CHORD ELECTRONICS DAVE

[Hifi Boy]

Ok, I have been listening to the 44.1kHz files compared to the 192kHz files of the same tracks on the below DVDA played through Blu2Dave. I am struggling to find any difference. However through Dave alone I do prefer the 192kHz files.

That would imply that we're reaching the limit of reconstructability. Also, it would imply that the 44.1kHz files have all the necessary information for "perfect" reconstruction of sound. At some point, with a high enough taps count, you won't be able to tell any difference at all, once the sound has been reconstructed, regardless of how high the initial recording sample rate was.

 

[Rob Watts]

There is one very important issue in the mix - non integer sample rate conversion (SRC), going from say 96k to 44.1. SRCs are impossible to make without measurable THD and noise issues, and on their own create timing issues too; so whenever I see the original recording is 96k, I will buy that over the 44.1.

Moreover, existing decimation filters (for going from 96k to 48k say) are also pretty poor, as the use of cheap half band filters are today universal; these create aliasing in the 20-24kHz band, and this damages timing too (by timing I mean transient timing error, the errors that the WTA filter is trying to reduce).

Another issue - going from 24 bit to 16 bit using regular triangular dither degrades depth perception and the only way that I have managed to do this completely transparently (so far) is with extremely aggressive noise shaping at 705/768 kHz - and aggressive noise shaping is categorically not something one can do at 44.1k.

Now with Davina I have designed decimation filters that guarantee aliasing being below -240 dB or so (against today's universal -6dB from half band), so we ought to solve that issue. But two questions remain:

1. For completely transparent decimation and reconstruction "all" we need to do is decimate without any aliasing, and reconstruct with a sinc function filter of sufficient tap length; but the signal must be bandwidth limited to 22 kHz for CD; will the action of bandwidth limiting itself create other sound quality issues? I suspect it won't, and that this issue will be a red herring - but nobody knows for sure. The Davina project will answer this question.

2. Can a method be produced that allows bit reduction without upsetting perception of depth? I have had much better performance in SQ using pseudo Gaussian dither over the usual triangular dither. Would dither that gets closer to ideal Gaussian give better sound quality re depth? Again, another question that Davina will answer.

Given the above issues, I still buy the original master recording over a integer decimated or SRC version.

 

[Jawed]

A lot of this discussion is with the assumption that DAVE's upsampling to 2048FS, the "native speed" of the pulse array, isn't itself introducing problems.

One thing we've learnt since Blu 2 arrived is that the pulse array is very greedy. It has so much quality that it's difficult for anything before it to sound good enough. Remember that Rob was surprised by the large increase in performance that 1 million taps produced. Between the 16FS that Blu 2 outputs and 2048FS that the pulse array uses, there is still another factor of 128x upsampling.

Blu 2 "fixes" the gross error to achieve "16-bit resolution" with 1 million taps at 16FS, but the pulse array appears to be so greedy that it wants "20-bit resolution" or perhaps "24-bit resolution".

Put another way: potentially we can't hear what the pulse array really sounds like because we don't have 2048FS 24-bit recordings. The 24-bit 16FS input on DAVE is a very tight constraint compared with the pulse array's capability.

Now playing: Tindersticks - Vertrauen II

 

[simorag]

The Phi have fixed most issues, going very close to reaching a mix of the best qualities of HD800S, LCD4 and Utopia IMHO, with no need for EQ or other tweaks. I can listen at full enjoyment to basically everything, from baroque to heavy metal, with no sense of something lacking, and much more so as the DAVE entered the rig.

Yet, there is this nasty residual sibilance, and before I give up to blaming the Phi or HD800S, I will rather prefer to try some other routes ... starting at the "source" of the problem

Further to my previous posts on this thread, and based on the kind suggestions and recommendations from other posters, I did some tests on my system, and I am glad to report back here (sorry for the OT ), hoping this is useful to anyone.

In short, the background is that I am greatly enjoying my Abyss Phi and DAVE combo, being a slight occasional sibilance (especially on female voices) the only complaint I have.

My issue with sibilance was there even on my previous rig (Audio GD NOS 11 V2 + HD800S), therefore is not produced by the DAVE or Phi themselves.

The components upstream the DAC are: a Surface Pro 3 Laptop running off batteries > USPCB > ISO Regen + LPS-1 combo > Phasure Lush USB Cable.

Since sibilance is reportedly associated to noise pollution, as confirmed by all the ferrite-fi posts here and SMPS grounding posts on CA, I purchased some ferrites (Topsinius) and the ifi Groundhog for grounding my LPS-1, and I compared the following chains:

1. Laptop > Lush > DAVE
2. Laptop > Lush + 10 ferrites > DAVE
3. Laptop > Lush +20 ferrites (i.e. cable covered from end-to-end) > DAVE
4. Laptop > ISO Regen + LPS-1 (grounded by Groundhog) > Lush + 20 ferrites > DAVE

The test tracks I used for benchmarking purposes were:

· Spanish Harlem by Rebecca Pidgeon (Chesky, 176/24)
· Texas Rangers (idem)
· Berlin (Stockfish Art of Recording 3, 192/24)
· Secret Love by Claire Martin (Linn Records, 96/24)
· Wind Song by Patricia Barber (HD Tracks download, 96/24)

All the tracks above exhibit some passages with annoying sibilance to my ears.

