually C/C++ as opposed to the FPGA bytecode or whatever runs on the Dev) is that with a different version even minor changes will have a slightly different run profile on the processors in the Devialet, and perhaps slightly different power requirements. It would seem unlikely that these would be significant enough to make a difference but it's certainly a possibility. Another factor could be how the firmware is compiled down. We don't even know how Devialet manage their firmware build environment but I'd hope they keep it extremely consistent between builds. Different compilation flags between builds could affect how the firmware runs on the processor.
Compilers are also improved and will create more optimized code. Simply using a new compiler version can already bring SQ improvements.
Posts: 842
Threads: 22
Joined: Dec 2014
Reputation:
14
(03-Aug-2016, 14:47)neXus 6 Wrote: This thread is indicative of a potential problem in DV's approach. A change to the OS (for want of a better term) should have no material impact on the sound structure. The upgrade might be useful for
- fixing bugs
- widening the scope of compatible devices
- dealing with a change in technology (e.g. USB2 to 3)
- some improvement in operating function whether that is messing about with the display, controller or even some operating efficiency (bypassing weak component, heat management, whatever) = change in quantity (not quality) of output
The whole purpose of the configurator is to take your base station (OS) and apply/tweak the functionality you need. This is like setting up a drilling rig on rock and then attaching all your tools for microscopic accuracy then someone, overnight, moving the rig to marshland. Some of us will get (or perceive) an improvement and others the reverse. However, we will have all lost our "control" reference and the only way (as pointed out) of knowing IF this is really an improvement is to switch back and forth between 9 and 10 (or 8 for those who prefer that over 9). It's nonsense not least because it's impractical.
The mere fact that there is any division at all on whether this step is a sound quality improvement means it's a fail IMHO.
Considering the Devialet is highly dependent on DSPing with it's volume control, SAM, cross overs etc etc. This is all software based it's not out of realm that an updated firmware with bug fixes will have an affect on sound quality.
But in reality, no one know what they change on each release. I do know that every time they release a firmware, owners on this forum are always all over they map on what they believe they hear.
Roon->HQPlayer->McIntosh C47->McIntosh MC301s>Transparent Super->Wilson Audio Sabrinas w/ Shunyata Denali, Rega RP8, Rega Apheta 2
Posts: 22
Threads: 1
Joined: Jul 2015
Reputation:
0
After listening to 10.0.02 for a few days, I went to 9.01 for its slightly more relaxed and natural (to me) sound. The 10.0.02 gave better soundstage and image separation but the bass and highs are a little tight to my taste.
Devailet 200, Lumin A1, Magico S5, Kimber ks6063, Rockport Capella, Benz Micro Ruby, Kimber ks1026
Posts: 12
Threads: 2
Joined: Jun 2015
Reputation:
0
(03-Aug-2016, 16:47)Vivialet Wrote: (03-Aug-2016, 16:16)Rufus McDufus Wrote: I used to have Linn equipment (Klimax DS) and exactly the same thing used to happen - new firmware would come out with (say) only functional changes or a bug fix or two, nothing to do with sound quality, yet many people would report SQ differences. My opinion (looking at it from a software point of view which was my trade a while back, though usually C/C++ as opposed to the FPGA bytecode or whatever runs on the Dev) is that with a different version even minor changes will have a slightly different run profile on the processors in the Devialet, and perhaps slightly different power requirements. It would seem unlikely that these would be significant enough to make a difference but it's certainly a possibility. Another factor could be how the firmware is compiled down. We don't even know how Devialet manage their firmware build environment but I'd hope they keep it extremely consistent between builds. Different compilation flags between builds could affect how the firmware runs on the processor.
Compilers are also improved and will create more optimized code. Simply using a new compiler version can already bring SQ improvements.
Going back to what Rufus said: compiling might change something in terms or run optimisation of the code which could alter the sound BUT in what way? The subjective nature of change (much referred to already on the thread) means the outcome is hardly quantitative: not least because they don't tell us what they've done!
Posts: 2,481
Threads: 41
Joined: Mar 2015
Reputation:
48
On more extended listening I went back to the old favourite 7.1.1 in the end.
I just wish I could have this (and stay with SAM1), but with the new features of 8, like DSD.
To my ears, each version since just seems a bit more 'aggressive' for want of s better word - despite seeming a bit more exciting at first. Be it technically right or wrong, bass just seems better was v7 and SAM1 too - maybe it's just the profile for my speakers who knows.
Anyway, will be sticking here for a bit, living without DSD ad new display is a very small price to pay for the sound I had when I bought the unit.
>>> 1st Place Award: Devialet, last decades most disappointing technology purchase. <<<
Posts: 842
Threads: 22
Joined: Dec 2014
Reputation:
14
(04-Aug-2016, 22:32)Hifi_swlon Wrote: On more extended listening I went back to the old favourite 7.1.1 in the end.
I just wish I could have this (and stay with SAM1), but with the new features of 8, like DSD.
To my ears, each version since just seems a bit more 'aggressive' for want of s better word - despite seeming a bit more exciting at first. Be it technically right or wrong, bass just seems better was v7 and SAM1 too - maybe it's just the profile for my speakers who knows.
Anyway, will be sticking here for a bit, living without DSD ad new display is a very small price to pay for the sound I had when I bought the unit.
I turned SAM off completely a few momths back. It didn't do much for bass extension with my speakers in fact made them a little tubby even at 10-20% and for what improvements it made with imaging it messed with the PRAT. Sucked some of the life out of the music for a better word.
Roon->HQPlayer->McIntosh C47->McIntosh MC301s>Transparent Super->Wilson Audio Sabrinas w/ Shunyata Denali, Rega RP8, Rega Apheta 2
Posts: 445
Threads: 7
Joined: Jul 2015
Reputation:
4
(04-Aug-2016, 22:32)Hifi_swlon Wrote: On more extended listening I went back to the old favourite 7.1.1 in the end.
I just wish I could have this (and stay with SAM1), but with the new features of 8, like DSD.
To my ears, each version since just seems a bit more 'aggressive' for want of s better word - despite seeming a bit more exciting at first. Be it technically right or wrong, bass just seems better was v7 and SAM1 too - maybe it's just the profile for my speakers who knows.
Anyway, will be sticking here for a bit, living without DSD ad new display is a very small price to pay for the sound I had when I bought the unit.
Maybe there's just too much processing going on with each higher version that adds new features.
Devialet Expert 200 (FW 7.1.3) / Magnepan 1.7i / Rel T5 / Foobar (WASAPI event 24bit) / AIR 3.0.1 public beta (best Air 3.x.x SQ by far)
Posts: 10
Threads: 0
Joined: Jul 2014
Reputation:
0
(04-Aug-2016, 22:32)Hifi_swlon Wrote: On more extended listening I went back to the old favourite 7.1.1 in the end.
I just wish I could have this (and stay with SAM1), but with the new features of 8, like DSD.
To my ears, each version since just seems a bit more 'aggressive' for want of s better word - despite seeming a bit more exciting at first. Be it technically right or wrong, bass just seems better was v7 and SAM1 too - maybe it's just the profile for my speakers who knows.
Anyway, will be sticking here for a bit, living without DSD ad new display is a very small price to pay for the sound I had when I bought the unit.
My thoughts exactly.
My resident firmware has been 7.1.1. for more than a year; its´a pity it does not work with DSD...
PSA PPPremier-Devialet 250-Sonus Faber Stradivari+SAM- Tellurium Black Diamond Cables- Audionet CD (ART G3 w/ EPS)-Michell Orbe (SME V, Koetsu Red Signat)-Audionet PAM V3 -MacBookPro -Audirvana
|