Resonance of case affected audio
20 years ago I had a Theta transport which performed heads and tails above anything I could find in its price class... and then the laser went out and had sadly gone out of production, with lesser replacements on offer. Before the Theta betrayed me, whilst looking inside one day, I found something that caused the music to take on a much greater sense of natural, analog-like ease that I'd not heard out of a digital in the 90's, without giving up any detail or definition. The cause was removing the case top. I had previously applied bitumen damping to the cover, which reduced some ringing and increased the sound stage. But removing the cover altogether was a revelation. Unfortunately at the time I could not think of a good solution to keep dust et al away from the transport drive and electrics. And I was a little concerned about health and safety. So I reluctantly screwed the cover back on.
It could have been a resonant interaction with the mechanics of the transport, and/or it could have been an interaction with some of the components on the PCB.
Reminded of that today, I wonder if my preamp would also respond well to removing or isolating the cover from the rest of the case. So I may try putting a thin layer of rubber between the case top and the main structure, screwing down only lightly. Since the cover wraps around on the sides, that will mean the side screw holes will not likely line up and will need to leave those unfixed.
Bruce
Theories are not so much answers as questions, to be supported or undermined by experience & testing.
Source: Audiolab 6000CDT > Calyx 24/192 DAC
Amplification: Pass-design B1rev2 pre-amplifier > Neurochrome Modulus 686.
Loudspeakers: Proac Response 1SC
Cables/stands: spkr: MIT MH-750 biwire; IC: HT Truthlink; Target stands, sand-filled; Excel Cat6A 23AWG UFTP & 1attack.de Cat.7 SFTP.
Other: Balanced AC transformer to hydra mains distr; Bass traps & Acoustic panels; Isolation: Inner tube & roller bearings; 3xZyxel ES104A switches in series w/Vreg upgrades.