iFi Audio iGalvanic3.0 - The Holy Grail of Audio
Nov 16, 2023 at 6:36 AM Post #724 of 745
My understanding is that in the following chain configurations:
A) Computer---(USB)---DAC, the digital signal is timed by the DAC that serves as the master;
B) Computer---(USB)---DDC---DAC, the digital signal is timed by the DDC that serves as the master. This is so provided that the USB connection is asynchronous.

Now, If one inserts an iFi iGalvanic3.0 in the A chain (between the two devices), is it correct to assume that the iGalvanic3.0 would become the master (clock-wise), similar to what the DDC becomes in B chain, since it does not only offer galvanic isolation but also Reclocks/Rebalances/Regenerates the signal?
I've read that iGalvanic3.0 is transparent/invisible to the computer (does not appear in the list of connected devices); does this mean that it will not become the master, however will somehow still fix the signal's timing and pass the better digital signal to the DAC, the latter of which still is the master in terms of signal timing?
 
Last edited:
Nov 16, 2023 at 7:27 AM Post #725 of 745
My understanding is that in the following chain configurations:
A) Computer---(USB)---DAC, the digital signal is timed by the DAC that serves as the master;
No, there’s actually two digital signals. The digital data signal from your computer to your DAC and then the digital signal from your DACs input to the input of your DAC’s DAC chip. The timing of the former is controlled by the computer and the timing of the latter is controlled by your DAC. In the case of asynchronous USB transfer, the timing of the former is completely independent and unrelated to the timing of the latter. In the case of synchronous USB transfer, the timing of the former is not independent or unrelated to the latter but is modified by your DAC’s clock recovery circuitry (typically involving some sort of PLL) after the signal is received, before the data is then passed to the input of the DAC chip.
B) Computer---(USB)---DDC---DAC, the digital signal is timed by the DDC that serves as the master. This is so provided that the USB connection is asynchronous.
As above, the USB digital signal is “timed” by the computer. The DDC then re-clocks it to the timing of the digital protocol its outputting. Your DAC then re-clocks it again in the case of an asynchronous protocol or modifies it with its clock recovery circuitry in the case of a synchronous protocol. This is the same as above, which is obvious if you think about it, as your DAC obviously doesn’t know (or care) whether it’s input signal has come from a computer, DDC or any other digital audio device.

The clock master of the digital signal entering your DAC chip (the only place it matters) is effectively ALWAYS your DAC, whatever your digital audio protocol/s or digital devices you have in the chain. The only difference is that your DAC is only modifying the embedded clock with synchronous protocols, rather than generating it from scratch in the case of asynchronous protocols.

G
 
Last edited:
Nov 16, 2023 at 7:38 AM Post #726 of 745
No, there’s actually two digital signals. The digital data signal from your computer to your DAC and then the digital signal from your DACs input to the input of your DAC’s DAC chip. The timing of the former is controlled by the computer and the timing of the latter is controlled by your DAC. In the case of asynchronous USB transfer, the timing of the former is completely independent and unrelated to the timing of the latter. In the case of synchronous USB transfer, the timing of the former is not independent or unrelated to the latter but is modified by your DAC’s clock recovery circuitry (typically involving some sort of PLL) after the signal is received, before the data is then passed to the input of the DAC chip.

As above, the USB digital signal is “timed” by the computer. The DDC then re-clocks it to the timing of the digital protocol its outputting. Your DAC then re-clocks it again in the case of an asynchronous protocol or modifies it with its clock recovery circuitry in the case of a synchronous protocol. This is the same as above, which is obvious if you think about it, as your DAC obviously doesn’t know (or care) whether it’s input signal has come from a computer, DDC or any other digital audio device.

The clock master of the digital signal entering your DAC chip (the only place it matters) is effectively ALWAYS your DAC, whatever your digital audio protocol/s or digital devices you have in the chain. The only difference is that your DAC is only modifying the embedded clock with synchronous protocols, rather than generating it from scratch in the case of asynchronous protocols.

G
Very interesting, first time I read about this distinction, this duality of the data audio signal within a DAC.

From your last sentence, is it correct to assume that it's better for the DAC to generate the clock timing from scratch rather than modifying the embedded clock with synchronous protocols?
 
Nov 16, 2023 at 7:53 AM Post #727 of 745
Very interesting, first time I read about this distinction, this duality of the data audio signal within a DAC.
That‘s probably not really the case, just that you didn’t realise it. USB 2.0 for example has fixed data rates up to 480Mbps, obviously that’s very different from the 1.4Mbps of say 16/44 digital audio, so a different clock is required for those two very different data rates.
From your last sentence, is it correct to assume that it's better for the DAC to generate the clock timing from scratch rather than modifying the embedded clock with synchronous protocols?
That very much depends on the DAC. Some DAC’s have very good clock recovery circuitry and there’s no difference between generating the clock itself or modifying (recovering) an embedded or external clock. In many cases though, your assumption would be correct, performance would be better when the DAC generates the clock from scratch. However, unless there’s something pretty seriously wrong with the DAC, this “better performance“ would be on paper only, it would not be audibly different/better.

G
 
Nov 16, 2023 at 8:14 AM Post #728 of 745
That‘s probably not really the case, just that you didn’t realise it. USB 2.0 for example has fixed data rates up to 480Mbps, obviously that’s very different from the 1.4Mbps of say 16/44 digital audio, so a different clock is required for those two very different data rates.

That very much depends on the DAC. Some DAC’s have very good clock recovery circuitry and there’s no difference between generating the clock itself or modifying (recovering) an embedded or external clock. In many cases though, your assumption would be correct, performance would be better when the DAC generates the clock from scratch. However, unless there’s something pretty seriously wrong with the DAC, this “better performance“ would be on paper only, it would not be audibly different/better.

