advertisement


Kef wireless 50's

I too think that there's a certain quality to Audirvana that is a bit special. Alas I have four issues with it: (1) it only runs on Mac OS and I want to go headless so that's either a Hackintosh solution or selling my NUC and getting an equivalent Mac Mini (which would then need an internal PSU upgrade), (2) TIDAL can only be controlled from A+ remote, (3) it keeps crashing. I had problems with Audirvana crashing even with a supposedly stable build. I am on the latest beta so that I can use the uRendu as a NAA but it crashes often - as to be expected to be fair, and (4) I really do prefer the Roon solution.

I think going 16/44 is a solid option. Nice move.

Are you thinking of using your TT DAC into the analog inputs on the Kef? I presume you are aware that it will convert it back into digital? The DARkO review seems to suggest that he had no degradation in his enjoyment of vinyl via. the Kef's analog inputs. Don't think he tried an external DAC though :)

Yes, TT DAC into the analogue to see a)which I prefer, and b) perhaps save some money. I wasn't aware of it being made digital again so there is no experiment here as Keith also highlights. Even so, this hasn't put me off. If the KEFs sound well, i'll be keeping them. Convenience plays a big factor in my house. I'm always switching between A+, youtube on i devices so the input selection is very appealing me to. I will buy a sub at some stage (another plus for the speakers), probably after the next house move. I am not suggesting they lack bass but a well integrated sub brings a huge smile to my face. Then finally iRC which has also been on my list for way too long now.

I have tried Roon several times and as much as I want to like it, it just isn't for me, even scrolling across irritates me, pedantic I know. I can see the fuss but i'm lucky to run A+ with the app and not once had a single issue. The app on an iPad Pro is wonderful to use - as such no going back for me in this dept.

I would look into why your app is crashing. Drop Damien an email, he's a very helpful chap and will likely have some suggestions for you.

Keep us updated :)
 
Pedro, you may want to give this a read (if you haven't already):

http://www.digitalaudioreview.net/2017/01/between-ad-and-da-the-power-of-pragmatism/

Ref: Audirvana, I've just signed up to Roon for a year so will stick with it for now.

This article has motivated me to hook up my Naim DAC v1 to the KEF analog inputs. Just ordered a DIN-to-phono cable so should be able to try this out at the weekend. I've no idea what to expect. I'm hoping that I prefer the in-built DAC.... I need to sell the v1!
 
Er...any signal fed to the KEF analog inputs are going to be converted to digital by an internal ADC. There is no bypass - everything goes through the DSP section before getting to power amps.

There surely can be no benefit in outputting the NAIM DAC to the analog inputs. Whatever is feeding the NAIM can go straight to the KEFs.
 
Er...any signal fed to the KEF analog inputs are going to be converted to digital by an internal ADC. There is no bypass - everything goes through the DSP section before getting to power amps.

There surely can be no benefit in outputting the NAIM DAC to the analog inputs. Whatever is feeding the NAIM can go straight to the KEFs.

Read back through this thread. I'm well aware of what the Kef does and how it handles analog inputs. Several Kef users (not on here but check Computer Audiophile and AVSforums) have settled on using their external DAC into the Kef compared to the Kef internal DAC - google it.

I'm not convinced but then I'm a scientist so I like to experiment. The DAC v1 has a superior asynchronous USB front-end with galvanic isolation, reclocking, higher quality DAC than the Kef. Whether that is realised as lower jitter, etc. who knows. And whether that 'difference' is palpable and whether it 'survives' the ADC->DAC process in the Kef.... who knows.

Sometimes empirical evidence is as important as the theory and so for the sake of £15 I'm going to try it for myself before I sell on the DAC v1.
 
So I've just realised that the DAC v1 has RCA pre-outs and I have a (very old) pair of RCA-RCA phonos in the garage. Possibly not a fair comparison but am going to give those a whirl tonight feeding the Aux line level inputs on the Kef.

The only reason I can think of for there to be any worthwhile difference is if there is a poorer implementation of the USB input on the Kef (e.g. The Naim V1 is based on the audiophilleo async USB solution which is very well regarded). Everything after that should be treated identically whether it's the Kef USB or the Naim analog line inputs.
 
Basic block diagram of the ls50w. Not sure what you would benefit from putting another conversion in the mix.

