The 4.6 verdict

Mike-B posted:

I spent a fews hours last evening with a Linn owner friend listening to my NDX;  he has been reading the posts on 4.6 & was interested.    We listened to 4.6, then 4.4 & then back to 4.6.   He was expecting the differences to be a lot more pronounced as a result of reading the forum posts,  but was very positive that 4.6 would be his choice.       That is a unanimous vote - Me,  Mrs Me & Mr Linn owner for 4.6 on NDX on my system.  

I think it'd be interesting to know in which areas you all felt 4.6 was better.

Alley Cat posted:

I think it'd be interesting to know in which areas you all felt 4.6 was better.

More or less everything,  no downsides at all - with NDX & on my system.     Mr+Mrs Linn owner & Mr+Mrs NDX agreed collectively - helped by a few too many mellow reds;   more full & detailed/timbre bass,  more/better mid range detail & tonal variation, sweater/less harsh treble,  nothing changed in sound stage width/depth but could identify/separate individual voices/instruments easier.  Mr Linn says musically it carries a tune better (Linn Tune Dem method) 

Chag... posted:

It might be time for Naim to tell the requirements for 4.6 including functionality and SQ, and explain the what and how of the scope of changes. 🤔

Chag -

Despite having worked on some computer signal processing systems, I doubt if I could relate the code changes in the DSP firmware to the resultant audio changes without having the experience of Naim's engineers.

It would probably take me a month or more (maybe even a lot more) just to fully understand the code and the relationship to the DAC and analogue filters, never mind the subtlety of the changes and how they interact with the different hardware configurations.  

Trevor Wilson posted:

Hello all

Just to say that we aren't ignoring this thread and we at Naim did perform extensive and considerable listening tests internally and in our own homes, PLUS the feedback from those that were involved at beta forum level. Whilst the feedback at that time was consistent and positive, and we ourselves are enjoying the new release -  we aren't ignoring our customers with a differing view on 4.6 release, just as we aren't ignoring those few customers on the 2.6 release that have raised concerns. To be clear and for the avoidance of implication the comments on 4.6 and the comments on 2.6 are separate systems.

These are very complex systems, with a multitude of variables and configurations - but considering our testing and listening I am frankly surprised that some people aren't happy, so this warrants us at Naim taking time to consider and analyse this.

Trevor

My wife and I have enjoyed the sound quality upgrade with 4.6 firmware to our NDS/555DR streamer ! But it is concerning a few do not .  Hope you guys can figure it out. 

Here I'm listening through NDS/555DR, 552DR, 300DR, full loom Super Lumia and SL2.
With 4.6 I was not convinced with the sound. More sparkle, more details, more soundstage but less goosebumps and less mojo. Singers were somehow further in the back and it was not as involving and intim as with 4.4. The sound didn't grap me as much with 4.6. and it got kind of boring after a while. It wasn't bad. Some aspekts I really liked (more details ...). But I thought I will go back to 4.4.
 
But then I tried moving the speakers. First further out from the wall. This didn't change things to the better. Then I thought maybe to much soundstage is the problem. So next I tried moving the speakers closer together. First I tried about 30 cm closer and "boom". Goosebumps were back, mojo again existing and singers sounded like they singing for you.
 
It is still early, I just listened to it for mayby a hour, but so far I guess I really like 4.6 now.
 
Happy listening
Stefan :-)

Hi Stefan,  interesting comment on speaker positioning.   When I first installed 4.6 on NDX (not NDS) I found it was slightly better to have the speakers moved out from rear wall.  After a few days I played around some more & I now have them back to very close to the original position.   

i think it is interesting to hear the beta tester view (as it gives insight we would never hear), but i think there can also be a unintentional bias.

I always thought the beta testers are told to keep their views to the other (beta forum), or are Naim now asking them to chime in and justify the change as good.

2.6/4.6 - Are a different sound (DSP) to the shipped units, so their are so many variables to contend with, for the home (i.e. room setup and familiarity, etc...) , plus some people just don't like change, especially when you had the setup perfect (moving speakers around is not an easy option and even then the sound is not liked).

I still think the simple solution is allow support for the shipped firmware equivalents of 2.6/4.6, then everyone is happy, or is the plan to keep offering DSP updates 

Obsydian posted:

I always thought the beta testers are told to keep their views to the other (beta forum), or are Naim now asking them to chime in and justify the change as good.

