Latest Thread Images

CHORD ELECTRONICS DAVE
Jan 20, 2016 at 3:19 PM Post #1,441 of 26,005
beolab said:
Hugo + w4s Remedy femto Re-clocker = remarkable combo!




Read through romaz long interview with Rob, and everything sounds astonishing, it is just one question i have from the question about femto clocks jitter resistants topic question:

I have a BlueSound Node II for convenience for my wife in the livingroom stereo. I have tried it together with my Hugo using Spidif and Toslink AQ Diamond cables.

The sound is a bit noisy ( not so black background , bright, a bit digital with not so great musicality or dept.
I think the Hugo sounds almost the same when i am using my Auralic Aries or CCK Ipad kit Hugo USB input .

But!! When i connected the W4s Remedy Re-clocker ( battery powered) to the Hugo using a Crystek CCHD-957 femto clock for Spidif / Toslink the sound is very similar to my MSB Analog !

( Headphones direct connected to the Hugo, no amp in btw)

(The Remedy Re-clocks all incoming signals like the MSB Analog does but the Remedy also up / down samples all signals to 96 khz, so it is not using Bit-perfect re-clocking)

But the sound is darker with a load of blackness and deeper detail and visceral bass and more smoother fluid musicality as a result!
The bass guitar in Brian Bromberg My Bass almost glowing in the ears and are more fluid with a greater dept and is much smoother.
= Lower jitter and lower RF noise floor from what Rob describes it.

There is more difference in the sound and more obviously with the Remedy, than if you disconnect the Hugo and just use the crappy BlueSound source directly connected to my amp.

So what i don't understand is if the Hugo are almost jitter resistant in the DLL , how come that i can hear a huge SQ increase if femto is not any good with jitter per cycle when It sounds like totally different dac SQ wise?

I want a good answer on this would be very interesting to hear and read more about.

So i will absolutely try the Remedy for my DAVE also and see if i can hear any diffrence, most likely i will not, but on the Hugo it is a huge difference like 30% i would say at least.
Click to expand...

I believe Rob's comments on Femto clocks relate to DACs and not necessarily to source components or external devices like the Remedy.  In music servers, my experience has been that the quality of the clock can make a difference although this probably doesn't apply when those servers are connected to the DAVE or Mojo.  This was definitely the case with my TotalDac.  
 
Despite Rob's comments that all of his DACs are resistant to jitter, here is his specific response to me:  "Source jitter has been eliminated on all of my DAC's from Mojo upwards".  This suggests that the jitter resistance of the Mojo and DAVE are at a different level compared to his older DACs like the Hugo and it may explain why something like the Remedy is helpful with the Hugo.  To my ears, the Mojo plays at a much higher level than the Hugo and there is rumor that the Hugo will soon be discontinued as a result.
 
Jan 20, 2016 at 3:41 PM Post #1,442 of 26,005
beolab said:
Hugo + w4s Remedy femto Re-clocker = remarkable combo!




....When i connected the W4s Remedy Re-clocker ( battery powered) to the Hugo using a Crystek CCHD-957 femto clock for Spidif / Toslink the sound is very similar to my MSB Analog !
Click to expand...

 
Very interesting but...
 
If MOJO is really immude to jitter (as opposed to HUGO), then may be it will be a better solution to buy a MOJO to replace our HUGO rather than buying a W4s Remedy to add to the HUGO ? 
 
Jan 20, 2016 at 3:55 PM Post #1,443 of 26,005
bmichels said:
Very interesting but...

If MOJO is really immude to jitter (as opposed to HUGO), then may be it will be a better solution to buy a MOJO to replace our HUGO rather than buying a W4s Remedy to add to the HUGO ? 
Click to expand...


Personally, this is how I would approach it.
 
Jan 20, 2016 at 5:48 PM Post #1,447 of 26,005
strygr said:
  For home stereo setup, Hugo is much more open and 3D in sound compared to Mojo.
 
Source? Raspberry Pi2 (with 5 A LPS) running Moode.
Click to expand...

As we know, Mojo is designed as a portable DAC for headphone use.  It can be used for 2 channel but the SQ is then significantly impacted by the quality of the cabling used and this type of cabling (3.5 mm to RCA) has limitations.
 
Jan 20, 2016 at 6:02 PM Post #1,448 of 26,005
sonic77 said:
I have yet to run the Mojo in my main set up as a dac, although I have used the Hugo in my main system regularly. Mainly because I don't have the rca conversion connection.
Click to expand...


I use know my Hugo to my Active speakers in the livingroom with Cardas SLVR + 8 pure silver conducters with Teflon screening, it is perfect, but when i connected the Remedy, its like a new DAC, much more relaxed and smoother, with great dept:wink:

Wating for Chords second batch will start this week, so my DAVE can arrive soon :wink:
 
Jan 20, 2016 at 6:02 PM Post #1,449 of 26,005
I see we are veering off Dave discussion but that's probably because some of us are just waiting for DAVE and others are happily listening to DAVE and don't have much more to say...
 
