Dear friends, I think I need a sanity check! I have just updated the RAM from two 4 GB cards to two 16 GB cards in my NUC and I am now listening to B.B. King live at Madison Square Gardens and I “feel” that I have greater clarity, increased soundstage and instrument separation! The rational portion of my brain has obviously popped out to get a coffee, because although I know that this is probably nothing but confirmation bias, it sure does sound good.
HifiMan HE500’s, Chord Hugo, RPi 4b, Ethernet via an unmanaged switch, Intel Nuc in a Arkasa case.
Sensible persons needed to guide me back to the path please!
Of course it does sound better, by my quick calculations you should see an improvement of 437.2% Good job! Must be sunny at your location and the coffee is strong……….enjoy it
Contrary to the others replying here, it is very well possible that your music souns better without taking bias confirmation into account.
I have experienced a couple of times changing a component in one of the devices improved sound, whilst the reason of change had nothing to do with audio.
Changing my network cabling from cat5 to cat6 made an impact. Not because I bought ‘golden cables’ but (I believe) my network was on the edge dealing with internet traffic and audio traffic.
Biggest difference with Audirvana, running on Win10 laptop, loading music from my NAS, sending it to HQplayer running on my Audiolinux server, upsampling to DSD512 and passing it on to HQplayer NAA on Holo Red. It did work om my cat5 network but it sounds better with cat6.
Bias, I don’t think so, as I only became ‘conscious’ of SQ change after a couple of days. I knew something had changed but first did not make the link.
Same experience wit CPU power. My previous streamer (pinkfaun 2.16x) was equipped with an AMD Ryzen processor (don’t recall exact number). I could HQplayer do upsample 44,1kHz/24bit to DSD512 , but not 48kHz with the same filters,
I.e the processor was at the edge of its power.
Bringing back upsampling to DSD256 improved SQ also for 44,1kHz sources.
(With my new Intel I9 13900 processor DSD512 sounds a tad better than DSD256).
All I want to say is that if you are running your equipment on the edge, it is possible that your SQ suffers.
I do believe that withe the right equipment this probably would be measurable ( think of error correction and such).
Most or all of these examples have a rational explanation and none of them have anything to do with increasing RAM in an OS that doesn’t swap and either has enough RAM or crashes when it doesn’t.
Thank you Dirk, I’m fairly certain that I “made” myself enjoy the changes, because I wanted them to be there; but I am grateful for your comments and I don’t disagree that even small changes can increase enjoyment.
No, but you’re allways welcome
I am not saying it is always better, but in the case with Audirvana that I described, it is absolutely noticeable and repeatable.
P.S. i really don’t know one would do a ‘blind test’ changing netwok cables. Person can of course leave the room, but the time in between would too far extended.
For the same reason, your claim of “noticeable and repeatable” differences is dubious, since it takes even longer for you to change things around and compare.
One could rig up two complete setups and have someone switch one cable to change between them. Sure, it’s a hassle, and I’m repeatedly on record saying that for practical reasons blind testing in hobbyist audio is by far not the panacea people make it out to be. (But this fact is at the same time not a valid reason for elevating subjective non-blind comparisons to being objective)