LS50_Block.png

The *only* conceivable difference, IMHO, is in how KEF have implemented the asynchronous USB input. As I said earlier, and as illustrated by the block diagram, once you get past the USB (or ADC process) everything else is identical. I know that the Naim DAC v1 has an excellent implementation of Async USB (if interested then google Audiophilleo for a description of why reclocking, ultra low noise, etc. are critical to the USB process).

Hence, the hypothesis is this: "The implementation of the async USB in the DAC v1 leads to an appreciable improvement via. the ADC compared to USB directly via. the async USB implemented in the Kef".

FWIW, everything leading up to the DAC v1 or KEF USB will be identical (e.g. custom NUC running ROCK feeding a microRendu with LPS-1 supply).

If I have the inclination I will experiment with setting Roon to upsample to DSD128 when using the Naim DAC v1 in the chain.
 
Short synopsis for now, I will do a more detailed write-up later (e.g. set-up, test protocol, tracks used, etc.):

Naim DAC v1->Analog Line Inputs

vs

KEF USB direct

There is a difference in the sound output from the Kefs. Some of that old Naim sound (precise, fast, forward, resolving) was evident through the Kef analog inputs. Suggests an almost completely transparent ADC in the Kef. Both inputs sound great. Interestingly on some test tracks the Naim v1 was more resolving in the higher frequencies. Overall I prefer the Kef USB direct as it presents a slightly more relaxed (warmer) sound than the Naim. I was tiring quite quickly of the Naim sound (which, incidentally, is what drove me to move to the Kef's in the first place). Conclusion? Don't write off using external DACs with the LS50W's just yet?
 
Setup

Front-end:

Room Optimised Core Kit (ROCK) installed on Intel NUC i5 in Antec ‘Plato X’ fanless case.
Music library on external SSD drive connected via. USB. 5v external linear power supply to to SSD drive. 15v linear power supply to the ROCK.
Silver-plated ethernet cable (DIY) to TP-Link pro switch (9v LPSU powered).
From switch via. Cat7 shielded cable to TP-Link 100CM fibre media converter. 1m fibre optic patch to second TP-Link 100CM FMC with 5v LPSU, 0.75m Audioquest Cinnamon ethernet cable to microRendu (7.5v powered but Uptone Audio LPS-1 ultra capacitor supply). Curious USB ‘regen’ link cable to Naim/KEF USB input.

LS50W:

Stock power cables to both speakers.
Chord C-Stream 5m ethernet ‘speaker cable’.
Temporarily sat on Monitor Audio speaker stands.

A/B Testing

Both the Aux input and USB input set to ’60’ (basically, keep turning the volume down for about 100 button presses, and then 60 button presses on vol up; each button press confirmed by light going off/on on top of right speaker).

The first 30 seconds of each track, repeated 3 times. All tracks listened through on either Naim or Kef before switching over. I make notes as I go along.

** All played at native sample rates **

Excuse Me Mr - Ben Harper (24/96) - bass, drive, attack
Ice - Sarah McLaughlin (16/44) - acoustic, vocals
Give Life Back to the Music - Daft Punk (24/88) - rhythm, musicality, bass
Jim Kelter-Drum Improvisation - Sheffield Lab Drum and Track Disc (24/96) - speed, precision, attack, resolution (** NB: I didn’t do any blind testing with this track **)
Inception (Mombasa) - Hans Zimmer (24/96) - speed, precision, attack
Big Jet Plane - Angus & Julia Stone (16/44) - mainly focusing on the coherence of Angus’ vocals over instruments.
Time - Pink Floyd (16/44) - lots of glassy/shimmery stuff with the bells at the beginning.

** My notes for each track get repetitive so I’m just going to summarise my main points **

DAC V1 with balanced DIN output to RCA input:

Bass is rock solid (e.g. Daft Punk, Excuse Me Mr).
Possibly too clinical; detracts slightly from the cohesion and ‘musicality’?
Precise.
Not sure whether its ‘musical’.
Fast.
Possibly better at resolving in upper frequencies.
Possibly more ‘timbre’.
Exceptional vocals.

Kef USB direct:
Very, very good snare drum hits. Like fireworks going off!! (Ben Harper, Drum Improvisation track)
Nice. Enjoying this! :)
Coherent.
Balanced.
** Pretty much most of the positives from the DAC v1 apply here too **

Of course, to either of the above you can also add: fantastic imaging, good soundstage, enjoyable!