 

No instructions from Naim to chime in. My understanding of the "purdah" relating to beta testing is that we are not supposed to reveal anything about new developments not in the public domain - new firmware coming next week etc. As far as I know none of us have breached this.

I see no harm in addressing some of the understandable concerns about the recent releases by laying to rest some misunderstandings and hopefully providing clarification and reassurance.

Bottom line is we are customers who are unpaid,  who have bought our own kit, and who want to help Naim to produce the best user experience.

We are not Naim stooges!

Pev posted:
Obsydian posted:

I always thought the beta testers are told to keep their views to the other (beta forum), or are Naim now asking them to chime in and justify the change as good.

 

No instructions from Naim to chime in. My understanding of the "purdah" relating to beta testing is that we are not supposed to reveal anything about new developments not in the public domain - new firmware coming next week etc. As far as I know none of us have breached this.

I see no harm in addressing some of the understandable concerns about the recent releases by laying to rest some misunderstandings and hopefully providing clarification and reassurance.

There have been a number of occasions when beta testers have offered reassurances to frustrated users, only to find that their suggestions have backfired, and caused even more frustration. Whether or not these posts have breached the NDA will often be a grey area, but in my view, it would be better, on balance, if beta testers kept schtum and let Naim handle their issues themself. 

Graham

Trevor,

Thanks for spending the time to explain that.  I've been a Naim user for about 22 years now and this is honestly the first time that something billed as an improvement by Naim has not worked out that way for me.

I have a fellow Naim user with a broadly similar system who likes 4.6 on NDS hopefully visiting in a few weeks.  He's already said that he likes the 4.6 sound so I am curious as to whether he feels the same with my system.  We plan to start on 4.4 and then update to 4.6.

If he says he likes the change then I'm willing to chalk it up to personal preferences, so it will be an interesting comparison.  If he doesn't like it and Naim would like to investigate further I'm happy to host a home visit.  I'm based in Aylesbury.  I realise it is difficult/nearly impossible for you to mimic every combination of system and scenario that your equipment is used in.

I'm only being vocal about it because I'm thinking "where's my fantastic Naim sound gone?!?"

Sounds like a good plan Graham. I could fully understand if it’s too much faff for your friend to bring his NDS head unit to you, but if he could do that then a direct A/B to directly compare would help eliminate the need for relying upon aural memory? Might help to dispel or add fuel to the query corrupted FW download concept perhaps?

Graham,

I see that you have an nDac/PS555DR in your system.

Have you compared the new firmware sound from the NDS direct into your Preamp as well as via the nDac?

If so, what were your findings i.e. was there a similar loss of SQ both with and without the nDac in the chain?

Steve

I have seen that there have been a lot of comments about the impact on the NDX and the NDS. And it's quite interesting but also concerning to see how it splits so clearly between people liking it and people disliking it. So it seems that under certain conditions things are working out less good. Or that people find it difficult to adjust to the new sound signature which might be objectively better.

However what I missed (and excuse me if it was posted), what if any was the impact for Muso owners. I also installed yesterday the new software on the Muso. And as I don't listen to the Muso as regularly as the main systems it's difficult for me to even understand if it improved / stayed the same or even declined.

I would appreciate some views from other Muso owners who listen more regularly to their Muso's.

Hi Bert,  I appreciate your comments related to Muso,  but this thread "The 4.6 Verdict"     Mu-so and the new Uniti range is very different firmware.    Maybe you should start a new thread on Muso 1.2 Verdict & gather more specific feedback. 

Bert Schurink posted:

I have seen that there have been a lot of comments about the impact on the NDX and the NDS. And it's quite interesting but also concerning to see how it splits so clearly between people liking it and people disliking it. So it seems that under certain conditions things are working out less good. Or that people find it difficult to adjust to the new sound signature which might be objectively better.

However what I missed (and excuse me if it was posted), what if any was the impact for Muso owners. I also installed yesterday the new software on the Muso. And as I don't listen to the Muso as regularly as the main systems it's difficult for me to even understand if it improved / stayed the same or even declined.

I would appreciate some views from other Muso owners who listen more regularly to their Muso's.