The results are easy to report as long as sibilance alone is concerned: I could not hear any improvement among the 4 configurations reported above.
This was further confirmed afterwards, when I compared my laptop to a dedicated streamer (Innuos Zen MkII) at a shop, and verified that the sibilance was there as well.
Based on this test, I have come to the conclusion that these tracks are inherently sibilant to my ears (which I tend to believe are overly sensitive to sibilance).

Coming to other aspects of the sound presentation, I did notice a slight improvement passing from (1) to (3), i.e. fully ferrited cable with no ISO Regen, in form of a more relaxed sound, blacker background.
I must say that I have no golden ears and these improvements were subtle if any, and could likely be the result of self-bias.

Adding the ISO Regen (powered by the Groundhog-grounded LPS-1) to the chain was more noticeable. The bass was deeper while staying well controlled, the perceived increase of transparency and background blackness were more apparent. Harmonics were also a bit richer.
Again, I would not call it a huge difference and I am no 100% sure I would pick it up immediately in a blind test.

I must say that the benefits of the ISO Regen were much more apparent with my NOS 11, which seem to confirm that the DAVE is more immune to the upstream components than other DACs out there.

The improvement over my laptop source produced by the Innuos Zen MkII streamer was easier to detect.
While the sibilance issues remained there, the overall increase of richness and weight of the music was quite obvious. When coming back to the laptop after listening to the Zen, the music seemed flatter and less alive.

Considering that the Zen MkII should be clearly outperformed by the Zenith MkII or other high end streamers like the Antipodes DX, Auralic Aries G2, etc., I have had the confirmation of what I already knew, i.e. that my laptop, while optimized via various software tweaks, is the weakest link of my present setup.

Unfortunately, my expectations of sibilance reduction coming from source components are now much lower than before, and I believe that once I have completed my chain with a good server, and possibly a Blu 2 (my current priority upgrade), I will need a very smooth headphone like the LCD-4 or HE1000 V2 to complement the Abyss Phi for sibilant-free vocals.

 

[yakaway]

Is there any discussion of the dave being replaced by a new flagship chord product

 

[Mython]

Is there any discussion of the dave being replaced by a new flagship chord product

Absolutely not.

 

[Jawed]

Based on this test, I have come to the conclusion that these tracks are inherently sibilant to my ears (which I tend to believe are overly sensitive to sibilance).

Yeah, I would say this is a fact of life, occasionally recordings are just spiky like that.

The improvement over my laptop source produced by the Innuos Zen MkII streamer was easier to detect.
While the sibilance issues remained there, the overall increase of richness and weight of the music was quite obvious. When coming back to the laptop after listening to the Zen, the music seemed flatter and less alive.

This seems like useful evidence that a battery-powered laptop is producing some kind of noise that's affecting DAVE.

Normally the theory is that a battery powered source for DAVE cannot cause a mains ground loop, and so there cannot be a current flowing from the source into DAVE that allows common-mode noise (RF noise) to cause a problem.

So either the cabling between your laptop and DAVE is somehow causing common-mode noise to get into DAVE. Or there is ordinary (differential-mode) noise. This is effectively jitter, which DAVE is supposed to have no problem with.

Could it be that the Zen MkII, which is mains powered (is that correct?), introduces a "friendly" mains loop?

It's really confusing, overall.

Unfortunately, my expectations of sibilance reduction coming from source components are now much lower than before, and I believe that once I have completed my chain with a good server, and possibly a Blu 2 (my current priority upgrade), I will need a very smooth headphone like the LCD-4 or HE1000 V2 to complement the Abyss Phi for sibilant-free vocals.

Is the Phi smoother than HD 800 S for sibilance?

Now playing: Paul Giger - Trogener Chilbilabe

 

[Triode User]

Is there any discussion of the dave being replaced by a new flagship chord product

Categorically "NO" and "Years away" paraphrasing earlier posts from the horses mouth.

 

[yakaway]

Does anyone know if the older ps audio perfectwave transport would be any good with the dave or even hugo 2. I'd love the blu, I just don't know if my cd collection is that relevant to spend that kind of money.

 

[yakaway]

And thank you for clarifying that there is no new product about to supplant the dave. I'm just astounded at the cost of these things, but I guess you get for what ou pay for if you do the research.

 

[Triode User]

And thank you for clarifying that there is new product about to supplant the dave. I'm just astounded at the cost of these things, but I guess you get for what ou pay for if you do the research.

