nige2000 wrote:
jkenny
posted about linux capable arm pc boards somewhere, its probably an area thats going to need exploration, along with a mqn for linux
maybe if we had mqn for linux then we can start testing these low power board types
Yes, I made this point privately to Gordon when he told me of his results Vs the NAD stack i.e that without all the PS & other mods on the PC it was probably not a fair comparison. But it does show up the weaknesses of the PC as an audio platform - we really do have to go to a lot of trouble to get a great sound out of it i.e a general purpose PC is probably not a great starting platform for audio?
As it seems that most of the SQ issues can be traced back to noise generation in the system (my current working hypothesis), it makes sense to try to come up with some base system that minimises this noise or is at least amenable to easy modification to minimise this noise. low power, minimum spec systems is one way to go - the systems on a chip or ARM based computers - along with minimum controlling OS.
Another way to go might be the networking approach where the audio rendering device is a minimum spec, low power, device with local memory storage for the audio files & specialised, minimum controlling software. However, there are many routes for noise intrusion into the audio rendering system & each connection point is a potential source i.e via the hardwired connection back to the streaming device (USB, network) or via wireless connection (WiFI, Bluetooth, optical).
One of the obvious ways to approach this is to use a standalone system which reads audio files off an SSD, render & transport the digital audio stream to a local DAC chip with minimum OS overhead & minimum PS noise issues. This might offer the best chance of SQ but storage & navigation of the audio libraries becomes one user issue.
There are lots of angles to consider in this whole can of worms most of which quickly leads to complexity & the need for something approaching an OS. Which kind of brings us back to where we started - if we need an OS then we need an PC (of some sort) which means we have to begin to deal with PS issues, anyway.
However, I do agree with Gordon, Windows is just too complex to get a good handle on the noise generation issues - his MQN software shows what can be done when focussing on a very small, probably crucial step in the audio chain but think about all the other elements that we have no control over in the OS which can give rise to noise generation. One approach to this dilemma is to make the PS so stable & imperturbable that these OS activities introduce no PS noise variation. I'm not sure we have fully achieved this yet & getting further along this route is becoming more & more difficult.
Anyway, Ignore me, I'm just rambling!!