clicks and clock adjustment too large

[Mike56]

I have continued to explore every possibility proposed to me by the vendor, manufacturer and others to improve/reduce the latency.
It varies, but i regularly get 5 minutes or so before latencymon decides the system may not be suitable for real time audio.
Is that as good as it gets and adequate or is there something fundamentally inadequate please?
Its interesting to note (to me anyway!) that it doesn't matter if i have the Flex connected (with or without the miniDSP driver loaded) or use the laptop speakers the results are similar. i explored this to try and identify the likely position of the failing and it seems this is the base audio performance of my new laptop.
I no longer hear clicks or dropouts playing music, after various software and firmware updates, but probably will now i mentioned it!
I would have imagined it would be better but maybe i am chasing rainbows?
I am not actually sure what is "playing" the audio files, other than media player, and when testing whatever REW links in to.
I have done all the simple audio trouble shoot and driver updates but nothing more fundamental in that area.
M

 

[Mike56]

OK, some success, i reset the audio setting from default at 48000Hz down to 44100Hz and got a seemingly satisfactory latency report for the computers internal speakers. i haven't tested when linked to the Flex yet, but this must be a clue at least.

 

[Mike56]

And as one might expect, when connected to the flex, the latency test was as bad as, if not worse than before, lasting seconds only.

 

[phofman]

OK, some success, i reset the audio setting from default at 48000Hz down to 44100Hz and got a seemingly satisfactory latency report for the computers internal speakers.

I do not know your current REW setup. But the config screenshot you posted in https://www.avnirvana.com/threads/clicks-and-clock-adjustment-too-large.14776/post-112950 says:

* Generate and capture by REW at 48kHz.
* Access the playback device via windows mixer at 48kHz. If the windows mixer is configured to run at a different rate, let it resample to the samplerate of the playback device.
* Access the capture device directly via WASAPI exclusive, and make it run at 48kHz.

Maybe in your "glitching" samplerate the windows mixer must resample because REW generates a different samplerate than the windows mixer is configured to run at? Do you need to run the playback side through the mixer, why not wasapi exclusive too?

 

[Mike56]

Many thanks for those thoughts! - and your kind attention to my struggles.
I think (but am never certain) that i have tried the correct settings.
When i use REW i have both the Mic (UMIK1) and the output to the Flex processor in exclusive mode, but that makes no difference, and i tried both.

It now seems the computer can play music through its own speakers satisfactorily at 24bit 44100Hz, but when USB'd to the MiniDsp flex as shown above it fails at that setting or the 32bit 96000 the flex would accept.
Those are the ones i have tried so far, and i can hear drop outs before running latencymon.

So i believe i have tried it in a mode where it would need to resample and one when it wouldn't? - and was proven to work internally.

I have done the tests just playing a music file to eliminate any REW incompatibility, and its the same file that works on the computer's speakers, but i don't believe i am seeing any limitation in REW, it seems to be a setting or capability of the computer. Having achieved satisfactory playback through the computer now, i feel there is a clue in that being achieved by the sample/bit settings, but don't know enough to go on from there!

 

[phofman]

First I would suggest to configure REW to use EXCL in both directions, to avoid any interference from the windows mixer.

 

[sam_adams]

When i use REW i have both the Mic (UMIK1) and the output to the Flex processor in exclusive mode

You are running into the USB two clocks problem.

 

[phofman]

You are running into the USB two clocks problem.

If that "exclusive" mode means wasapi exclusive, IMO there should be no clicks caused by separate clocks for each direction. But the clock difference could still be too large for REW, exceeding some inner hard-coded limit in REW.

 

[Mike56]

Thanks folks!
I have set REW with exclusive mode as per attached, i dont know of anything other than Wasapi exclusive this can be

But at the moment i don't perceive this to be the current or only problem, because music will not play without dropouts and associated noises from the computer to the processor (FLEX), so it is probably not a problem with REW but with computer to Flex compatibility or settings. i solved the internal speakers playback with a reset to sample rate/bit depth, maybe this could be the same for the USB output. if i get music to play successfully then i can try REW again!

 

[phofman]

Then it does look like a problem with your system, unrelated to REW.

 

[Mike56]

So "all" i need to do is find out what it is.
There is a faint interest to know what the problem is and a much stronger interest to resolve it and get on with things.
My biggest clue so far is that with the computers speakers only resetting the sample rate/bit depth transformed the latency from seconds to 10 minutes before it declared unacceptable and each reading much less. it would be foolish to assume that is the route to follow, but this evening i will try all the options available to set bit depth/sample rate and see where that leads me. i suspect two areas inside the computer not working seamlessly due to how they build it/set it.

 

[Mike56]

You are running into the USB two clocks problem.

Interesting thought, i don't know what to do about it, but how come others don't report this problem? - does everyone just ignore it, can i set up audio playback differently?

 

[phofman]

Please what does the USB two clock problem actually mean?

 

[reddress]

It sounds like a sample rate or driver issue on the new laptop. Try setting the UMIK-1 and Windows audio to the same sample rate (e.g., 48kHz) in sound settings. Also, check if exclusive mode is enabled in Windows sound properties. If the issue persists, reinstall the MiniDSP and REW drivers. Hope this helps!

 

[Mike56]

Morning Redress and thanks.
I think i have tried all that and lots more driver and system updates and checks, a clean reboot, factory reset BIOS update all drivers check etc etc.
It seems to me from latencymon reports that the NT Kernel activity causes interruptions.
I wonder if i can separate the paths/activity - NB, i don't know what i am doing, but attach the reports that gives this clue. if i were to set the affinity for all audio matters to a CPU area unused by the high activity?
Here attached todays latency reports with PC plugged in and WIFI off, which gives best results so far, and all settings optimised.
M