Notwithstanding Mike-B's suggestion that you start a new thread (probably a Good Idea), worth stating from my perspective that I haven't found much if any quantifiable subjective difference to the SQ of my Muso since updating the firmware - though that could be because it's situated in my kitchen, and therefore by definition we're tending to be getting on with stuff whilst listening, as opposed to the kind of critical listening on my main Naim system. 

The Muso still sounds great, and I have noted a couple of effects from the upgrade, as others have observed - iRadio now takes a bit longer to tune in than it did, but the connection never drops (I'm using Wifi to service the Muso), and multi rooming seems more stable - pre-upgrade I would occasionally find that the multiroom option fell over. So it's all good so far as I see it... 

Warning, longish posting:

Are you sitting comfortably? Then I will begin…

I was thinking about the poor sound of my NDS4.6 over the weekend and one thing struck me – that is, I was using a Lindy 31887 - 5m USB 2.0 Cable - Type A to Mini-B, that I bought from Amazon in August 2015 which then allowed me to update the NDS using my desktop (Dell Optiplex i7) as the USB would be long enough.

I have run the 4.6 update from this PC twice – and each time it ran to completion OK without error but as I reported, the sound quality, to me, was dire, extremely poor. As I have mentioned before, I simply didn’t believe my not liking this sound was anything to do with preferences – or speakers, or speaker adjustments, or Snaxo adjustments – or whatever. My view was that there was probably something wrong with the new firmware that had ended up in my NDS – some kind of ‘corruption’ -- just a guess of course.

I started wondering whether 5m USB cable was somehow problematic and so last Friday I decided to buy a much shorter one – less than 1m in fact – and this was duly delivered over the weekend.

This meant of course I had to use a different computer to drive the update and so I borrowed my daughter’s laptop – an old Dell Latitude E5420 – still running Windows 7!! – and with very few installed application on it. I had been busy with work till late last night when I decided to install 4.6 again – must have been close to 0100hrs - this time using the shorter USB and from the Dell laptop. It was a bit awkward to position the laptop relative to the NDS without disturbing cable dressing etc.,, but I somehow managed in the end.

I installed the USB driver linked to in the 4.6 update instructions from Naim and started the update process – the first part -  updater application - installed OK and so I proceeded to the 2nd part where you start the NDS firmware update itself. But just before the screen where one selects the appropriate COM port – the software reported an error: ”CreateProcess failed; code 740

A little panic here – but I noticed that the “Streamer Updater” was in my applications list anyhow suggesting that had installed OK – or so I assumed. So, I invoked that – and this ran through the process to the end OK.

When the process completed, i couldn’t really have a meaningful listening session as it was quite late (or early this morning). But I listened to play one of the tracks that I found unlistenable after the two previous 4.6 update attempts.

From this brief session I sensed something was different – very different. The somewhat plastic sound I had experienced after the previous 4.6 updates seemed cured and the bass seemed to be in tune and rather more integrated.

After sending a mail to Naim support about the error that I encountered, I retired to bed. Too tired to do anything else really.

This morning there was a response from Naim about the error. Provided the NDS thinks it has 4.6 firmware – and then all is OK – suggesting that the error was probably 'spurious' – dunno. In his message, the Naim support guy also included a link to the USB driver – in the 4.4(?)  Update instructions document.

Purely out of neurosis, I uninstalled the one I had and instead installed this other one (apparently later in terms of create date) suggested by Naim. I ran the update again – by just invoking the ‘Streamer updater’ since this was already installed. I did wonder whether I should start from zero again – but decided there was no point in doing that. This ran to completion without any drama. Phew!

Then I did the superstitious voodoo “factory reset”.

I only intended to listen briefly as I was in the middle of work – but I ended up listening for much longer.

Seems the ‘draw’ is back.

My theory is that the initial firmware was somehow corrupt. My guess is that 2 things may have caused that: (a) my desktop has a LOT of peripherals connected to it via USB ports – so perhaps this got in the way and (b) perhaps the 5m long USB cable didn’t help.

You will of course appreciate this is pure conjecture on my part – I don’t know for sure. And it may be one or the other, and not both.

BUT: To anyone else having poor SQ from 4.6, I would suggest use the simplest update setup you can manage - from a laptop that is as ‘uncrowded’ as possible – and/or possibly using quite a short USB cable.

