VAC/audio problems on 200D

USB headsets to digital audio workstation software...
G3ZQH
Posts: 111
Joined: Mon Apr 10, 2017 5:21 pm

VAC/audio problems on 200D

Postby G3ZQH » Mon Jan 20, 2020 12:14 pm

200D P2v1.9, Thetis 2.6.9, PC Windows 10, VAC 4.60 on my set-up

I need some help with Thetis stability, particularly when using digi modes (JTDX, WSJT). I have intermittent problems with VAC audio which causes Thetis to crash. This is usually accompanied by under/overflows in VAC and a visual distortion in the waterfall. I have attached a few screenshots of FT8 waterfall to show one normal and two types of 'bad' waterfall. These problems occur at erratic intervals, sometimes several times a day or other times I can leave the rig on for many days and it is stable. I can't find any pattern at all. I'm using VAC by Eugene Muzychenko to generate the virtual cables.

Any suggestions form VAC audio experts much appreciated!

73 Dave G3ZQH
Attachments
JTDX Thetis bad.png
JTDX Thetis bad.png (624.71 KiB) Viewed 10147 times
JTDX Thetis bad1.jpg
JTDX Thetis bad1.jpg (267.31 KiB) Viewed 10147 times
JTDX Thetis good.png
JTDX Thetis good.png (650.75 KiB) Viewed 10147 times
Last edited by G3ZQH on Mon Jan 20, 2020 4:06 pm, edited 1 time in total.
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: VAC/audio problems on 200D

Postby w-u-2-o » Mon Jan 20, 2020 3:40 pm

Dave,

Helpful hint: next time you post attachments, if you post them as JPG, GIF or PNG and click the "place inline" button they will show up right in your post and we don't have to download them and open them in Word or something. The easiest way to do that is to use the "Snipping Tool" app built into Windows.

If you are having runaway under/overflows, that will cause this problem.

Runaway under/overflows are caused by the resampler being unhappy with your chosen VAC1 settings.

You have two choices:

1. Find settings that the resampler likes. This will provide superior audio.
2. Click the "force" checkboxes (and makes sure the rate is set to "1"), this will disable the resampler. This is the easy way and works for most people who don't care about squeezing the last ounce of performance out of the software.

Post a screen shot of your VAC1 settings (in a graphics format, e.g. JPG or PNG) if you like.

73,

Scott
G3ZQH
Posts: 111
Joined: Mon Apr 10, 2017 5:21 pm

Re: VAC/audio problems on 200D

Postby G3ZQH » Mon Jan 20, 2020 4:32 pm

Hi Scott
Thanks, I have made the attachments visible in my post now.

I have evaluated two basic VAC settings over several months- one with the resampler fully on (all boxes unticked) and one with the 'force' boxes ticked. I didn't notice much difference in stability between these two settings over periods of days or weeks. Both were unstable and prone to the same type of VAC problems. I have attached my VAC1 settings. My RX gain setting at -25 dB seems unusual, but it needs this amount of attenuation to stay on scale for WSJT/JTDX.

I have also tried various audio drivers: MME, WDM-KS, Windows direct sound. None of these resolved the problem.

Are there any settings within the VAC control panel (Muzychenko) that need tweaking? I have just used the defaults.

73 Dave G3ZQH

Annotation 2020-01-20 162129.png
Annotation 2020-01-20 162129.png (42.93 KiB) Viewed 10146 times
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: VAC/audio problems on 200D

Postby W1AEX » Mon Jan 20, 2020 8:51 pm

G3ZQH wrote:200D P2v1.9, Thetis 2.6.9, PC Windows 10, VAC 4.60 on my set-up

Hi Dave,

I use VAC and it performs flawlessly for me to transport my TX audio from my UMC202HD USB interface into Thetis, but I am by no means an expert. I don't run digital modes with my ANAN but with my hardware/software installation I know I could not get stable VAC operation with a Buffer Size as small as yours, with or without the resampler. A size of 1024 has been perfectly stable while anything smaller is not. The other thing I wonder about is the Mono/Stereo setting. I have always had Stereo checked, and as far as I know have never run with it unchecked. Scott is definitely the expert with all the VAC stuff so I certainly defer to him on all that.

