Article: Audiophile Style Products of the Decade

[vortecjr]

Chris those are optical Toslink. 

[vortecjr]

1 minute ago, Matias said:

MSB has been isolating their DAC inputs with their proprietary ProISL fiber as well.

 

http://www.msbtechnology.com/accessories/prousb/

In this category (expensive/proprietary) you could also include Playback Design. What i’m working will be inexpensive and we may release it as open hardware...TBD.

[Qhwoeprktiyns]

ECDesigns will be affordable as well. Their adapters will be compatible with toslink inputs of most DACs. 

 

Can you tell us more about your project ? 

[The Computer Audiophile]

14 minutes ago, Matias said:

MSB has been isolating their DAC inputs with their proprietary ProISL fiber as well.

 

http://www.msbtechnology.com/accessories/prousb/

 

Or Bel Canto Black system went further and its "preamp" converts everything to 2 fibers and sends to the dac+amplifier monos.

 

http://www.belcantodesign.com/home/black/the-system/

 

 

Ayre has used optical isolation inside DACs. If I remember right, they said it has a small negative effect. I could be wrong though. 

 

 

 

6 minutes ago, hopkins said:

ECDesigns will be affordable as well. Their adapters will be compatible with toslink inputs of most DACs. 

 

Can you tell us more about your project ? 

 

Toslink is so limited. I don't think I'd go back to it. 

[Qhwoeprktiyns]

6 minutes ago, The Computer Audiophile said:

 

Ayre has used optical isolation inside DACs. If I remember right, they said it has a small negative effect. I could be wrong though. 

 

 

 

 

Toslink is so limited. I don't think I'd go back to it. 

 

Well, that is another interesting debate

For the coming decade I believe we will see "source immunity" and "format immunity" meaning DACs capable of offering similar quality for Redbook and high res. Let's see in 10 years (if we are still around) what will be in your "products of the decade".

 

Vis à vis the négative effects of optical, they really depend on the implementation. The toslink cable is also a source of jitter signal degradation from what I understand. 

[Matias]

5 minutes ago, The Computer Audiophile said:

 

Ayre has used optical isolation inside DACs. If I remember right, they said it has a small negative effect. I could be wrong though. 

Yes, QB-9 DSD started with optical isolation of the USB port inside the unit. I always wondered why other manufacturers did not also implement this.

[The Computer Audiophile]

Just now, hopkins said:

 

Well, that is another interesting debate

For the coming decade I believe we will see "source immunity" and "format immunity" meaning DACs capable of offering similar quality for Redbook and high res. Let's see in 10 years (if we are still around) what will be in your "products of the decade".

Source immunity would be really cool. The Lumin X1 I just reviewed has fiber Ethernet input.

 

I don't believe it has happened in analog yet. 

[The Computer Audiophile]

Just now, Matias said:

Yes, QB-9 DSD started with optical isolation of the USB port inside the unit. I always wondered why other manufacturers did not also implement this.

Sometimes the cure also has issues not present in the original issue. Perhaps the opto isolator has issues or other manufacturers didn't think it was necessary or the cost was higher. Who knows. 

[vortecjr]

Ayre was using an opto-isolator on the i2s signals generated by the USB receiver.

[Ralf11]

do any DACs other than the Lumin X1 have fiber Ethernet input?

[Jud]

37 minutes ago, Matias said:

Yes, QB-9 DSD started with optical isolation of the USB port inside the unit. I always wondered why other manufacturers did not also implement this.

 

At least some optical isolators are themselves electrically noisy when they do the conversion back to electrical, though I would guess this wasn't the case with the Ayre units.

[Qhwoeprktiyns]

I would like to add something, to "tie" these different comments together. What Chri's 3 selections have in commmon, it seems to me, is "simplicity". As Duke Ellington once said: "Simplicity is a most complex form", and that is true for audio systems as well - there is certainly a lot of complexity involved in the design of these systems. Now lets hope we see even more progess in the coming years, with outstanding audio quality of course.

[Summit]

I agree on Sonore and Roon but not Qobuz, no way. Qobuz isn’t even possible to stream in many countries. I have been a member of Qobuz since 2013 and still can’t stream anything from them. I think that an Audiophile Style Products of the Decade should be something all can get if they want and besides good sound should have excellent service.

[vortecjr]

I started a new thread to discuss optical solution:

 

 

[Ropet]

I wonder why Quboz is not available in the Nordics/Scandinavia? Neither for streaming or downloads. This is the message you get when we try to enter:

 

Qobuz isn't available in your country.

 

Qobuz isn't available in your country.

[barrows]

On 12/20/2019 at 6:31 PM, mansr said:

Such designs can achieve jitter performance very close to that of a plain crystal oscillator, close enough to be of no concern for audio purposes.

While I am well aware of the approach you describe, I have yet to hear an SPDIF interface which sounds as good as a properly implemented USB one (although I would expect that if the async SPDIF receivers, as apparently used by Ayre and dCS to be perhaps be equal if they are truly async).  And I prefer to disable all PLLs and use synchronous clocking in my ESS based DIY DACs for this same reason, better sound.  So with my ESS DACs the masterclock, at 45.1584 MHz is the same clock used for the USB interface (I play only DSD 256, so no need for a 48 KHz base XO clock).  While I understand your belief, and the measurements, which suggest that a PLL based approach, and/or async re-clocking at the DAC approach is of "no concern", my ears tell me something different: that a good (low phase noise at low frequencies) master XO, at the DAC chip, acting as master for both the source (USB in my case) and the DAC (allowing for bit clock and master clock to be in sync) just sounds better.  I can audition both approaches in the same (DIY) DAC, and the synchronous clocking approaching plain sounds better.  Yes, this is subjective evaluation for sure, both approaches yield very good measured results with the ESS 9038 Pro.  Although I readily admit the differences here are are small, but they are significant enough in an audiophile sense for me to have a clear preference.