It may just work for you -- or it may not -- but worth trying anyhow.

Good luck!

Enjoy

Ken

ken c posted:

Warning, longish posting:

Are you sitting comfortably? Then I will begin…

I was thinking about the poor sound of my NDS4.6 over the weekend and one thing struck me – that is, I was using a Lindy 31887 - 5m USB 2.0 Cable - Type A to Mini-B, that I bought from Amazon in August 2015 which then allowed me to update the NDS using my desktop (Dell Optiplex i7) as the USB would be long enough.

I have run the 4.6 update from this PC twice – and each time it ran to completion OK without error but as I reported, the sound quality, to me, was dire, extremely poor. As I have mentioned before, I simply didn’t believe my not liking this sound was anything to do with preferences – or speakers, or speaker adjustments, or Snaxo adjustments – or whatever. My view was that there was probably something wrong with the new firmware that had ended up in my NDS – some kind of ‘corruption’ -- just a guess of course.

I started wondering whether 5m USB cable was somehow problematic and so last Friday I decided to buy a much shorter one – less than 1m in fact – and this was duly delivered over the weekend.

This meant of course I had to use a different computer to drive the update and so I borrowed my daughter’s laptop – an old Dell Latitude E5420 – still running Windows 7!! – and with very few installed application on it. I had been busy with work till late last night when I decided to install 4.6 again – must have been close to 0100hrs - this time using the shorter USB and from the Dell laptop. It was a bit awkward to position the laptop relative to the NDS without disturbing cable dressing etc.,, but I somehow managed in the end.

I installed the USB driver linked to in the 4.6 update instructions from Naim and started the update process – the first part -  updater application - installed OK and so I proceeded to the 2nd part where you start the NDS firmware update itself. But just before the screen where one selects the appropriate COM port – the software reported an error: ”CreateProcess failed; code 740

A little panic here – but I noticed that the “Streamer Updater” was in my applications list anyhow suggesting that had installed OK – or so I assumed. So, I invoked that – and this ran through the process to the end OK.

When the process completed, i couldn’t really have a meaningful listening session as it was quite late (or early this morning). But I listened to play one of the tracks that I found unlistenable after the two previous 4.6 update attempts.

From this brief session I sensed something was different – very different. The somewhat plastic sound I had experienced after the previous 4.6 updates seemed cured and the bass seemed to be in tune and rather more integrated.

After sending a mail to Naim support about the error that I encountered, I retired to bed. Too tired to do anything else really.

This morning there was a response from Naim about the error. Provided the NDS thinks it has 4.6 firmware – and then all is OK – suggesting that the error was probably 'spurious' – dunno. In his message, the Naim support guy also included a link to the USB driver – in the 4.4(?)  Update instructions document.

Purely out of neurosis, I uninstalled the one I had and instead installed this other one (apparently later in terms of create date) suggested by Naim. I ran the update again – by just invoking the ‘Streamer updater’ since this was already installed. I did wonder whether I should start from zero again – but decided there was no point in doing that. This ran to completion without any drama. Phew!

Then I did the superstitious voodoo “factory reset”.

I only intended to listen briefly as I was in the middle of work – but I ended up listening for much longer.

Seems the ‘draw’ is back.

My theory is that the initial firmware was somehow corrupt. My guess is that 2 things may have caused that: (a) my desktop has a LOT of peripherals connected to it via USB ports – so perhaps this got in the way and (b) perhaps the 5m long USB cable didn’t help.

You will of course appreciate this is pure conjecture on my part – I don’t know for sure. And it may be one or the other, and not both.

BUT: To anyone else having poor SQ from 4.6, I would suggest use the simplest update setup you can manage - from a laptop that is as ‘uncrowded’ as possible – and/or possibly using quite a short USB cable.

It may just work for you -- or it may not -- but worth trying anyhow.

Good luck!

Enjoy

Ken

Wow Ken ! Just Wow!  So glad you got it fixed . And it was somehow corrupted the way you originally  installed it. You must feel awesome that you fixed this ! Congratulations! 

I’d give it a few days before coming to a conclusion. If the original install showed that your system was running v4.6 then it must have installed correctly one would assume.