So based on what I've read in this forum about Dave and some of the other comments Rob Watts made about Hugo and Mojo, this is my suspicion. Notice how Chord has never mentioned the tap length of Mojo. I suspect this is because Mojo has <26,000 in tap length. I think Rob Watts decided to use the FPGA power to optimize for different things in the Mojo so Hugo and Mojo would sound different. Hugo supposedly has 16fs WTA oversampling filter and then eventually getting to 2048fs linear interpolation oversampling filter and 26000 taps and 5th order noise shaper. Dave supposedly has 256fs WTA oversampling filter and then eventually getting to 2048fs linear interpolation oversampling filter and 164000 taps and 17th order noise shaper. I suspect Mojo has say 32fs/64fs oversampling filter, <<26,000 taps (say 15,000 taps) and 6th or 7th order noise shaper. Rob Watts already said that the noise shaper can alter the noise floor and we know Mojo has a lower noise floor than Hugo. He also said that lowering the noise floor sometimes makes the sound too warm and soft but he can fix that by increasing the oversampling with WTA filters. But something's got to give if you're devoting your FPGA to do these tasks and obviously, that would have to be the tap length. Unfortunately, since Chord marketing has been focusing on tap length, they can't possibly market Mojo as having a lower tap length for a different smoother (possibly better) sound... Obviously, there are other design differences between the two DACs so I'm just commenting on the computation part...
 
As for source jitter, I thought Rob Watts said that all his DACs are immune to source-jitter, as are most DACs. The issue is that RF noise and correlated current/signal/ground noise can still get through into the DAC so it generates internal jitter at the DAC level and has nothing to do with removing jitter from the incoming source. It's hard to know if Hugo or Mojo or Dave are better with jitter one way or another as there are so many factors at play. What I did wonder with respect to W4S remedy is that the Remedy is ultimately not a bit-perfect solution. So most of the time, signals are being upsampled from 44kHz to 96kHz using non-WTA algorithm and probably insufficient tap length. The femto-clock may improve timing but if most DACs are immune to source jitter, the entire Remedy+femto-clock design is mostly reducing RF noise and correlated current/signal noise. Hence, I think we are dealing with many different signal changes from the Remedy+femto-clock that we can't isolate the sonic change necessarily to the femto-clock. Whether it sounds better or not is yet another matter. As long as one enjoys the sound, why not use it?
 
Jan 20, 2016 at 6:22 PM Post #1,450 of 26,005
ecwl said:
I see we are veering off Dave discussion but that's probably because some of us are just waiting for DAVE and others are happily listening to DAVE and don't have much more to say...

So based on what I've read in this forum about Dave and some of the other comments Rob Watts made about Hugo and Mojo, this is my suspicion. Notice how Chord has never mentioned the tap length of Mojo. I suspect this is because Mojo has <26,000 in tap length. I think Rob Watts decided to use the FPGA power to optimize for different things in the Mojo so Hugo and Mojo would sound different. Hugo supposedly has 16fs WTA oversampling filter and then eventually getting to 2048fs linear interpolation oversampling filter and 26000 taps and 5th order noise shaper. Dave supposedly has 256fs WTA oversampling filter and then eventually getting to 2048fs linear interpolation oversampling filter and 164000 taps and 17th order noise shaper. I suspect Mojo has say 32fs/64fs oversampling filter, <<26,000 taps (say 15,000 taps) and 6th or 7th order noise shaper. Rob Watts already said that the noise shaper can alter the noise floor and we know Mojo has a lower noise floor than Hugo. He also said that lowering the noise floor sometimes makes the sound too warm and soft but he can fix that by increasing the oversampling with WTA filters. But something's got to give if you're devoting your FPGA to do these tasks and obviously, that would have to be the tap length. Unfortunately, since Chord marketing has been focusing on tap length, they can't possibly market Mojo as having a lower tap length for a different smoother (possibly better) sound... Obviously, there are other design differences between the two DACs so I'm just commenting on the computation part...

As for source jitter, I thought Rob Watts said that all his DACs are immune to source-jitter, as are most DACs. The issue is that RF noise and correlated current/signal/ground noise can still get through into the DAC so it generates internal jitter at the DAC level and has nothing to do with removing jitter from the incoming source. It's hard to know if Hugo or Mojo or Dave are better with jitter one way or another as there are so many factors at play. What I did wonder with respect to W4S remedy is that the Remedy is ultimately not a bit-perfect solution. So most of the time, signals are being upsampled from 44kHz to 96kHz using non-WTA algorithm and probably insufficient tap length. The femto-clock may improve timing but if most DACs are immune to source jitter, the entire Remedy+femto-clock design is mostly reducing RF noise and correlated current/signal noise. Hence, I think we are dealing with many different signal changes from the Remedy+femto-clock that we can't isolate the sonic change necessarily to the femto-clock. Whether it sounds better or not is yet another matter. As long as one enjoys the sound, why not use it?
Click to expand...


