Watts Up...?
Mar 28, 2021 at 12:29 PM Post #2,221 of 4,753
I am just wondering if this product can effectively break the ground loop of DAVE and reducing the need for multiple ferrite?

https://ifi-audio.com/products/gnd-defender/
I presume that gadget just lifts the ground. I think it depends on what the issue is. If it‘s a traditional ground loop, it can work. But if it’s power supply related leakage current loop RF noise, it may or may not work. So that’s my way of saying probably not? My guess is that for most systems, the best solution would be
https://ifi-audio.com/products/groundhog/
Where the groundhog is plugged into the USB source that’s connected to DAVE/M-Scaler. That should hopefully remove the ground loop noise and prevent it from getting into DAVE/M-Scaler? Not really sure.
 
Mar 28, 2021 at 1:00 PM Post #2,222 of 4,753
Since I could not find anyone who posted a link with the new solid core ferrites (not the clamp on one, as Rob sugested that you need those because for some reason they do not work with the mscaler, but only with blu2) I began a search for ferrites that go up to 2Ghz. The biggest problem was that almost all manufacturers only provide datasheets going up to 1Ghz. But those ones I ordered do look promising. Atleast they should fit the cable and seem to be the right type of ferrite material. I can post my listening results once they are mounted.
https://www.mouser.de/ProductDetail/710-74270061
 
Mar 28, 2021 at 1:03 PM Post #2,223 of 4,753
The Linn Organik is a very interesting product, also very expensive, but it does have streaming (including Roon) and speaker/room correction capability built in. And optical ethernet, HDMI, and plenty more.
I think Linn nowadays is a very different company than Chord and has been for a decade now.

Both started out as heavy engineering companies that produce state of the art products that they know how to make with the focus on optimal audio quality.

But over time, Linn devoted more and more of its engineering resources into all-in-one systems and focussed more on solving what my favorite photography blogger Thom Hogan calls ”user problems”: wanting an HDMI input, wanting high-end optical Ethernet, speaker/room correction capability, or even having 1 box instead of multiple boxes.

Whereas Chord devoted its efforts to solving audio engineering problems that users don’t even realize they have: transient and timbral accuracy, noise floor modulation. This is why IMHO, Chord products sound better.

The problem with the Chord approach is that just because the product is better, the users may not appreciate the improvement, as evidenced by all the forum posts that says Chord DACs are nothing special. We see this with some Apple products too. Chord (and sometimes Apple) are basically solving “user problems“ that users don’t even know they have. But there is no guarantee users would appreciate these solutions. Audio is particularly hard because our ears can get used to anything. I’d be perfectly happy listening to music on my iPhone speakers, even with all its distortions. Moreover, by not selling all-in-one products, Chord is at the mercy of customers pairing their Chord products with pretty much anything else which may have noise floor modulations (class D amplifiers), electronics with poor transient accuracy and distortions, etc. And no customers liked to be told that what they current own is “problematic” or that their ears are so used to distorted sound that they can’t hear the difference. So they’d rather say, well Chord DACs sound like everything else, because on their system, they can’t hear the difference.

Whereas user problems that users recognize, like not having an all-in-one box, no speaker/room correction, no HDMI input, no 5GHz WiFi in an apartment where there is lots of 2.4GHz WiFi are problems that users can readily point to that they might feel Chord is not solving for them.

Linn is also smart in that it insists most of the time that their dealer network really supports their products and dedicate rooms to their products. At least that’s at the stores where I’ve seen Linn products (Vancouver & Montreal). Whereas most Chord dealers who pair them with speakers often stock lots of other brands and allow users to mix and match whatever they want (because that’s how people buy components). Imagine pairing an amplifier with low SNR or high harmonic distortion with Chord DAC vs pretty much any other DAC, you’re not going to hear as much of a difference/improvements.

I think it is also why Chord is much more successful in the headphone space because at most headphone stores, the demo is your portable battery powered device into Chord DAC vs other DACs into your headphones. There are significantly fewer variables that deteriorates the sound.

