I'm probably going to get blasted for this, but I don't believe in instant switching ABX testing. I think that the brain fills in acoustics that it expects to hear and I prefer to have a few seconds delay wash out period. This hypothesis is loosely supported by a recent meta-analysis of high definition audio discrimination studies in the AES journal, and is something that I think needs to be addressed in future ABX methodology tests. Instant switching increases the chance of false negatives.
The paper can be checked out here: http://www.aes.org/tmpFiles/elib/20160806/18296.pdf
I've put some clippings below:
<snip/>
What the forest plot shows is that studies with training appear to show a greater ability to discern high definition content, and overall there appears to be a a small ability to discern high definition content. The paper is interesting and not without limitations, so take with a grain of salt and read for yourself. Interesting, among the studies in the training subgroup the longer stimuli and longer intervals between stimuli appeared to correlate with greater ability to discern content.
The results should be taken in a somewhat guarded manner, as the subgroups have small sample sizes. The study also assesses most studies as having high risk of bias. It just isn't the case that audio studies involving people are generally well conducted. There is also lots of heterogeneity in the study methods of studies included in the meta-analysis, but I think the preparation work to make code the studies in a binary fashion, discern or not discern, was done well.
Anyway, go check out the study. It is well worth a read.
You won't get blasted by me!
The box I'm piddling around with is just a tool intended to take another person out of the loop and make sure that "random" is actually random ... I know that personally I find some effects only apparent after protracted listening (e.g. they cause fatigue, sooner or later than some other "thing"), and nothing specifically prevents the box being used in that manner.
And instant vs. delayed switching, as well as short-loops vs. entire albums is something it does already, so the choice is there for the user (assuming there's ever anyone using it but me ...).
Interesting material.