Take a look at the jitter on the usb input on the topping d90 akm 4499 dac. It really couldn't give two shits what anyone thinks about the quality of the usb source.
I'd suggest anyone who thinks either the smsl or soncoz are affected by jitter on the usb input is deluding themselves. My mbp has an awful usb output, I could hear my mouse and trackpad interaction through it on an mdac and Brooklyn, not any more.
I haven't kept up with this forum for several years, but I was pointed to this thread by a friend and I'm pretty surprised by this statement, the belittlement to anyone who has a different view and the assumption that they're ultimately delusional is almost verging on ignorance. I assume that wasn't the intention as I remember PFM being a great community!
Any digital input, regardless of whether it's USB, SPDIF or even I2s direct, is not solely affected by jitter and this has been known for a long time. Yes you can re-clock it and yes you can put that process behind an isolated layer, however, that doesn't mean all underlying technical challenges of high-speed digital circuits are resolved or we would see it being used in military and medical projects. I think there would be of benefit to invest time reading EE books around digital circuits before you make such a holistic claim or don't make the claim at all. As this is very well understood and could cause other people to make ill-informed investment decisions.
I cannot imagine how poorly your system was implemented prior, where you could hear mouse/trackpad interaction through your hi-fi. On that basis, it's really difficult for me personally to trust your subjective statements and more insight into your chain would be appreciated. I'm quite confident that if you cannot hear a difference between two dacs which have different power supplies, DAC chips, oscillators and output stages, there is an underlying cause. I'd suggest you look into further optimisations of your system as out of the 100+ components within each dac will certainly be leaving a footprint.
In regards to your statements about the Ian Canada Dac, I find those interesting, as I too own it. With some slight differences. I have Ian's un-released prototype boards for the larger ultracaps across all my voltage rails, use accusilicon clocks rather than crystek and Bisesik transformers. For me personally, it has zero digital glare, in comparisons that I have made to dacs at a variety of price points albeit only a few. The only way, I could maximise the Ian DAC was through using a well-designed USB input board that has an isolation barrier, utilised xmos 208 and NDK clocks (NDK's aren't active, use the FIFO board as a master). If you feed it I2s through the GPIO, you are limiting it significantly for several reasons (one of them being jitter). I have tried with a PI and a USBbridge Sig, but both now sit in a cupboard.
I hope to have manufacturers prototype AK4499 DAC in the next 4-6 weeks for testing. It's using a discrete oscillator, op-amp output stage and a xmos implementation and will be interesting to see if my thoughts reflect yours and maybe even a dac-off compared to off-the-shelf chi-fi would be of great benefit.
If you'd like any help in looking to improve your Ian dac, to achieve what it should do or provide insight into why you MBP may be introducing sound through your previous dacs. Please let me know, happy to help