But ultimately, why can’t Chord do both what Chord is currently doing and what some of Linn is doing? My take is that you only have so much engineering resource. So you can devote it to say creating 2Go & 2Yu as your optimal streamer solution but you’ll have to forego adding 5GHz wifi, Bluetooth AAC support, HDMI input, room and speaker correction, etc. Moreover, adding these user-desired features may mean audio performance compromises which Chord doesn’t want.

Moreover, users are finicky. I might want 5GHz WiFi and Bluetooth AAC support in Hugo 2 & 2Go, but I might not want to pay for HDMI input and room/speaker correction. You might want to pay for room+speaker correction and optical Ethernet but you don’t want to pay for 5GHz Wifi & Bluetooth AAC support. So what is a company supposed to make? And how much does the company charge?

The worst I have heard from many companies is that they would do an online poll on forums like ours and they’ll get all of us saying we would totally buy whatever product for a specific price with a specific feature set so in their polls, they can estimate that 10,000 will buy the product and when the actual product comes out, only 100 people would buy it, even though in the forums, at least 500 people said they would. Just because we say here we want a specific product at a specific price doesn’t mean when the product comes out we actually pay for it.

Another aspect to me that is interesting specifically about speaker and room correction is that Linn and Devialet (or even Dirac/Audyssey) never sells the product as “your room and your speakers and your speaker/listening seat positioning are crap” so you need our products to make things better because that antagonizes customers. It’s always “our products can make your great speakers and room sound even better”. Except of course, simple logic tells you that you can’t digitally correct everything that’s physically wrong acoustically with your system. For instance, Devialet allows you to dial in how much speaker correction (and the bass frequency response) you want to engage because all digital room/speaker correction are a compromise.

At the end of the day, I’m very happy listening to music from my Chord system right now (with digital convolution filters in Roon to correct my speaker/room issues). I’m grateful for the products that Chord makes. I think it’s a tough market. I hope they continue to succeed. And for others who prefer other brands and are happy with their stereo systems (or headphones), kudos to them.
 
Mar 28, 2021 at 1:52 PM Post #2,227 of 4,753


When Rob hears technical nonsense... 😂
 
Mar 29, 2021 at 4:33 AM Post #2,229 of 4,753
I am talking about sound quality and musical performance. To be honest I don't care about anything else.
Why shouldn’t the Linn Organik compete with Chord on the basis of sound quality and musical performance? Apart from their in-house engineering and manufacturing skills, they have been making award winning classical recordings for decades, so they do know about sound quality.
 
Last edited:
Mar 29, 2021 at 7:24 AM Post #2,231 of 4,753
Why shouldn’t the Linn Organik compete with Chord on the basis of sound quality and musical performance? Apart from their in-house engineering and manufacturing skills, they have been making award winning classical recordings for decades, so they do know about sound quality.
So making recordings means you can be a DAC designer?

You have no idea of the work involved in doing a DAC design from the bottom up. It's very easy to cobble together a PWM flip-flop FPGA based DAC design that will measure half decently. It's another thing entirely to create a state of the art design - this takes a huge amount of effort, with man decades of work, in coding, verification, research, and listening tests - if you are interested in maximising transparency and musicality. If of course you merely want to create good marketing copy to sell some expensive products to gullible individuals and "reviewers" who just want a good story then it's very easy to do. But in reality it's very much more difficult if you are interested in chasing the absolute performance and pushing the envelope.

I have been working on DACs for 33 years now, and developing my own discrete DACs for over 30 years. And I am still learning new things. I am constantly amazed at how very small errors have huge SQ impact. But the scheme that Linn allude to in their video is a very basic conversion scheme - something I was doing over thirty years ago (I sound like a whining old codger now). And the problems of PWM with analogue moving average filters are impossible to eliminate - the essential problem is that they have poor transient timing problems, and innate HF distortion. The reason being is that PWM modulation has a timing that means that high amplitude signals come earlier than low amplitude signals; this creates transient timing errors, which if you have read my many posts you will know that will damage sound quality, as the timing of transients is an essential cue for the brain to make sense of the music. This scheme (together with many other schemes) were rejected when I finished the first version of pulse array in 1995 - which doesn't suffer from this problem at all.

