@amir_asr
You are arguing against the very nature of how your perception works. Not only experts in audio will disagree with you, so will those in the medical community who research the same. No way your brain has the capacity to remember every bit of fidelity in music you listen to hours and days later. It is impossible.
I think the subject of audio memory is fascinating. And somewhat misunderstood. We certainly have very high bandwidth short-term memory (the type that works with fast/uninterrupted switching). We also have long term audio memory (lower bandwidth so our brains can store more of it). We have intermediate levels as well as different encoding and storage processes (I'd have to dig up relevant papers) but these two are enough for this discussion.
I had an interesting experience a while back that cased me to re-assess two aspects of catechism: that amps that measure similarly well sound the same when performing within their limits; and that longer term audio memory is uninformative.
I replaced one integrated amp (Krell KAV-300iL) with another (Micromega M150) when the former was damaged beyond repair. Standard suite of measurements for both with similar performance at normal listening levels, etc. Both class AB and similar output. Listening sighted and subjectively the Micromega resented different in stereo imaging (less width/depth) and bass timbre. The vendor recommended the dreaded break-in (I'm ok with that, my ear/brain needs it at least) and I'd get used to it, obviously.
But I didn't. When I played new music it sounded pretty good. But when I played familiar stuff, it sounded 'wrong'. I can't tell you which was more accurate, of course that's not possible. But, based on long-term audio memory, the subjective impressions persisted over a few months of listening and my brain didn't adjust to the different sound. The old amp would work for 10-20 minutes or so before static built up and I could compare every now and then after a fashion.
As the Micromega had XLR-out, I bought a secondhand power amp (Krell KAV-2250) and plugged it in. Stereo image and bass returned. So same source, same DAC/pre (the Micromega) but different power amp stage. I ran some measurements (Fuzzmeasure with mic at listening position) to see if euphonic second-harmonic distortion (or similar) was sweetening the bass. Not visible. Also ran room correction (Sonarworks) for FR and left-right imbalance. Not that either.
That leaves a bunch of more esoteric stuff. The Micromega has a different power supply and likely more negative feedback. Some think the latter affects stereo imaging. On the bass side, Krell tend to go overboard on the power transformer (2KVA in the 2250). I'll speculate that the room (with some lateral/oblique mode nulls around 70-90 Hz) pushes against the speaker, and the Micromega doesn't have the current to push back as effectively. But, speculation is all.
Anyway, my takeaway is that long term audio memory is a more complex story, it certainly has resilience and differentiation in my experience (but the efficacy for a reviewer who listens to many system will be a different story). Bass is pretty straightforward (watts are good, but current is better, if you'll excuse the vernacular). Stereo image is the complex product of many factors, starting with the recording, but I wouldn't rule out the amp-speaker-room system as contributor.
According to ASR lore, this can be explained by sighted bias. I was (weirdly) biased against my new amp (I know, a bit contradictory). What I could hear, consistently over several months, was neurosis. While that logic is effectively hermetic, why not test, controlling for visual bias? Well, logistics (I'd need a comparator box to fast-switch, or a friend to slow-switch, or similar) so while I pondered the possibilities a storm took out the Micromega (lighting blew up the water main and fried everything on the ethernet network). Can't win.