The only noticeable difference was with the drum improvisation track. This is a fantastic demo of a hifi setup and, frankly, I have never heard this track sound so good. Fair play to Kef, this is a stonker of a music system. So, I spent about 2 hrs this morning trying to nail this difference down. This is a track laced with symbols, triangles, and - basically - anything that can be hit/kicked/thumped. The decay of a triangle being hit or the resounding echo of a kettle drum was more noticeable with the Naim. It sounded wonderful and three dimensional. There was a precision to how Naim was resolving these sounds that was very impressive. I found myself preferring this track played on the Naim but, and its a big but, we are talking fractions of differences here. The KEF also rendered this track beautifully.

Apart from this one track - which is quite technical - I preferred how the KEF USB rendered the others. They were more musical, 'listenable' and enjoyable.

‘Blind’ A/B testing

Now, onto the blind A/B testing. For this all that I wanted to confirm was that I could detect a difference between them and that I wasn't applying any kind of subconscious bias. Same parameters and tracks as above except this time I recruited my 7-year old son to switch over the USB cable to either the Naim or the Kef, and to set the Kef to the right input. He was a bit reluctant at first until I explained that I would need to do what he told me to do. Hmmmm…. Well, he took that to heart and was bossing me about for an hour or so. I explained that his job was to trick me so that I didn’t know which ‘box’ he’d plugged it in. And, I said that he didn’t even need to change it over.. now that would be really sneaky! Well, he decided to banish me to the house (my hifi kit is in my man cave in the garden) each time he did the switchover. The first time around I got a rather puzzled look from my wife. By the 4th time I sheepishly walked in she drily commented “back again huh”. Bless him though, he did a great job, didn’t let on what he’d done, and - more importantly - didn’t break anything!!!

In total, we had five 'switcharounds' (or 'switch-a-roos').

I sat through the 30 seconds of the same tracks as before but only once (you can only keep a 7-yr old engaged for so long!). By the third time around I was pretty sure I was listening to the Naim/analog. That precision and resolution I’d noticed during open listening was there in spades. In fact, it was sounding pretty damned good - better than I remembered actually. Anyway, I stuck to the plan and we did another two switcheroos. By the end of it I'd noted:

1: Analog
2: USB
3: Analog
4: Analog
5: USB

Any this is what my son had actually done:

1: USB
2: Analog
3: USB
4: Analog
5: USB

So much for me having any kind of ‘golden ears’!!!!! Ha, this completely threw me - blind A/B testing does appear to have some merit to it. I was quite surprised that I’d completely got the input wrong but that I had, mostly, been right that I was listening to a different input.

For what was left of this morning I had the Naim DAC via. the analog inputs playing in the background whilst I was working. I’m afraid to say that I found myself either turning the volume down or changing tracks as I was finding extended listening quite tiring/irritating. This is precisely what my listening experience was like with the DAC v1->NAP200->Tannoy setup prior to the Kef’s arrival. Don't get me wrong, it is a BETTER sound than I was getting with the Naim/Tannoy setup, but not good enough.

After lunch I went back to the simpler set-up with my microRendu directly into the KEF USB input. It’s been playing ever since and the volume has gradually creeped back up to my normal listening levels. At the end of the day that’s what counts. So, whilst I really do appreciate the verve with which the Naim DAC goes about its business, and how impressed I may be with the transparency of the Kef’s ADC, I just don’t get on with the Naim sound.

Summary

So, my original hypothesis is false “The implementation of the async USB in the DAC v1 leads to an appreciable improvement via. the ADC compared to USB directly via. the async USB implemented in the Kef”. If however the hypothesis was there was an appreciable difference then it would hold true. Alas, to these ears anyway, the Naim DAC v1 is too forward for my tastes.

To others who have external DACs and who are wondering whether there’s any worth in trying out the Kef’s analog inputs all I’ll say is this:

Is there a noticeable difference over the USB input? Yes.

Is there any degradation in SQ using an external DAC to analog inputs? No. The KEF ADC process appears to be very transparent.

Is there an improvement on the sound quality over the USB input? Depends. If you already love the sound of your DAC then you may find that this transfers nicely into the KEF set-up plus you get all that goodness of phase-aligned and timed crossover, room corrections, and that wonderful dual-DAC/dual-mono set-up driving each speaker!!! For me, this experiment has confirmed what I suspected: I’m not a fan of the Naim sound. And that has nothing to do with the KEFs.

If folks are interested I can go into (a lot) more detail on my notes from each track.