Another issue with PWM schemes is that the noise shaper output is limited to the PWM switching frequency, which using a 2048FS master clock and 32 cycles would be around 3MHz. This guarantees that small signal accuracy will be impaired - there is no way you can achieve anything like the 350dB accuracy (this performance is essential for transparency) that I get on my pulse array modulators (as the noise shaper produces a new value every 9.6 nS not every 354nS with a PWM noise shaper). Poor small signal accuracy results in a flat soundstage and degraded detail resolution.

I am sorry if this post upsets you but I get very angry and frustrated when people just assume that a brand name automatically means better performance.
 
Mar 29, 2021 at 8:55 AM Post #2,233 of 4,753
So making recordings means you can be a DAC designer?

You have no idea of the work involved in doing a DAC design from the bottom up. It's very easy to cobble together a PWM flip-flop FPGA based DAC design that will measure half decently. It's another thing entirely to create a state of the art design - this takes a huge amount of effort, with man decades of work, in coding, verification, research, and listening tests - if you are interested in maximising transparency and musicality. If of course you merely want to create good marketing copy to sell some expensive products to gullible individuals and "reviewers" who just want a good story then it's very easy to do. But in reality it's very much more difficult if you are interested in chasing the absolute performance and pushing the envelope.

I have been working on DACs for 33 years now, and developing my own discrete DACs for over 30 years. And I am still learning new things. I am constantly amazed at how very small errors have huge SQ impact. But the scheme that Linn allude to in their video is a very basic conversion scheme - something I was doing over thirty years ago (I sound like a whining old codger now). And the problems of PWM with analogue moving average filters are impossible to eliminate - the essential problem is that they have poor transient timing problems, and innate HF distortion. The reason being is that PWM modulation has a timing that means that high amplitude signals come earlier than low amplitude signals; this creates transient timing errors, which if you have read my many posts you will know that will damage sound quality, as the timing of transients is an essential cue for the brain to make sense of the music. This scheme (together with many other schemes) were rejected when I finished the first version of pulse array in 1995 - which doesn't suffer from this problem at all.

Another issue with PWM schemes is that the noise shaper output is limited to the PWM switching frequency, which using a 2048FS master clock and 32 cycles would be around 3MHz. This guarantees that small signal accuracy will be impaired - there is no way you can achieve anything like the 350dB accuracy (this performance is essential for transparency) that I get on my pulse array modulators (as the noise shaper produces a new value every 9.6 nS not every 354nS with a PWM noise shaper). Poor small signal accuracy results in a flat soundstage and degraded detail resolution.

I am sorry if this post upsets you but I get very angry and frustrated when people just assume that a brand name automatically means better performance.
Interesting to read ,so when,when, when? can we expect THE SOTA product in ONE and the same package, that includes ALL you´ve learnt in all these years and WITHOUT any of the compromises in both Qutest, H2 and TT2 or even Dave.
Why does a competing product like dCS´s Bartok measure132dB SNR and TT2 127dB?
And some others even boast 144dB SNR?
Isn´t SNR very important in DAC design?

Bartok apparently works fine without the need for add ons like expensive third party BNC cables or separate upscaler boxes like the Mscaler, to function at its best. And has enough power/current to effortlessly drive even the most difficult to drive SOTA headphones on the market with ultimate transparency according to some reviews and no lack of power/current even with non compressed large scale classic material?
And even includes streamer/player functions.
I am asking this with the experience that yes, Dave/HMS is about as good as I´ve heard digital sound, and NO I have not auditioned the Bartok yet.

But the just posted comparison between TT2/HMS and Bartok with HEDDPHONE direct to both over at Headphonics.com makes want to do just that!

The now more than 5 year old Dave to my ears definitely needs an external headphone amp with Susvara the best planar headphone I´ve heard.
But with headphone amps basically in the same price range as Dave itself, I was very impressed indeed.
But only then!

Will your next SOTA product like some competitors already allow,have volume adjustment that works at half dB steps instead of as now 1 dB? Imho the Devil lives in the details, and 1dB up or down in volume is quite a step isn´t it?

