yes;
except timing
enforces the data, which IS NOT resent (in the case of AUDIO); error correction handles the process, and ultimately
estimations take over.
The less ‘estimations’, the more ‘honest’ a system can be to the source material..
The most obvious improvements can be noted across a range of genres, whether a user is looking for ‘placement of instruments’, or maybe ‘transients’(power/speed), to ‘bass note quality/texture’ etc..
I do not have the ears (training) to consciously spot ‘the attack on the violin’ changes, the way a conductor-who has ear training of the highest calibre, will easily ‘see’.
The changes
are there, and speed is 100% felt in the recording, when the transients are in the right places. This is why a better cable can translate to better musical playback in our systems; as the more accurate passing of zeros and ones (as an analogue wavelength that USB cables utilise), the interpretation of the peaks and troughs (zeros and ones), by TIMING, is crucial to digital data coming out ‘as close to perfect‘ as possible.
Close to perfect?
Thats right; I do not know of, and certainly have never seen a setup that has 1:1 in this regard. Error correction and ‘guestimation’ is the rule of the day... (good enough for our ears, certainly)..
Getting as ‘close to 1:1’ as possible is aided by a cable that passed data ‘wave’ in perfect time alignment,- as otherwise it is standard for the wavelength to wind up with mismatched eyelet reads, where zeros BECOME ones and VICE VERSA....
So the cable itself, by timing errors in the receiving circuitry, will ’create’ errors, and combined with the fact that the data itself
isn’t resent, it is just ‘made up’ to make up for some of the ‘misses’ along the way..
When made up data is near a rock and roll transient peak, it will KILL it! (either extend the peak or miss it, ‘in part’). The many ways in which errors can amass in ways that alter the final sound output; is
many (CD spec allowed for up to 1411 errors per frame(?edit checked this; I think I was getting my CIRC confused with CRC, but I believe I was refering to the 75 sectors per second, each being 2048bytes, where 1400 odd errors, which also had ‘loose tolerances’ as to what is required/acceptable when playing with ‘44 thousand samples per second’ would govern our sound

, and ‘on the fly’ reading of zeros and ones from a round disc (requiring speed adjustment to remain linear/constant) CRC can only compensate so much; beyond that is when the logic would either stop playing (good equipment typically), or ‘be heavily
altered’ (bargain brands that were happy just magically getting it to all work), and would pass it along. Ever wondered why the $3k DVD player stopped on scratched discs but the supermarket DVD player could ‘play anything’?
Those great players had a lot of effort put into them to enforce as ‘close to perfect’ sound as could be had.
Usually involving clocks.
As someone who had a Superclock 2 mod in a Denon DCD-S10 (with 18x backgate cap upgrades), I have always seen the clock chip as ‘part of the triangle’ of things that matter.
Back in the day it was ‘Transport quality’,’DAC’(circuit) and ‘Clock’. This ’triangle‘ relationship, kinda like with cameras; the triangle of ‘lens’, ‘sensor’, and ‘software’, getting any one of the pillars wrong and the others can become near worthless....
Nowadays we have TRANSPORT and DAC and AMP as our ‘typical‘ headfi ‘source’ to sort..
The M11+ is an excellent transport, a great DAC and a good(/average) amp. (i downplay the amp cause it is about right for the pricepoint, perhaps it may have been improved if FiiO had used their typical homebrew,
tuned solutions

). Feeding the M11+ into an outboard amp- it scales way above what the unit alone can do. The amp section is seriously holding back its ‘best performance’ capability..
.. rave aside,;
Timing is key: and is why the worlds worst USB cable (on the basis that digital audio consists of zeros and ones), isn’t a good idea.