| OvenWerks | Eickmeyer: the liquorix kernel has a group of irqs called PCIe BW notif | 20:02 |
|---|---|---|
| OvenWerks | I don't know what they are, there are 6 of them plus another with s-PCIe which has a priority of 85 | 20:03 |
| OvenWerks | So more than I have cores for (only 4 cores no hyperthreads) | 20:03 |
| Eickmeyer | So, that tells me it's possible the liquorix kernel is patched, not configured differently. I doubt the kernel team will like that. | 20:07 |
| Eickmeyer | Rather, not ONLY configured differently. | 20:07 |
| OvenWerks | Ya we don't want to go there. The difference (at least for me) is minimal anyway | 20:07 |
| Eickmeyer | Yeah. | 20:08 |
| OvenWerks | Though as I said it is also different kernel versions | 20:09 |
| OvenWerks | the stock is 4.* and the liq is 5.* | 20:09 |
| Eickmeyer | That's correct in 18.04, not in 19.04 and 19.10. | 20:09 |
| studiobot | <teward001> thought we were on 5.* in 19.04+ | 20:25 |
| Eickmeyer | @teward001: That's correct, but OvenWerks likes to stick with 18.04. | 20:26 |
| OvenWerks | Interesting... need to be more apples to apples. When I was testing our kernel before (LL) I had FF open with a few YT tabs open (maybe 40 tabs all togther) but when I was looking at liq, FF was not open. | 20:27 |
| OvenWerks | when I run our LL kernel just with Ardour open I had no xruns during play (1 for transport move) but with liq I actually had 5 or 6. This with jack running 32/2 | 20:28 |
| OvenWerks | So it makes a bigger difference whatelse is using CPU than which kernel. | 20:29 |
| OvenWerks | It may also make a bigger difference what HW is in use. Liq may be mo bedda with USB audio devices than with my "obsolete but better than anything new in the price range" PCI audio device. | 20:30 |
| Eickmeyer | Thing is, for someone like me, USB is the only option. | 20:36 |
| OvenWerks | But for me, I don't have a reasonable USB 2.0 Audio device to try. | 20:43 |
| OvenWerks | I am not sure the ones I have are USB 2.0 and not USB 1.1 | 20:43 |
| OvenWerks | After all they are 1/2 i/o and 16bit at 44k1 or 48k which fits in with USB1.1 quite well. | 20:44 |
| wonko | Eickmeyer: Same | 20:56 |
| Eickmeyer | OvenWerks: My Behringer interface goes clear to 192k | 20:58 |
| Eickmeyer | Is it necessary? no, but it can. | 20:58 |
| wonko | Eickmeyer: so does my KA6 | 21:10 |
| wonko | but i'm running it at 48k | 21:10 |
| wonko | because OvenWerks told me to. :-P | 21:10 |
| * OvenWerks pet bats were complaining... | 21:13 | |
| Eickmeyer | Yeah, I run mine at 48k because my HDMI capture card requires 48k. | 21:13 |
| OvenWerks | so much as there is one, 48k is the standard | 21:13 |
| OvenWerks | the 44k1 was always just a hack to make CDs able to hold 1 hours of program | 21:14 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!