Kernel Optimizations and Rpi 2

[hondagx35]

[Frost_dk]

Ok.. Thanks for your reply. I really appreciate it
Ill guess that it is working now then, but but the changes in sound quality is very subtle.
One last question.. Can you point me to some information about the optimizations itself, if possible? Or is it a bit too "hardcore" for at normal linux user?
Thanks

[hondagx35]

Hi Frost_dk,

the script changes the following:

- MPDs process priority.
This is done with the linux command nice / renice.
Because of your isolated core, this may be not noticeable.

- CIFs process priority.
If you have your music on a SMB/CIFS share this may be noticeable.

- Kernel scheduler latency.
Very complex thing, also the default schedulers have changed with kernels.

- nrpacks
nrpacks which defines how often your USB controller will throw an IRQ/how many frames the USB core will prepare in advance for the USB controller to process on its own. So it is manly for USD DACs.
This parameter has been removed by a patch which reworks the packet size calculation logic on newer kernels.


Frank

[Frost_dk]

[Frost_dk]

[Frost_dk]

Hello Again
I have now been using a Rpi3 (recognized as a Rpi2) for a while, and I think I owe you (Frank) a short update.

I CAN hear a difference when using the different kernel optimizations. It is not a HUGE difference, but clearly noticeable.
The setting OrionV2 is the most "damped" one with great transparency in the midrange. The 2 OrionV3's has more "sparkle", but has not the full transparency as the V2..

If you ever need someone to listen/test some future changes in the optimizations, then I will be more than happy to try it for you!

Kind regards,
Rune Frost