But good for you if it now sounds better.

FWIW I used a notebook with Windows 7, downloaded the usb drive for v4.4 from the Naim site, downloaded 4.6 and followed the instructions  using 50cm USB cable to connect to the NDS. Carried out a factory reset following the installation. Ran with 4.6 for a couple of days before deciding I didn’t like the changes to the sound and reverted back to 4.4.

Hi Ken,  I can't believe the USB cable made any difference, it either connects or it doesn't,  & I use a 5m Lindy as well & thats been 100% OK over a few years of beta testing.  You post does open up some ponderings & for sure I can't disagree with your theories,  I too am left wondering if your first install failed to complete or corrupt something.  As a matter of course I've always checked the DSP & other info in the NDX display after a new install,  it would have been interesting to look at that in your previous install.  

I ran into the problems with the 'wrong' USB driver link, long boring story but I went direct to SiLabs & got v6.7.5 so I can understand your stumbling block in that area.   

Anyway it seems like it's happy days

lot's of BS about the way to update the streamer etc. It really makes zero difference if the laptop is full of crap. Just connect a pc device with a suitable usb cable to the streamer and update the thing. There is really lot's of voodoo problems here on this Naim voodoo forum. Read the damn manuals, read the faq's, wash the ears etc. 

 

Jonn posted:

I’d give it a few days before coming to a conclusion. If the original install showed that your system was running v4.6 then it must have installed correctly one would assume.

But good for you if it now sounds better.

FWIW I used a notebook with Windows 7, downloaded the usb drive for v4.4 from the Naim site, downloaded 4.6 and followed the instructions  using 50cm USB cable to connect to the NDS. Carried out a factory reset following the installation. Ran with 4.6 for a couple of days before deciding I didn’t like the changes to the sound and reverted back to 4.4.

i am hoping of course that noting will change now on the NDS to make it sound as bad as it did before this last iteration - I'm also hoping that i'm not dreaming!

None of what I did was based on any logic or any deep knowledge of the underlying issues - I just had a hunch and tried it and it seemed to 'do something'

hope things will be better for you soon...

enjoy

ken

Robiwan posted:

lot's of BS about the way to update the streamer etc. It really makes zero difference if the laptop is full of crap. Just connect a pc device with a suitable usb cable to the streamer and update the thing. There is really lot's of voodoo problems here on this Naim voodoo forum. Read the damn manuals, read the faq's, wash the ears etc. 

 

Why are you taking this personal Robiwan ? Ken just reported the facts of his experience. I thought it was nice of him to share !  

 

 

 

 

 

Interesting Ken... that would imply that the actual firmware is sent via the serial connection with no checksums, or other means of verifying that the content is not corrupted. Seems a risky approach from Naim, but I suppose anything is possible,

I always update from a MacBook with a 0.5m USB, and a 1m Ethernet to the local switch at the bottom of my rack, so it’s all nice short connections. Perhaps some form of firmware corruption is possible?

Maybe someone with a better understanding of implementing an IIR filter on a SHARC DSP can explain further, but it seems extremely unlikely that a corruption could hit a non-critical byte or two that wouldn’t stop it working completely, but would materially impact the filter response such that the sound is so badly impacted?

In the NDS white paper, Naim state that the actual filter is implemented in 5 lines of assembly code, but I guess these are SIMD type operations, so they could each address loads of data. Could a corrupted data block containing coefficients for the filter be corrupted, but the actual execution inctructions be intact? Just trying to rationalise this in my own mind.

Glad the music is back though Ken, whatever the explanation.

Dave.

 

Robiwan posted:

lot's of BS about the way to update the streamer etc. It really makes zero difference if the laptop is full of crap. Just connect a pc device with a suitable usb cable to the streamer and update the thing. There is really lot's of voodoo problems here on this Naim voodoo forum. Read the damn manuals, read the faq's, wash the ears etc. 

 

wow!

enjoy...

ken

ken c posted:
Robiwan posted:

lot's of BS about the way to update the streamer etc. It really makes zero difference if the laptop is full of crap. Just connect a pc device with a suitable usb cable to the streamer and update the thing. There is really lot's of voodoo problems here on this Naim voodoo forum. Read the damn manuals, read the faq's, wash the ears etc. 