Will it have 40 elements intead of Dave´s current 20?
I suspect that the obvious improvement Dave brings over the other three Mscaler capable dacs lies in partly in that ingredient.

Will it upscale even higher than the current 32/768 Mscaler does?
Will it have more taps than Mscaler´s 1M?
And if so how many more?

I am still thinking of something you said at Canjam in Singapore 2018, ie the number of taps needed to fully resolve 24bits would according to digital theory be 256M taps and you did not think that would be achievable in your lifetime.

Any news to share as far as number of taps are concerned?

Having said all this ,for me personally, Mscaler 1 M taps has been the most transformational SQ improvement with digital I have yet heard.

Finally since some of the competition like dCS allow for firmware upgrading I hope that whatever you´ve got up your sleeve will allow me to keep my 1M taps of my Mscaler and only need to add those to whatever number the new products bring.

I still don´t understand why the 49000 or 94000 taps of Qutest and TT2 or the 146000 taps of Dave had to be dumpted with Mscaler?

PS Whatever is in store ,I hope we will be spared both more annoying little coloured balls and 80s Star Treck design.
Cheers CC
 
Last edited:
Mar 29, 2021 at 10:05 AM Post #2,234 of 4,753
I am asking this with the experience that yes, Dave/HMS is about as good as I´ve heard digital sound, and NO I have not auditioned the Bartok yet.
I have heard the Rossini DAC.
It was at Windsor Hi-end show a few years ago, partnered with Audio Research valve amps and Wilson Audio Yvette speakers, I stayed for two sessions of about 20 mins per session.
That has so far been the most awesome sound I ever heard to this day.
The previous year, in a similar setup (different Wilson Audio speaker), a Metronome DAC was used instead. Although it sounded excellent, but the Rossini was better, noticabley.
Frankly I do understand, that this is "Watt's up" thread, and Mr. Watts pride in his work (deservedly) - however to blatantly declare "I am the best and others can not compete" is a little hard to swallow.
Linn is a very professional manufacturer - a Hi-end manufacturer going back a bit longer than 33 years and they too have devoted a lifetime of experience in their craft.
I doubt that they just " Cobble things together" , stick their name on it and sell it for thousands - they have lasted this long because of the quality of their products.
Snake-Oil Merchants ?? they are not.
 
Mar 29, 2021 at 10:07 AM Post #2,235 of 4,753
I don't know where you are getting the numbers for dCS Bartok, but looking at dCS website they quote:

residual noise
24-bit data: Better than –113dB0, 20Hz - 20kHz unweighted. (6V output setting)

That number is flat, so AWt would be around -116dB or nearly 4 times more noise than TT2.

As to your assertion that Bartok will drive anything, then again look at their specs:

headphone outputs
1 stereo balanced pair on 1x 4-way male XLR connector. 1 stereo unbalanced pair on 1x 6.35mm (1/4”) 3-pole jack. Full-scale output levels are 1.4W rms into 33Ω, 0.15W rms into 300Ω. Output levels are 0, -10, -20, -30dB, set in the menu. Minimum headphone impedance is 33Ω.

So you can't use a load below 33 ohms! Also, 33 ohms is only 1.4W - against TT2's 10W into 33 (balanced OP). Dave will match Bartok in drive from the SE headphone OP, but can deliver 0.5A RMS so will drive loads down to 8 ohms.

SNR or residual noise of -127 dB is perfectly fine, and will not create SQ problems - so long as the noise is fixed. But of course all non chord DACs do not have a fixed noise level, as it increases with higher signal voltages. Chord DACs are unique in that the noise is fixed - no noise floor modulation, and this has huge SQ benefits. I could easily improve the noise levels - but to do so would increase distortion. I am not prepared to sacrifice sound quality for meaningless specs, as 127 dB is inaudible.

As to what I do in the future - it will be governed by one simple rule - whatever works best for the best sound quality. If that's one box or ten, then so be it. I am not prepared to sacrifice SQ to suit convenience - and if you want convenience then there are plenty of other alternatives open to you.
 

Users who are viewing this thread

Back
Top