I think you are on the right track!

The Mojo have the newer FPGA Xing 7 and Hugo got version 6 so most likely Rob had more processing power on tap so it most likely got 26k-45k taps or more is my estimation also, but Chord are afraid to mention this because then the Hugo would drop in selling volumes, because most of the Chord buyers are stirring them blind on the Tap rate figure. ( Interpolating filter rate)

In this interwiev https://m.youtube.com/watch?v=3e7SRXP3RHI from CES 2016, Rob mentions 44 DSP cores in Mojo and the Dave got 166 cores with 166.000 taps = around
45.000 taps in the Mojo perhaps!





Hugo + Remedy combo:

Yes everyone with a Hugo should try it out with an Remedy, its like a supercharger on a V8 engine no question about it, but i will most likely not need it with DAVE, but i will do some serious listening with it connected and see if i can hear any improvement! :wink:
 
Jan 20, 2016 at 10:09 PM Post #1,451 of 26,005
beolab said:
I think you are on the right track!

The Mojo have the newer FPGA Xing 7 and Hugo got version 6 so most likely Rob had more processing power on tap so it most likely got 26k-45k taps or more is my estimation also, but Chord are afraid to mention this because then the Hugo would drop in selling volumes, because most of the Chord buyers are stirring them blind on the Tap rate figure. ( Interpolating filter rate)

In this interwiev https://m.youtube.com/watch?v=3e7SRXP3RHI from CES 2016, Rob mentions 44 DSP cores in Mojo and the Dave got 166 cores with 166.000 taps = around
45.000 taps in the Mojo perhaps!





Hugo + Remedy combo:

Yes everyone with a Hugo should try it out with an Remedy, its like a supercharger on a V8 engine no question about it, but i will most likely not need it with DAVE, but i will do some serious listening with it connected and see if i can hear any improvement!
wink.gif
Click to expand...


Hmm, this is getting interesting. If i am not completely wrong Rob himself has consistently stated the great importance of tap length for improved SQ?
And if Mojo has almost twice the tap length of Hugo combined with a lower noise floor and immunity to jitter, I am led to  suspect that it is indeed a better more accurate headphone DAC than Hugo.
I for one will definitely compare the two with  acoustic music only, reference material asap.
Then again if Remedy  or the  Intona galvanic  isolator or some other problem solving  product will REALLY improve Hugo via speakers at home. I will try those too .
But in theory I find it difficult to believe that downsampling say 24/192 or DXD to 24/96 as the Remedy obviously does, can really improve SQ?
But of course I am still very interested to hear if there is no going back to either, after having auditioned DAVE.
 
Jan 20, 2016 at 10:11 PM Post #1,452 of 26,005
beolab said:
I think you are on the right track!

The Mojo have the newer FPGA Xing 7 and Hugo got version 6 so most likely Rob had more processing power on tap so it most likely got 26k-45k taps or more is my estimation also, but Chord are afraid to mention this because then the Hugo would drop in selling volumes, because most of the Chord buyers are stirring them blind on the Tap rate figure. ( Interpolating filter rate)

In this interwiev https://m.youtube.com/watch?v=3e7SRXP3RHI from CES 2016, Rob mentions 44 DSP cores in Mojo and the Dave got 166 cores with 166.000 taps = around
45.000 taps in the Mojo perhaps!





Hugo + Remedy combo:

Yes everyone with a Hugo should try it out with an Remedy, its like a supercharger on a V8 engine no question about it, but i will most likely not need it with DAVE, but i will do some serious listening with it connected and see if i can hear any improvement! :wink:
Click to expand...


Hugo uses the Spartan-6 LX9 (2, actually, one for the inputs). Not being an engineer, I don't know what all the specs mean, but that FPGA has 9152 login cells, 16 DSP slices, and maximum I/O of 200.

Mojo uses the Artix-7 A15T which has 33,280 logic cells, 90 DSP slices, and max I/O of 250. Only one chip perhaps, so it may be doing more duty in the Mojo.

Anyone know which chip(s) in DAVE?

Interesting reading:
https://forums.xilinx.com/t5/Xcell-Daily-Blog/Battery-powered-Chord-Hugo-mobile-DAC-headphone-amp-implements/ba-p/541775

http://theproaudiowebblog.com/rob-watts-interview.html

Enjoying my Mojo but someday...DAVE!
 
Jan 21, 2016 at 12:40 AM Post #1,455 of 26,005
Dave has one FPGA, the 34 extra cores are custom coded using FPGA fabric (using gates to make DSP cores) - so 132 DSP slices plus 34 fabric = 166.
 
All the DAC's - Mojo, 2 Qute, Hugo, Hugo TT and Dave have identical source jitter rejection, that is the DAC eliminates it. Any SQ differences in sources is not down to jitter but source RF noise and correlated noise. The galvanic isolation (USB and optical) on Dave eliminates these problems - as far as the DAC is concerned - too.
 
Rob
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 10 (members: 1, guests: 9)
Back
Top