(22-Feb-2016, 03:23)Pim van Vliet Wrote: [ -> ]Here's something to contemplate. If it's so obvious that an incoming signal needs 'work done', shouldn't a good DAC maker do that themselves? I guess what I'm trying to say is that if (and that's a big if) all DAC makers solved the problems that are now solved by Regen, Mutec etc. There wouldn't be a market for them.
And that theory could also explain why some people hear bigger differences on their equipment than others, after installing such devices. After all, the best DAC makers out there should know how to deal with 'difficult' data.
At this point I think it's worth pointing out that Devialet themselves claim that none of these tweaks in the digital domain are capable of having any impact on the sound. In other words they are already solving the "problem".
Guillaume
(22-Feb-2016, 03:23)Pim van Vliet Wrote: [ -> ]Here's something to contemplate. If it's so obvious that an incoming signal needs 'work done', shouldn't a good DAC maker do that themselves? I guess what I'm trying to say is that if (and that's a big if) all DAC makers solved the problems that are now solved by Regen, Mutec etc. There wouldn't be a market for them.
And that theory could also explain why some people hear bigger differences on their equipment than others, after installing such devices. After all, the best DAC makers out there should know how to deal with 'difficult' data.
The problem of the DAC doing everything can not be always and robustly solved.
There is no reason why the DAC, Devialet in our case, should not be properly isolated from cable borne interference. In my opinion it is a poor design if it is badly affected since it is basic well understood engineering. It is my view that the Devialet is well engineered.
When it comes to jitter things are more complex, re-clocking sometimes comes with unacceptable latency. Firstly there is not good agreement as to where the audibility threshold is with jitter with wildly varying claims. In fact jitter is the only mechanism I am aware of which can cause the output analogue signal to differ from the recorded original in a properly engineered digital system.
Jitter in the record ADC will mean the digital file is -not- an accurate representation of the analogue input, but there is nothing we can do about that.
So, does the Devialet re-clock the input data to an inaudibly low level of jitter without latency, or not?
I have no idea personally, but the Mutec has a very good jitter spec, so there will probably be a theoretical improvement.
Whether the improvement is an audible one is another story, but at least it is a technically valid tweak, which is rare in hifi IME.
Isn't the TotalDac D1 basically an optimised Cubox, together with a very high specification re-clocker? I mention this because with all the debate about re-clocking, one thing that there is very good evidence for is that the TotalDac D1 works splendidly well with a Devialet.
(I am also wondering if all this should move onto the Mutec thread, as we have rather hijacked poor old Antoine's system thread .........)
(21-Feb-2016, 18:01)Our Wrote: [ -> ]Hi Antoine,
Congratulations on your purchase. Hope you will enjoy it :-)
Thanks!
(21-Feb-2016, 19:06)f1eng Wrote: [ -> ]I checked out the Mutec web site after reading Jan's comments here, I bought one too after reading Karl-Heinz Fink's comments on their web site.
I have not used it much yet but I did not think t was a risky purchase given the quality of the engineering.
It does seem like a reversion to D-Premier in any ways! I used my Metric Halo recorder as a firewire to AES/EBU converter on that, and HDMI for TV and films.
I have never tried the USB input on my 800.
Nice, I read that comment too on the Mutec site.
Perhaps a future update will prove the added worth of the 200/250/400/800/900 platform over the old D-Premier platform. So far, today there is none for me personally, not anymore.
[align=left]
(21-Feb-2016, 23:49)thumb5 Wrote: [ -> ]Up to a point, yes. The potential for noise injection is pretty clear. I'm not sure I buy the argument about inconsistencies in the timing of the clock generated by the USB host, especially in the case where the DAC (Devialet) is in asynchronous mode. But I must add it's not something I've experimented with in any listening tests with so I'm only talking about what I would expect from a knowledge of how the hardware and software work.
Since Antoine said he's using the Mutec as a USB to AES/EBU converter, that doesn't say anything (I think) about whether there would be any sound quality improvement to be gained by re-clocking the USB. Again...I might be missing something, that's just my --possibly naive -- reading of the story so far.
I too have trouble understanding what the added benefit of using HQ clocks is for the USB receiver itself. Intona writes it doesn't matter:
Quote:Ahem, this clocking thing on USB is an ambivalent issue that I recognize on many forums. All transceiver and hub chips have their own PLLs inside and they all rely on internal RC oscillators to derive the bit clock. We didn't see any change on eye diagrams of the USB bus whether the transceivers where clocked with +-500ps Jittter or +-5ps Jitter on our LeCroy SDA. Anyway, we took special attention to all design aspects and price economy was never a subject.
Everyone else bringing out audiophile level USB 'optimizers' writes it does matter or does incorporate HQ clocks local to the USB receiver/chipset. I honestly don't know if it does matter but it won't hurt either.
The re-clocking the Mutec does is done to the signal after the USB receiver, I guess the S/PDIF or AES3 signal since those have clock information embedded. Other USB optimizers "repacketize" so receive the packets on input and reconstruct those again for output. This way they "regenerate" the USB signal using a local clock. Their circuitry will also make sure impedance matching occurs and they frequently inject clean(ed up) 5V. The Intona does this as well but also provides full galvanic isolation.
This is what Mutec writes about it what it does for USB:
Quote:USB interfaces are the most commonly utilized devices for digital audio transmission in a computer-based music playback environment. But the convenience and ease of use often times come at a price of lackluster audio quality to the critical ears of high-end consumer audio aficionados and of professional users. It’s no secret that even the best D/A converters are not capable of reproducing and converting audio via USB at the highest possible level, which is why using digital audio inputs such as AES3 or S/P-DIF yields significantly better results and is generally preferred.
