[Igor375375]

Please share any news why Gaia has better USB implementation than Amanero.

I'm not sure I'm right. I like my Amanero, just looking into the issue for now. Do you recommend in addition to my Audio GD R-28 ( 2022 ) to buy a DI20HE rather than a GAIA ?

 

[gimmeheadroom]

R-28 will benefit from DI-20 (not to mention HE version), as it [sic the R28] doesn't have isolation on the USB port. Galvanic USB isolation is only offered in top-end R-8/R-7 MK2 models. These are identical in design with DI-20, but still benefit from DI-20, it is other issue.

Agree with @sajunky, especially if you have Audio-gd gear, DI-20HE is the best possible DDC. If you don't want to spend the extra money, the DI-20 is a great value.

I'm not sure I'm right. I like my Amanero, just looking into the issue for now. Do you recommend in addition to my Audio GD R-28 ( 2022 ) to buy a DI20HE rather than a GAIA ?

Yes, that's what we're saying.

 

[Igor375375]

Agree with @sajunky, especially if you have Audio-gd gear, DI-20HE is the best possible DDC. If you don't want to spend the extra money, the DI-20 is a great value.

• Thank you )))

 

[jimmychan]

For those who are interested in the Audio-GD DI-20HE vs GAIA, you could find the test result here:

https://goldensound.audio/2021/10/04/denafrips-gaia-measurements-with-terminator-clock-sync-test/

" Overall, the Gaia is an excellently performing DDC.
It falls ever so slightly behind some other choices like the DI20HE and SU6 in performance, but the added features, and matching aesthetics with Denafrips dacs may be more important to many."

I hope this will help.

 

[gimmeheadroom]

"The test results" sounds a little too enthusiastic and final. Maybe "a review" would be more accurate. There are endless review sites and opinions and those comments are far from definitive.

 

[JaMo]

Yes, A sober compare by listening to the devices is always the way to go. I haven't heard the Gaia but I am sure it is a nice re-clocker/digital converter. If a Master clock is an important part in Your rig and Your rigs performance, I think that the Gaia has strange choices of external clock inputs with 49.152 MHz and 45.1548 MHz. They are not alone with this but it is a bit unusual. The DI20/HE re-clocks USB and SPDIF based on a good internal XO's but the wonders happens when a really good external XO of 10MHz (Master clock) is connected and conducts the "duck march of bits". When we are at it I think a quick look at the Gustard U18 DDC is relevant. It does a really good job connected to a 10MHz Master clock. The U18 is an Only-USB digital interface but a great one for the money (<USD500 with global delivery)
my 2cents/J

 

[sajunky]

If a Master clock is an important part in Your rig and Your rigs performance, I think that the Gaia has strange choices of external clock inputs with 49.152 MHz and 45.1548 MHz. They are not alone with this but it is a bit unusual. The DI20/HE re-clocks USB and SPDIF based on a good internal XO's but the wonders happens when a really good external XO of 10MHz (Master clock) is connected and conducts the "duck march of bits".

Gaia external clock ports only for works with Terminator, it is their own solution to avoid clock synthesiser. It makes a set not upgradable, clocks on the Terminator are final. All others have accepted 10MHz clock as a standard external source, it gives flexibility. In the case of R-28 it doesn't make difference (there is no external clock input), but for a quality of power supply it is difficult to beat DI-20, not to mention HE version, as it is far above competitors.

I have to say that neither R-28 v2022 nor any model of DI-20 (HE) do reclocking on the USB port. The output of DI-20 is driven directly from one of the internal oscilators and USB host transfers are synchronised to these internal clocks. The same with R28 2022 version. Internal clocks drive R2R ladder directly. In theory, R-28 do not need a DDC, as there is no added jitter on the USB connection, right? It is a prime textbook example where it should be no difference, but a quality of power supply matters. Getting rid of ground loops noise is the main purpose of a DDC. All reclocking or no reclocking or how it is clocked debate is a secondary issue.

 

[gimmeheadroom]

Gaia external clock ports only for works with Terminator, it is their own solution to avoid clock synthesiser. It makes a set not upgradable, clocks on the Terminator are final. All others have accepted 10MHz clock as a standard external source, it gives flexibility. In the case of R-28 it doesn't make difference (there is no external clock input), but for a quality of power supply it is difficult to beat DI-20, not to mention HE version, as it is far above competitors.

I have to say that neither R-28 v2022 nor any model of DI-20 (HE) do reclocking on the USB port. The output of DI-20 is driven directly from one of the internal oscilators and USB host transfers are synchronised to these internal clocks. The same with R28 2022 version. Internal clocks drive R2R directly. In theory, R-28 do not need a DDC, as there is no added jitter on the USB connection, right? It is a prime textbook example where it should be no difference, but a quality of power supply matters. Getting rid of ground loops noise is the main purpose of a DDC. All reclocking or no reclocking or how it is clocked debate is a secondary issue.

But, I guess non-USB connections on the R28 could benefit from reclocking i.e. AES/EBU, coax, etc.

I don't have a 10 MHz clock to drive my DI-20s, but in fact running the DI-20/HE to the Mutec MC-3+ USB over AES does add some sparkle and clarity/air.

 

[sajunky]

But, I guess non-USB connections on the R28 could benefit from reclocking i.e. AES/EBU, coax, etc.

I don't have a 10 MHz clock to drive my DI-20s, but in fact running the DI-20/HE to the Mutec MC-3+ USB over AES does add some sparkle and clarity/air.

Yes, USB is only one interface with asynchronous data delivery (slaved to a DDC/DAC clock). All other interfaces follow a source clock, synchronising its own clock to the source clock. It means that a final quality depends on the source clock.