G
Thanks for taking the time to elaborate. Sure, there are different quality DACs on the market. I'm trying to determine numerous things implicating clock/signal timing, including:
1. Whether a USB asynchronous connection DAC-to-computer would be preferable over a synchronous one (e.g. via Toslink found on some Apple computers for example) due to the fact that computer timing accuracy is reportedly not up to the level of an audiophile audio system (on the other hand, an optical Toslink connection would have the advantage of not carrying electrical noise).
2. If a DDC or other similar device (e.g. iFi iUSB Micro or Nano or even iGalvanic) controls timing then it has to be superior than the DAC's internal clock circuitry to be worth using it.
3. If a CD player (used as a CD transport) has often inferior clock to a DAC (and given it's connected via synchronous coax or toslink to the DAC, thus being the master) then it will degrade the DAC's otherwise superior (typically) timing and put to the test its clock recovery circuitry.
 
Last edited:
Nov 16, 2023 at 9:03 AM Post #729 of 745
Sure, there are different quality DACs on the market.
True, but then “different quality” can mean almost anything. For example, a high quality, very expensive DAC does not necessarily mean it has better clock recovery circuitry than a far cheaper DAC. It may just have a far more expensive case and/or attempt a non-conventional topology, that may actually have no better or even poorer performance.
I'm trying to determine numerous things implicating clock/signal timing, including:
1. Whether a USB asynchronous connection DAC-to-computer would be preferable over a synchronous one (e.g. via Toslink found on some Apple computers for example) due to the fact that computer timing accuracy is reportedly not up to the level of an audiophile audio system (on the other hand, an optical Toslink connection would have te advantage of not carrying electrical noise).
Computer timing accuracy with synchronous transfer is irrelevant, as it will be modified/corrected by the DAC’s clock recovery circuitry. The difference between that and asynchronous transfer will be inaudible unless you happen to have a DAC with particularly horrendous clock recovery circuitry. Not carrying electrical noise/interference should also be irrelevant as inherently digital audio is immune to it, the only exception would be a need for cables longer than the protocol recommends or again, an incompetently designed DAC that doesn’t prevent electrical noise/interference on the digital input from somehow seeping into the analogue side of the DAC.
2. If a DDC or other similar device (e.g. iFi iUSB Micro or Nano or even iGalvanic) controls timing then it has to be superior than the DAC's internal clock circuitry to be worth using it.
It doesn’t control the timing though. It only controls the timing from its output to the input of the DAC. The timing up to the input of the DDC (or other device) is controlled by the computer and the timing of the signal entering the DAC chip is controlled/modified by the DAC.
3. If a CD player (used as a CD transport) has often inferior clock (and given it's connected via synchronous coax or toslink to the DAC, thus being the master) to a DAC then it will degrade the DAC's otherwise superior timing and put to the test its clock recovery circuitry.
True, but again, the clock recovery circuitry would have to be really horrendous for that degradation to be audible. Eg. Significantly worse than the clock recovery circuitry found in relatively cheap consumer devices from around 30 years ago.

G
 
Nov 16, 2023 at 9:09 AM Post #730 of 745
True, but then “different quality” can mean almost anything. For example, a high quality, very expensive DAC does not necessarily mean it has better clock recovery circuitry than a far cheaper DAC. It may just have a far more expensive case and/or attempt a non-conventional topology, that may actually have no better or even poorer performance.

Computer timing accuracy with synchronous transfer is irrelevant, as it will be modified/corrected by the DAC’s clock recovery circuitry. The difference between that and asynchronous transfer will be inaudible unless you happen to have a DAC with particularly horrendous clock recovery circuitry. Not carrying electrical noise/interference should also be irrelevant as inherently digital audio is immune to it, the only exception would be a need for cables longer than the protocol recommends or again, an incompetently designed DAC that doesn’t prevent electrical noise/interference on the digital input from somehow seeping into the analogue side of the DAC.

It doesn’t control the timing though. It only controls the timing from its output to the input of the DAC. The timing up to the input of the DDC (or other device) is controlled by the computer and the timing of the signal entering the DAC chip is controlled/modified by the DAC.

True, but again, the clock recovery circuitry would have to be really horrendous for that degradation to be audible. Eg. Significantly worse than the clock recovery circuitry found in relatively cheap consumer devices from around 30 years ago.

G

Not carrying electrical noise/interference should also be irrelevant as inherently digital audio is immune to it
There are be many (knowledgeable) people who would disagree on that statement. :)

Thanks for all the time to reply, I appreciate it.
 
Nov 16, 2023 at 9:31 AM Post #731 of 745
There are be many (knowledgeable) people who would disagree on that statement. :)
As this is an audiophile forum, you can be pretty much assured some/many people would disagree, even amongst those claiming to be knowledgeable. However, their agreement or disagreement doesn’t change the actual facts of why digital audio was invented.

G
 
Nov 28, 2023 at 4:26 AM Post #732 of 745
i`d like to know if the device itself requires usb power to work, or it will still work if you use a custom usb cable that is data only, no 5v power ?
im referring to the connection from the pc to the igalvanic, does the device require the 5v power that comes thorough the cable ?
 
Last edited:
Nov 28, 2023 at 4:28 AM Post #733 of 745
i`d like to know if the device itself requires usb power to work, or it will still work if you use a custom usb cable that is data only, no 5v power ?
The iFi iGalvanic3.0 uses a USB 3.0 cable to connect to the source via which it receives both data and power (for its operation). Just use a USB 3.0 cable, does not have to be high-end (e.g. Amazon Basics would be fine).
 

Users who are viewing this thread

Back
Top