 

As for I2S I am not a fan of using this interfaces for between boxes as currently implemented by most.  I2S might work really well if it used a master clock at the DAC, and then sent that clock back to the source, but this is not how it is implemented (excepting perhaps MSB's proprietary approach).  Of course if one really believes re-clocking everything into a new clock domain at the DAC and PLLs are perfect, then sure, it would be fine.  And yes, we are phrasing a bit differently...  Since it is easily possible to just have one master, close to the DAC conversion stage (chip or discrete, whatever), I prefer having one master in that location, and sending that master back to the source (whatever it may be) and then re-align everything via a Potato chip and the master clock right there, directly before the conversion.  Then we need no PLLs, async re-clocking, or any other nonsense.  Just the best possible XO at the DAC conversion stage, with everything else synched to that single clock (or two of them if one must deal with the two base frequencies).

[mwhitak]

I'm testing Dante (AOIP) to a Burl DAC (with Dante card installed) as a possible alternative interconnection method to all this USB/SPDIF madness. PC serves files through Dante virtual sound card or dedicated PCIe card, delivered direct over Ethernet (optical isolators used), with everything reclocked at the DAC. There is also an input for an external clock if you want to up the anti.

[The Computer Audiophile]

25 minutes ago, mwhitak said:

I'm testing Dante (AOIP) to a Burl DAC (with Dante card installed) as a possible alternative interconnection method to all this USB/SPDIF madness. PC serves files through Dante virtual sound card or dedicated PCIe card, delivered direct over Ethernet (optical isolators used), with everything reclocked at the DAC. There is also an input for an external clock if you want to up the anti.

Is there a sample rate limit using Dante in your setup?

[audiobomber]

5 hours ago, barrows said:

While I am well aware of the approach you describe, I have yet to hear an SPDIF interface which sounds as good as a properly implemented USB one (although I would expect that if the async SPDIF receivers, as apparently used by Ayre and dCS to be perhaps be equal if they are truly async).  And I prefer to disable all PLLs and use synchronous clocking in my ESS based DIY DACs for this same reason, better sound.  So with my ESS DACs the masterclock, at 45.1584 MHz is the same clock used for the USB interface (I play only DSD 256, so no need for a 48 KHz base XO clock).  While I understand your belief, and the measurements, which suggest that a PLL based approach, and/or async re-clocking at the DAC approach is of "no concern", my ears tell me something different: that a good (low phase noise at low frequencies) master XO, at the DAC chip, acting as master for both the source (USB in my case) and the DAC (allowing for bit clock and master clock to be in sync) just sounds better.  I can audition both approaches in the same (DIY) DAC, and the synchronous clocking approaching plain sounds better.  Yes, this is subjective evaluation for sure, both approaches yield very good measured results with the ESS 9038 Pro.  Although I readily admit the differences here are are small, but they are significant enough in an audiophile sense for me to have a clear preference.

 

As for I2S I am not a fan of using this interfaces for between boxes as currently implemented by most.  I2S might work really well if it used a master clock at the DAC, and then sent that clock back to the source, but this is not how it is implemented (excepting perhaps MSB's proprietary approach).  Of course if one really believes re-clocking everything into a new clock domain at the DAC and PLLs are perfect, then sure, it would be fine.  And yes, we are phrasing a bit differently...  Since it is easily possible to just have one master, close to the DAC conversion stage (chip or discrete, whatever), I prefer having one master in that location, and sending that master back to the source (whatever it may be) and then re-align everything via a Potato chip and the master clock right there, directly before the conversion.  Then we need no PLLs, async re-clocking, or any other nonsense.  Just the best possible XO at the DAC conversion stage, with everything else synched to that single clock (or two of them if one must deal with the two base frequencies).

Any plans for a commercial DAC?

[mwhitak]

1 hour ago, The Computer Audiophile said:

Is there a sample rate limit using Dante in your setup?

Since Dante is a pro audio setup....I believe the maximum is 32 bits/ 192khz.

[The Computer Audiophile]

53 minutes ago, mwhitak said:

Since Dante is a pro audio setup....I believe the maximum is 32 bits/ 192khz.

Yeah, I’m pretty familiar with the similar technologies. I thought there was a 96k limit, but that could’ve been a while ago. 
 

Here is an interesting interview I did with Domonique from Merging about ARS67 and Ravenna etc...

 

 

[mwhitak]

3 hours ago, The Computer Audiophile said:

Yeah, I’m pretty familiar with the similar technologies. I thought there was a 96k limit, but that could’ve been a while ago. 
 

Here is an interesting interview I did with Domonique from Merging about ARS67 and Ravenna etc...

 

 

 

Interesting...thanks Chris ! Perhaps it's time for AudiophileStyle to do a write up on the latest developments in AES67/AOIP ? 

[The Computer Audiophile]

9 minutes ago, mwhitak said:

 

Interesting...thanks Chris ! Perhaps it's time for AudiophileStyle to do a write up on the latest developments in AES67/AOIP ? 

Yeah, that would be good to do. 

[Arg]

I will not endorse as Product of the Decade any product that is not available to large swathes of audio consumers. Please check the global distribution extent of products before making awards.

[The Computer Audiophile]

1 hour ago, Arg said:

I will not endorse as Product of the Decade any product that is not available to large swathes of audio consumers. Please check the global distribution extent of products before making awards.

I certainly hear what you’re saying but I believe this is a pretty large group of people. Certainly not as many people as some other services, but that’s ok in my book. 
 

Austria

Belgium

Ireland

Italy

France

Germany

Luxembourg

Netherlands

Spain

Switzerland

United Kingdom

USA