This largely depends on the DAC and the amount of jitter that the DAC sees on it’s (clock) source. Jitter is the timing deviation in a digital or clock signal. It’s not only about 1’s and 0’s, it’s about when they arrive and if the DAC’s internal clocking circuitry has the “time” to correct or ignore this timing inconsistency. A DAC chip that is fed a jittery clock source has higher distortion. This can definitely be measured and perceived.
Digital audio signals such as SPDIF or AES-EBU carry both the timing and data information. Conceptually, you could hook up a DAC with a separate wordlock generator (if it has that feature). In this scenario, jitter of the source device (considering it’s also locked to the same wordclock) is now irrelevant and as long as the data signal comes in a timely manner, the DAC’s operating quality will be solely dependent on the jitter of the wordclock generator. The source is effectively only sending data, not timing information.
In practice most modern DACs have a very stable internal clock generator that gets locked to a wordclock or AES/SPDIF clock source in a very forgiving manner. In other words, a good DAC is not affected by jitter and performs at the same level of distortion regardless of clock source. In a lower quality DAC differences of the connected source could theoretically be perceived in the same manner that the same DAC may operate when connected to a jittery clock.
With USB-audio things get more complicated, but I think you’re mostly interested in the concept of why a digital source may matter for sound reproduction quality rather than the exact details.
I honestly don’t know. Price point is hardly relevant. I’m more into pro audio and have noticed that on recent higher end devices clock source makes little difference anymore.
It’s a design choice to trade of stability for jitter rejection, not necessarily a matter of cost. Digital audio products have used the same type of conversion chips for years and we have seen a lot of newer designs lately that include onboard digital PLLs or manage this problem by SRC. These ICs may be expensive but I don’t know how that is reflected in standalone DAC devices.
Tbh, I think any digital source connected to a decent DAC will give you decent results. I’m just trying to explain that in digital audio, the entire system is to be considered. A digital source may not have a “sound”, that does not mean the next device in the chain will perform optimally which may unfairly contribute to the “sound” attributed to a certain digital source device.