In the meantime, I'm going to start chipping away at a backlog of jobs that have piled up :)
 
So much for me having any kind of ‘golden ears’!!!!! Ha, this completely threw me - blind A/B testing does appear to have some merit to it. I was quite surprised that I’d completely got the input wrong but that I had, mostly, been right that I was listening to a different input.

Not sure I agree with the second conclusion. You correctly guessed that there had been a change 3 times out of 5 - that is as close to random as you can get with an odd number of tries.
 
Not sure I agree with the second conclusion. You correctly guessed that there had been a change 3 times out of 5 - that is as close to random as you can get with an odd number of tries.

:) From a sample size of 5? Is that statistically significant?
 
<sigh> I take the time to do this experiment, post it on here, and the only person who has bothered to respond is a pedant..... It's a big world out there y'know....
 
Great write up Rik, i had the passive versions and thought they were smashing really.
Same goes for your findings on the different inputs ADC etc, it pretty hard to tell difference between Dacs if they are level matched, anyway i've enjoyed your findings :)
 
Had a pair of these for about 3 weeks now and thought I'd add my thoughts.

On paper, and financially, these are a big step down on the numerous seperate set-ups I have had over the last two decade or so. I even had a pair of Avanatgarde Zero 1 Pro digital active horns so have been down a similar 'all-in-one' road before. My room is about 12ft x 9ft and the speakers fire down the room - photo here:





I only use Tidal for my music and so have been using the version built in to the Kef app.
It's not very good - limited options, can be very flakey, slow to respond and has no gapless playback - but I can live with it if it means not having to buy another box and cable for the streaming. If it gets too annoying I'll buy another Auralic (Mini or Aries).

I should add that I can't upgrade the firmware as I don't have a suitable laptop and it can't be done direct from an iPad. Possibly the newer firmware improves things.

I started out with them on some Something Solid open frame stands but changed to some Dyanudio Stand 2 filled with sand and that improved things no end.

The dsp sound adjustments have proven invaluable in tuning them to my room - I have the bass down by 1db and the treble up by 1db.
Today I added a Rel T5 subwoofer, which I bought very cheaply on here, and that has improved things further, not just in terms of slam, but more generally in terms of dynamics, soundstaging and openness. Got my eye on a top of the range Rel sub now.:D

Anyway in terms of sound quality thay are seriosuly impressive. It is a very well balanced sound - nothing stands out and if there are any weaknesses they aren't obvious. Highlights are the great levels of detail and insight, but not at all bright or fatiguing. Soundstaging and imaging are particularly impressive - the old veils being lifted trick.

Bass, without the subwoofer, was as good as you could reasonably expect from the cabinet size - maybe not the last word in snap and speed but a decent amount of heft.

For the money, indeed even if you have a budget 3 or 4 times greater as I used to have, they really are a superb solution, especially if you mainly stream your music. Hopefully Kef give them the support they need and iron out the little annoyances over the next few months.
 
One criticism I have of fast AB testing/blind testing is it can take quite a long time, maybe days or weeks, to really get to know a product or a change. All the stuff we listen to tends to be good, so we are looking for fine differences and where these differences reside may not be immediately obvious. As an example take MP3 vs. Lossless. In a fast AB or worse an on-the-fly switch I may be caught out, but once I learned what to listen to I can tell them apart pretty easily (follow the reverb tails and listen to guitar distortion, MP3 can do neither well!). As such I am no fan of this type of quick switch dem and would place little weight on your not "passing" the test. If you really want to know which is best play each for a week and see which one makes you want to play more music/investigate new or seldom played titles. That is the best test IMO. I can sometimes not tell if I've improved something, but after a bit of time it becomes very clear if I've made it worse! Despite that I admire your dedication!

PS The Kefs sound like a winner I like the standard LS50 (assuming a nice big powerful amp is nearby), so I'd expect the actives to be rather good!
 
Today I added a Rel T5 subwoofer, which I bought very cheaply on here, and that has improved things further, not just in terms of slam, but more generally in terms of dynamics, soundstaging and openness. Got my eye on a top of the range Rel sub now.:D

Interesting. What size is your room? Mine is 3.3 x 6m (or 11x20ft in old money).

I don't feel that I'm lacking bass but I do sometimes wonder whether handing off some of the lower frequency stuff to a sub would give the Kef's a better frequency response above, say, 80hz. What have you got the sub xover set at?
 


advertisement


Back
Top