It’s rather amazing to read so many threads and posts on this forum that debate the merits of two opposing schools of thought; 1) digital is perfection, bits are transmitted perfectly (or certainly after any needed error correction) and hence obsessing over digital cables (ethernet, usb, optical) ethernet network switches, etc. is a waste of time and money, e.g. audiophoolery, or 2) digital music reproduction systems are susceptible to many of the same issues observable in analog audio systems and cables, switches, power supplies, clocks, etc. can all have a material impact sound quality.
Bits will always be bits and are seldom messed with…the digital environment is pretty robust.
It’s the amount of data processing and emi/rfi associated with this that can pollute the analogue infrastructure that is much more sensitive to interference - that is why everything matters.
No doubt this thread will end up like all the others with such good intentions to just have open minded discussions.
Unfortunately I believe that (a) bits is bits in digital domain, and (b) all those factors like lps, interconnects, etc can matter a very meaningful amount the moment you are in the analog domain. And DACs have a portion of their functionality in the digital domain, and a C portion in the analog domain. So the way the question is phrased, I’m a little bit both. But in the sense I think you mean the question, I’m decidedly “bitsisbits”.
Digital to digital is like having a printed menu in a restaurant given to you and when the waiter comes over you point to the items you want and you get the message across to him to get your order.
interference in the analogue domain is like when the waiter asks how do you want your steak cooked and you say audibly how you like it - as medium rare. Now depending on on the amount of other people also ordering and chatting away in the same restaurant he might hear just rare or just medium, and while your steak will arrive it might not be quite the way you ordered it…and even if the waiter got it wrong maybe the chef messed it up anyway.
some will happy to just eat it anyway…others will send it back and wait some more. personal choice.
Imagine you have a factory with a production chain that produces an end product.
Needed materials are gonna be delivered in packages. It’s the materials that go into the production chain, in our case in digital material.
Ideally the material delivered comes in parcels without any dirt from the transport or former packing. But if there is dirt in and in the packages, you need to clean it before you put it in the machines of the production chain. Otherwise you would have the dirt in the end product, although the materials delivered themselves are always (bit)perfect.
If bits are just bits, then NOTHING before the digital to analog conversion in the DAC matters. NOTHING! That would include the input used to get the bits into the DAC.
My personal experience has found that to not be the case.
Yes, I’d like for there to have been a third option of ‘I don’t know.’
2 Likes
Bill_Janssen
(Wigwam wool socks now on asymmetrical isolation feet!)
10
Classic false dichotomy. Sure, bits are bits, but there are other things which matter, too. Not everything matters, though. Reality is gray, not black or white. Something we have a hard time with these days, I guess.
There is a receiver (different for any type of input above) in the DAC. What type? What brand? How is it implemented?
That receiver is in an electronic design. How is that design? Route on the PCB, components used. How well can the receiver cope with electrical noise produced by the PC, streamer and cable. etc. etc.
That electronic design is fed by a power supply. What kind of power supply? One mutual for analogue and digital side or separated. Switched or linear. Good or bad design. Cheap or expensive transformer.
What DAC chips are used and how are they implemented? 1,2 4, 8 … Balanced, unbalanced.
How is the analogue filtering done?
How is the clocking done?
How is the temperature stability / dependency? PPM of the DAC chip(s) and other components
And the list goes on and on and on.
So bits are bits until they reach the DAC. After that point it is totally up to the designer of the DAC with ALL components involved.
I totally agree, bits are bits until they reach the DAC. The question here refers to the whole chain: “digital cables (ethernet, usb, optical) ethernet network switches, etc.”
Bits are bits, so is a molecule of water. A molecule of water is a molecule of water, but once you get a cup of water, or a pale of water and you feed into a rubber hose, the way the water(a random large collection of water molecules) flows in a rubber hose is influenced by so many factors, that’s what fluid dynamics is all about. The water pressure in the hose can vary so much due to many factors.
Funny how people find ways to compare dataflow with something in the analog domain.
When water flows, it will change its consistency, sequence, structure, composition and the result in the end is unpredictable, chaos.
When data flows, it will not change a bit (pun intended), the result is predictable.
Data flow is encoded in bits, the tiniest of bit is in reality an electron negatively charged, hence data is 0 or 1 where 1 is the presence of an electron and 0 is the absence of an electron. (see What Exactly is in a 1 bit of digital memory? https://medium.com/coderscorner/what-exactly-is-in-a-1-bit-of-digital-memory-d5395f9001a6 )
So 1 amp of current is equivalent to 6.24 x 10¹⁸ electrons passing through a single point within 1 second. So the analogy of water molecules is quite close.
