[Solved!] M-Track Plus Mk II

Talk about your MIDI interfaces, microphones, keyboards...

Moderators: MattKingUSA, khz

kinleyd
Established Member
Posts: 14
Joined: Sun Apr 26, 2015 10:26 am

Re: M-Track Plus

Post by kinleyd »

Thanks @eric71. I may have tried that already, but will remember to try it again when I get back to jack2.
kinleyd
Established Member
Posts: 14
Joined: Sun Apr 26, 2015 10:26 am

Re: M-Track Plus

Post by kinleyd »

Bummer. KXStudio uses jack2 and cadence, and I'd forgotten that cadence calculates latency differently from qjackctl. For example for this setting:

Sample rate: 96000
Buffer size: 256 samples
Periods/buffer: 4

cadence reports a block latency of 2.7 ms whereas qjackctl (which is what I'd settled on) reports 10.7 ms. So my initial reaction to the lower latency when running kxstudio was misplaced. Also, xruns were still there, with or without Server Synchronous Mode in jack2. Darn!
Last edited by kinleyd on Thu Oct 29, 2015 6:14 pm, edited 1 time in total.
kinleyd
Established Member
Posts: 14
Joined: Sun Apr 26, 2015 10:26 am

Re: M-Track Plus

Post by kinleyd »

falkTX wrote:the cadence reported latency is the *roundtrip* one, qjackctl gives a hw approximate (takes in account number of periods and extra latency).
both values are correct, they display different things.
Thanks for the clarification, falkTX. I did notice that changing the number of periods in cadence didn't have any effect on the latency calculation.

A quick question for you: Is the latency that is commonly talked about, for example the <10 ms latency that is generally recommended, also based on roundtrip calculations alone as calculated in cadence? I'm hoping it is, because then the latency I'm getting isn't too bad, even if I get the odd xrun.
kinleyd
Established Member
Posts: 14
Joined: Sun Apr 26, 2015 10:26 am

Re: M-Track Plus Mk II

Post by kinleyd »

Tried a number of additional things (including upgrading my BIOS, on the off-chance) but without too much headway. So I've had to settle on this:

Sample rate: 96000
Frames/period: 64
Periods/buffer: 16

For this setting qjackctl reports a latency of 10.7 ms, and I experience 1 xrun every 5 minutes on average, which it seems I'll have to settle for. I will now return focus to learning how to play the guitar better! :)

Over and out.
kinleyd
Established Member
Posts: 14
Joined: Sun Apr 26, 2015 10:26 am

[Solved!] Re: M-Track Plus Mk II

Post by kinleyd »

Well, darn. After all that I've finally discovered that the problem was with setting qjackctl to use hw:Plus (the m-track plus) as the capture device, AND hw:PCH (the Intel HDA) as the playback device in Duplex mode:
Capture and playback set to hw:Plus (The M-Track) and hw:PCH (Intel HDA)
Capture and playback set to hw:Plus (The M-Track) and hw:PCH (Intel HDA)
image1.png (56.59 KiB) Viewed 774 times
Including the playback device appears to work, except for the xruns.

Setting it as capture only:
Capture only, set to hw:Plus
Capture only, set to hw:Plus
image2.png (58.88 KiB) Viewed 774 times
and creating an alsa_out device to direct playback to in a script following the load of qjackctl:
script to to create an alsa_out device after qjackctl loads
script to to create an alsa_out device after qjackctl loads
image3.png (46.08 KiB) Viewed 774 times
with some configuration in patchbay to persist the alsa_out device did the job. No more xruns at normal loads even at a latency of 2.67 to 4 ms! I'm not sure why duplex mode doesn't work properly but at least I've banished the xruns. And learnt a lot of stuff in the process. :)
Post Reply