With Mutec you can also feed DI-20 with external clock from Mutec. It helps avoiding use of PLL for clock synchronisation. To be exact, PLL is still used for decoding serial data, but a final clock synchronisation is based on the external clock input. Also an ideal scenario, almost zero-jitter, similar result to the asynchronous data delivery method.

 

[gimmeheadroom]

Yes, USB is only one interface with asynchronous data delivery (slaved to a DDC/DAC clock). All other interfaces follow a source clock, synchronising its own clock to the source clock. It means that a final quality depends on the source clock.

With Mutec you can also feed DI-20 with external clock from Mutec. It helps avoiding use of PLL for clock synchronisation. To be exact, PLL is still used for decoding serial data, but a final clock synchronisation is based on the external clock input. Also an ideal scenario, almost zero-jitter, similar result to the asynchronous data delivery method.

You meant the REF10 Mutec clock? It's on my list but it might be a while, that's a lot of money.

 

[sajunky]

You meant the REF10 Mutec clock? It's on my list but it might be a while, that's a lot of money.

Do MC-3+USB have 10MHz BNC clock output? You can use. If not, you can use any 10MHz external clock to synchronise both MC-3+USB and DI-20 with the same clock source.

 

[gimmeheadroom]

Do MC-3+USB have 10MHz BNC clock output? You can use. If not, you can use any 10MHz external clock to synchronise both MC-3+USB and DI-20 with the same clock source.

It does not, but it has a 10 MHz clock input- but I don't have a 10 MHz clock. It has multiple clock outputs but it's WCLK and multiples, 10 MHz isn't possible.

I do use the DI-20 word clock into my Brooklyn DAC+ which is kinda cool. Now that I think of it, I don't know if I saw a spec for the DI-20/HE clock output. Do you know if it's 50 or 75 ohms? A 75 ohm BNC connector works fine.

I did not try running the DI-20HE WCLK into the Mutec, I assumed the Mutec internal clock would be better. But it might be a good idea to use the DI-20HE as a master clock when recording with one of my TASCAMs off the Mutec which runs off the DI-20HE...

 

[sajunky]

It does not, but it has a 10 MHz clock input- but I don't have a 10 MHz clock. It has multiple clock outputs but it's WCLK and multiples, 10 MHz isn't possible.

I do use the DI-20 word clock into my Brooklyn DAC+ which is kinda cool. Now that I think of it, I don't know if I saw a spec for the DI-20/HE clock output. Do you know if it's 50 or 75 ohms? A 75 ohm BNC connector works fine.





I did not try running the DI-20HE WCLK into the Mutec, I assumed the Mutec internal clock would be better. But it might be a good idea to use the DI-20HE as a master clock when recording with one of my TASCAMs off the Mutec which runs off the DI-20HE...

DI-20 clock inputs/outputs have impedance 50 Ohms, but signal levels are different. The input follows 10MHz standard, the output use LVDS logic levels. For me it is strange, as it is not differential output, but it says such way in the manual:

CLK OUT ：
Outputs 256fs main clock or LRCK/WCLK (3.3V LVDS @ 50 Ohm) . User-selectable .
INPUT S/PDIF ：
RCA 75 Ohm coaxial audio input (0.5VPP @ 75 Ohm) .
INPUT 10MHz ：
External 10MHz clock input (0.3VPP -3VPP @ 50 Ohm)

75 Ohm BNC connector looks very similar to the 50 Ohm, they mate together, but there are different. Some companies mix them allowing user configurable option, but it is not a proper way. When it is for feeding a clock, impedance matching must be followed from the source through cable to the receiver. If it is impossible 100% matching, the a cable must be very short.

You cannot use DI-20 clock output as a master clock source for the upstream device (Mutec in this case), as it is dependent on the input sample rate. Such feedback could result in instability. The clock output in DI-20 is solely designated for the downstream devices like a DAC. It would be great to have a standard 10MHz clock output, it could be used in both directions.

 

[gimmeheadroom]

DI-20 clock inputs/outputs have impedance 50 Ohms, but signal levels are different. The input follows 10MHz standard, the output use LVDS logic levels. For me it is strange, as it is not differential output, but it says such way in the manual:


75 Ohm BNC connector looks very similar to the 50 Ohm, they mate together, but there are different. Some companies mix them allowing user configurable option, but it is not a proper way. When it is for feeding a clock, impedance matching must be followed from the source through cable to the receiver. If it is impossible 100% matching, the a cable must be very short.

Yes, I know. But I didn't find the spec until you posted it. Thanks for that.

You cannot use DI-20 clock output as a master clock source for the upstream device (Mutec in this case), as it is dependent on the input sample rate. Such feedback could result in instability. The clock output in DI-20 is solely designated for the downstream devices like a DAC. It would be great to have a standard 10MHz clock output, it could be used in both directions.

I did not find the clock spec on the specs page, but I found the info you quoted on the user manual page Indeed, the DI-20's clock out worked fine with the Brooklyn (even if the wrong impedance) until I played DSD.

These incompatibles are annoying. It would be nice if the hifi companies would use the pro standard 75 ohms.

 

[sajunky]

These incompatibles are annoying. It would be nice if the hifi companies would use the pro standard 75 ohms.

75 Ohms is a standard for TV Antenaes, security systems. It was also adopted by S/PDIF many years ago, but implementation is lousy, even RCA connector is completely inadequate.

50 Ohm BNC connector is a generic type for radio communication in military, HAM, general industry, measuring equipment like oscilloscopes, e.t.c. There are quality cables on the market for this standard, even those conforming to the strict military specs, you don't need to buy "audiophile type" for $xxx. It is why 50 Ohm BNC is adopted for 10 MHz clock distribution in audio, I am very happy with.