 

wow!

enjoy...

ken

Bit too much venom in Robiwan’s post !  Don’t let it bother you Ken. 

Great news Ken, very pleased for you.

I would hope that Naim take note of your 4.6 'journey' and try to understand if there are any possible explanations of 4.6 SQ disparity (particularly on NDS) from your trials and tribulations.

I have always maintained there in something inconsistent with the 4.6 firmware update that can't be explained away by 'personal preference' as I cannot remember such a split in attitudes from any of the pervious updates. Having said this, it seems to me that 4.4 to 4.6 has had the biggest impact (positive in my view) on SQ, so I am sure personal preference is playing a part in the disparity of opinion, but I suspect there is more to it. Anyway, Trevor has said Naim are investigating.

In the meantime, as you say Ken, enjoy!

DaveBk posted:

Interesting Ken... that would imply that the actual firmware is sent via the serial connection with no checksums, or other means of verifying that the content is not corrupted. Seems a risky approach from Naim, but I suppose anything is possible,

I always update from a MacBook with a 0.5m USB, and a 1m Ethernet to the local switch at the bottom of my rack, so it’s all nice short connections. Perhaps some form of firmware corruption is possible?

Maybe someone with a better understanding of implementing an IIR filter on a SHARC DSP can explain further, but it seems extremely unlikely that a corruption could hit a non-critical byte or two that wouldn’t stop it working completely, but would materially impact the filter response such that the sound is so badly impacted?

In the NDS white paper, Naim state that the actual filter is implemented in 5 lines of assembly code, but I guess these are SIMD type operations, so they could each address loads of data. Could a corrupted data block containing coefficients for the filter be corrupted, but the actual execution inctructions be intact? Just trying to rationalise this in my own mind.

Glad the music is back though Ken, whatever the explanation.

Dave.

 

Thanks Dave. I must admit I wasn't expecting what I did to make any difference at all, given that 2 previous iterations failed to produce the right SQ for me. In fact initially I really had to pinch myself to be sure that the nasty artefacts I was hearing before had gone. I have no idea what made the difference in the end -- I have no deep knowledge of the underlying issues, i.e. whether corruption is possible etc

One thing I cannot remember is whether I checked the version the first 2 times that I updated -- pity -- as that would been useful information.

enjoy

ken

musicfan51 posted:
Robiwan posted:

lot's of BS about the way to update the streamer etc. It really makes zero difference if the laptop is full of crap. Just connect a pc device with a suitable usb cable to the streamer and update the thing. There is really lot's of voodoo problems here on this Naim voodoo forum. Read the damn manuals, read the faq's, wash the ears etc. 

 

Why are you taking this personal Robiwan ? Ken just reported the facts of his experience. I thought it was nice of him to share !  

 

 

 

 

 

nothing personal it's in general for this voodoo forum lack of any logical common sense and silly advices.

 

 

JogoVogo posted:
nigelb posted:
JogoVogo posted:

Sorry,

bur cut off the important middle frequencies for all that bad speaker is not a experience...

....and now in a language some of us might recognise please...

Magnificent!

Lost frequency.

No, still don't understand the banter.

nigelb posted:

Great news Ken, very pleased for you.

I would hope that Naim take note of your 4.6 'journey' and try to understand if there are any possible explanations of 4.6 SQ disparity (particularly on NDS) from your trials and tribulations.

I have always maintained there in something inconsistent with the 4.6 firmware update that can't be explained away by 'personal preference' as I cannot remember such a split in attitudes from any of the pervious updates. Having said this, it seems to me that 4.4 to 4.6 has had the biggest impact (positive in my view) on SQ, so I am sure personal preference is playing a part in the disparity of opinion, but I suspect there is more to it. Anyway, Trevor has said Naim are investigating.

In the meantime, as you say Ken, enjoy!

Thanks Nigel. 3rd time lucky I guess. I'm sure you remember that I also simply didn't think this was anything to do with personal preferences. I also found it difficult to believe Naim would release something that was no negatively different from the Naim signature that I fool myself to think I know well from my acquaintance with the brand since late 70's.  But other than what I have conjectured, I have no idea what actually made 4.6 work on my NDS this time around...

enjoy...

ken

 

Likes (2)
hungryhalibutAndrew Lea
×
×
×
×