We have invested into our own research to uncover the underlying principles and the root of the problem with high-end audio via USB. In a nutshell, electromagnetic interference of any kind, but particularly those caused by audio computers and other IT devices, will induce noise and seriously disturb the audio transmission. In order to eliminate these disturbances and avoid their damaging effect on the conversion process as much as possible we have developed a new signal-isolating, asynchronous USB interface for the MC-3+ Smart Clock USB. It has been designed as a dedicated building block with a power supply that’s independent from the main board of the unit. In addition, the USB interface is galvanically isolated from the other parts of the circuitry and employs its own ultra-low noise audio oscillators that once again are autonomous and isolated from the other processes. In combination with IT-based high-speed isolators we have achieved exceptional reduction of the impact of noise interference within the USB data stream yielding a USB interface that’s virtually immune to any interference problems caused by PCs, laptops or music servers. At the same time, there are no limitations regarding the possible clock rates for playback from your music library with the MC-3+USB supporting all rates possible up to high-speed USB as well as DSD/DoP streams up to DSD/DoP256.
All of these features in combination with the subsequent re-clocking process, proprietary to MUTEC, leads to a significant acoustic improvement of connected digital audio devices that usually manifests itself in increased width of the sound stage, more precise imaging, and with a more musical representation of the recorded space and the instruments within it. Stock consumer audio devices, laptops, and audio computers can now achieve a new level of sonic perfection that was previously reserved for equipment with much higher price tags.
So to sum up it's benefits are:
- galvanically isolating the USB input from the rest of the circuitry to shut off electromagnetic interference from the source
- separate, isolated ultra-low noise audio oscillators for USB circuitry
- conversion to S/PDIF and AES3 and subsequent reclocking using a HQ clock
This may be one as well: pulling the USB circuitry from the DAC, to prevent electromagnetic interference inside the DAC. I've read Berkeley Audio Designs deliberately does not implment USB inside of their DAC's but as an external box because of this.
(22-Feb-2016, 12:22)GuillaumeB Wrote: [ -> ] (22-Feb-2016, 03:23)Pim van Vliet Wrote: [ -> ]Here's something to contemplate. If it's so obvious that an incoming signal needs 'work done', shouldn't a good DAC maker do that themselves? I guess what I'm trying to say is that if (and that's a big if) all DAC makers solved the problems that are now solved by Regen, Mutec etc. There wouldn't be a market for them.
And that theory could also explain why some people hear bigger differences on their equipment than others, after installing such devices. After all, the best DAC makers out there should know how to deal with 'difficult' data.
At this point I think it's worth pointing out that Devialet themselves claim that none of these tweaks in the digital domain are capable of having any impact on the sound. In other words they are already solving the "problem".
Guillaume
Exactly, and if they don't acknowledge possible issues or potential improvements they'll never fix these issues or implement the improvements.
I do know now from experience that the USB receiver inside the Devialet (and/or it's processing after) can be improved upon significantly, the Mutec and the other things I've tried prove this, but Devialet is far from alone in this. Even the highest-end DAC's, especially those designed a few years back (like the Devialet) aren't there yet. It's like S/PDIF in the 90's/early 00's, some of you may remember products like the Genesis Digital Lens and the Theta TLC. Products now obsolete since serious hardware designers today do acknowledge the importance of low jitter designs.
I guess and am sure the same will happen for USB since it seems it's not going away anytime soon.
(22-Feb-2016, 14:04)Confused Wrote: [ -> ]Isn't the TotalDac D1 basically an optimised Cubox, together with a very high specification re-clocker? I mention this because with all the debate about re-clocking, one thing that there is very good evidence for is that the TotalDac D1 works splendidly well with a Devialet.
I think it is.
(22-Feb-2016, 14:04)Confused Wrote: [ -> ](I am also wondering if all this should move onto the Mutec thread, as we have rather hijacked poor old Antoine's system thread .........)
I don't mind.
Although, it may be better if it's (discussed) there so more people can find the information in the future.
(22-Feb-2016, 16:00)Antoine Wrote: [ -> ] (22-Feb-2016, 14:04)Confused Wrote: [ -> ]Isn't the TotalDac D1 basically an optimised Cubox, together with a very high specification re-clocker? I mention this because with all the debate about re-clocking, one thing that there is very good evidence for is that the TotalDac D1 works splendidly well with a Devialet.
I think it is.
I would say that the external power supply is an important component too, this is the same power supply that is used with their über high-end DACs.
The Cubox itself is completely shielded from the rest of the box and of course the connection from Cubox to re-clocker is the totaldac USB filter.
A lot of attention in the totaldac design is paid to minimising noise, including the grounding design. When selecting inputs all other inputs are physically disconnected - you can hear the clicks as you toggle between inputs.
There is also some clever stuff going on with the AES output, I know that when I sent my d1 server back for its free upgrade Vincent "tweaked" this.
Sorry - I'm hijacking the thread and I should know better!
Guillaume
No problem Guillaume, I personally don't consider it hijacking. Also a thorough description gives more deserved credit to the device.
(22-Feb-2016, 17:19)Antoine Wrote: [ -> ]No problem Guillaume, I personally don't consider it hijacking. Also a thorough description gives more deserved credit to the device.
Cool.
By the way when Jan visited me several months ago he brought the previous version of the Mutec along. We did quite a lot of A-B testing (which I detest btw!) and yes the Mutec did make a small noticeable effect. Not enough to go out and buy one mind you, particularly as I would need a second AES/EBU cable... I wonder if the "new" Mutec will make a bigger difference...
Guillaume