Presumably you meant ‘no new product’.

 

[Jawed]

Does anyone know if the older ps audio perfectwave transport would be any good with the dave or even hugo 2. I'd love the blu, I just don't know if my cd collection is that relevant to spend that kind of money.

Before I bought DAVE, I expected my mostly "low brow" music to not benefit from a top DAC. It turns out that it makes everything much better. Even stuff recorded badly on cassette tape originally still comes through with a new kind of vigour, honesty, insight and passion. The magic is recorded into the CD, but getting it out of the CD is extremely difficult...

Many would argue that for not much more than the cost of DAVE you can get something a lot better in the shape of Blu 2 + Hugo 2.

It's a risk: you could buy Hugo 2 now, get to know the way Chord DACs make music (it might take a while, they really are that different) and then upgrade by adding the Blu 2.

Or you could go for DAVE then add Blu 2 later.

I'm just astounded at the cost of these things, but I guess you get for what ou pay for if you do the research.

See if you can ask a dealer to let you use it at home for a week or two, if you have any doubts about it.

In the end you have to decide what you can stomach. This is stuff that should last 20-30 years. Eventually Chord will release new products that make this sound quality cheaper to buy, but all the signs are that it will be years away.

You could buy Hugo 2 now, while you wait until that miraculous future arrives (who knows when). But then you might accidentally buy a Blu 2. And then sell the Hugo 2 to buy a DAVE. And then think to yourself, "why did I wait so long?"

Now playing: My Morning Jacket - Honest Man

 

[yakaway]

I actually own the Hugo 2.
Bought it for its great quality, versatility and also for its relative portability (and I do mean relative).
I understand the mojo has more coloration and is a good all arounder, but the price was right so I opted for the Hugo 2. I'm glad I did

I got into this hobby about a month ago and went from HE 560s, to tesla t1 gen 2's. When my Focal Utopias came, I knew the others have to go back! There was just no comparison.

Its been a busy month. Added the stax 009s with the SRM-7tII (which I am not happy with, but that'll have to be discussed elsewhere I suppose), and Abyss 1266's are on their way from Europe (on the list for Phi upgrade).

To keep this short, the Hugo 2 and Utopias are wonderful. I actually thought I was craazy for not being satisifed with the way I was hearing music... Until now.

But to go back to this idea that Blu 2 and Hugo 2 have a better sound than Blu 2 and Dave is just mind blowing. And, yes I will likely be purchasing a Blu 2 at some point soon.

Just bought a PS audio transport (price was right) and so I want to spend a minute with that configuration.

Can anyone offer an explanation to this notion? Are there multiple articles describing this. Any mention or input from Chord directly? Is this one person's impression or does this have merit?

If I were a DAVE owner, I'm not quite sure how I'd feel about this. Better performance for less money is great, unless you are the guy who spent more for a dave.

I suppose it will be better from every other source, so it's probably unjustifiable for the cost.

 

[Hifi Boy]

With all that talk about ferrites, made me wonder, is it possible to somehow add them to the next product? Let's say there is a standalone MScaler. Would it be possible to have a built in batch of ferrites, or whatever would offer good isolation right in the product itself?

Then we wouldn't have to make our cables look weird by putting them on the cables.

 

[zimzim2001]

I actually own the Hugo 2.
Bought it for its great quality, versatility and also for its relative portability (and I do mean relative).
I understand the mojo has more coloration and is a good all arounder, but the price was right so I opted for the Hugo 2. I'm glad I did

I got into this hobby about a month ago and went from HE 560s, to tesla t1 gen 2's. When my Focal Utopias came, I knew the others have to go back! There was just no comparison.

Its been a busy month. Added the stax 009s with the SRM-7tII (which I am not happy with, but that'll have to be discussed elsewhere I suppose), and Abyss 1266's are on their way from Europe (on the list for Phi upgrade).

To keep this short, the Hugo 2 and Utopias are wonderful. I actually thought I was craazy for not being satisifed with the way I was hearing music... Until now.

But to go back to this idea that Blu 2 and Hugo 2 have a better sound than Blu 2 and Dave is just mind blowing. And, yes I will likely be purchasing a Blu 2 at some point soon.

Just bought a PS audio transport (price was right) and so I want to spend a minute with that configuration.

Can anyone offer an explanation to this notion? Are there multiple articles describing this. Any mention or input from Chord directly? Is this one person's impression or does this have merit?

If I were a DAVE owner, I'm not quite sure how I'd feel about this. Better performance for less money is great, unless you are the guy who spent more for a dave.

I suppose it will be better from every other source, so it's probably unjustifiable for the cost.

The point was that Hugo2 + Blu2 is better than Dave alone, not that Hugo2 + Blu2 is better than Dave + Blu2.
Hugo2<Dave<Hugo2+Blu2<Dave+Blu2