Quite possibly Tim
But more likely we have different perspectives of what listening for differences and preference are.
Quick switching is definitely useful for identifying differences, but the key criteria in this situation is the listener being trained to understand and then listen for the parameter-trait.
Even then this just provides a way to analyse subjectively and quantify what is heard, however it does not necessarily equate to one's own preference; as an example what we think we want does not mean it matches our long term preference.
This is also compounded by what we think is better quality; a better approach is not to consider one better than the other but to analyse and quantify the differences against a true reference.
If looking to buy a product then considering preferences also has to look at one's self perception-satisfaction-cognitive behaviour, and not just to identify what one thinks has better bass-imaging-tonality-more realistic-etc.
In this way we are more sensitive to aspects that are a type of cognitive dissonance to us as an individual, and are as important as the initial enjoyment when considering long term listening.
That aside, also it is worth considering the approach for preference tests, such as done at Harman involve multiple products with trained listeners with rating set parameters and not the ABX switching scenario (which is usually used as mentioned in the other thread to prove if two products sound different debate, most scientific JND studies use other methodologies).
Now coming onto the context of subtle differences being picked up on cognitive behaviour but not consistently or with high confidence with quick switching.
What we are talking about here are very subtle differences that may fail ABX type comparison for reasons I mentioned in the ABX Obsolete thread (bearing in mind I still stress this is unknown although as I pointed out other just noticable difference tests are used in science studies) relating to confusion due to multiple reference points/anchoring/etc, along with bias that is applicable to all such perception tests including ABX.
Along with the above, we also have the challenge of training one to actively hear and be able to identify the parameter; example such as distortion/artifacts/etc.
So it is quite plausible for a factor not to be identified but also still affect us, and this is then compounded by tolerance and threshold with duration.
A very simple example is to consider listener fatigue where the person may not be able to identify the factor involved and indeed the product is rated good, but longer listening sessions then cause the listenes tolerance to reach its threshold and trigger a form of cognitive dissonance.
Another example would be % distortion, where the listener really needs to be trained to identify the % value, but it is fair to say even untrained listeners can and will have long term tolerance problems at certain thresholds.
This is part of the reason (or one anyway) why some audiophiles change parts of their system frequently, especially when they have the itch-dissatisfaction-feel it can be better.
Now I appreciate you may not have that itch-dissonance, which comes back to the original context with regards to the conclusion for you that different DACs make no difference in your system.
The suggestion here is that if two DACs still seem inaudibly different and yet are using two very different filter-noise shaping then it may make sense to investigate why.
Of course the DACs you have tested all may be using the internal DAC chip and also the integrated filter-noise shaping function, so they are all still near enough the same.
But if they are different filters and the DACs seem audibly identical, a longer term listening session may be required if they are using different filters to be aware of cognitive behaviour or possibly becoming aware of certain factors one were not initially (such as maybe one having higher uncorrelated jitter subjectively affecting clarity).
As we are going way off topic, might be worth focusing discussion then on the DACs tried and their filters, if you feel it is worth carrying on this side of the discussion.
But in summary it is unfortunate that really all we have relating to long term preference-cognitive dissonance in this example-JND audible differences and limitations-etc are all either speculation or anecdotal experiences when taken in context of audio perception and preference.
Hope this helps to clarify.
Cheers
Orb