One other thing. I also have a 200D and am completely unaware of a P2 v1.9 firmware! Where can that be found???

73, Rob W1AEX
"One thing I am certain of is that there is too much certainty in the world."
G3ZQH
Posts: 111
Joined: Mon Apr 10, 2017 5:21 pm

Re: VAC/audio problems on 200D

Postby G3ZQH » Mon Jan 20, 2020 11:13 pm

Hi Rob
That is helpful. I'll try running at 1024 buffer size and see if it improves stability. Separately (not wanting to change two things at the same time) I'll investigate mono/stereo as well. I'll report back on this.

I'm generally very happy with the recent improvements to the Thetis user interface (many thanks to Richie MW0LGE for his phenomenal work) and it is only this small problem with audio stability still bugging me.

Apologies- it should be v 1.8 - fat fingers...

73 Dave G3ZQH
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: VAC/audio problems on 200D

Postby w-u-2-o » Tue Jan 21, 2020 12:03 am

Rob is nailing it. Increase your buffer setting. The additional latency won't bother you for digi op's. Check the force boxes if necessary. And check the stereo box.
User avatar
ramdor
Posts: 1468
Joined: Wed Jul 03, 2019 3:07 pm
Contact:

Re: VAC/audio problems on 200D

Postby ramdor » Tue Jan 21, 2020 1:06 am

Just to muddy the water a tad, even with stable vac settings and increased buffer sizes, I sometimes (perhaps once/twice a week) had issues with Muzychenko's vac and it would give similar results to the top image seen on both the anan+thetis and sunsdr2pro+expertsdr. A quick restart of his audio engine would sort it out. II have not however seen anything like that with his new 4.62 version. The second image looks like re-sampler unable to stabilise.

Richie.
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: VAC/audio problems on 200D

Postby w-u-2-o » Tue Jan 21, 2020 3:13 am

I'll add that I greatly prefer Voicemeeter Potato to Muzy VAC. It's easier to use, and provides better results. The only thing Voicemeeter won't do is transfer IQ data to CW Skimmer.
G3ZQH
Posts: 111
Joined: Mon Apr 10, 2017 5:21 pm

Re: VAC/audio problems on 200D

Postby G3ZQH » Tue Jan 21, 2020 5:10 pm

Hi Scott, Rob, Richie
Great information, thanks to all. I’m using VAC 4.60 and had earlier problems installing 4.62 but will try that again. If that doesn’t work I’ll move to Voicemeeter Potato as Scott suggests (may be better anyway). My 200D was stable today for 8 hours with only one ‘audio’ crash, so that is a big improvement by simply altering the buffer to 1024.

One thing I would like for SSB phone operation is to be able to monitor my own TX audio without the current annoying slight delay that makes it impossible to speak and monitor at the same time. From what I understand, I’ll need to have different settings for this so it can wait until I have RX stable.

I’ll report back on the other VAC changes soon.

73 Dave G3ZQH
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: VAC/audio problems on 200D

Postby w-u-2-o » Tue Jan 21, 2020 5:18 pm

G3ZQH wrote:My 200D was stable today for 8 hours with only one ‘audio’ crash, so that is a big improvement by simply altering the buffer to 1024.
Bump it to 2048 and see if it's better still.

One thing I would like for SSB phone operation is to be able to monitor my own TX audio without the current annoying slight delay that makes it impossible to speak and monitor at the same time. From what I understand, I’ll need to have different settings for this so it can wait until I have RX stable.
You will not be able to achieve this. The lowest mic-to-ant latency that is achievable is on the order of 30 to 40ms with the best possible settings, and that is without any audio processing such as the leveler, EQ, CFC, etc. 30ms is still too long of a delay to be unnoticeable and when you add all the cool audio processing stuff it becomes much longer than that. Also, if you are using PureSignal then the MON output also includes the pre-distortion. In order to hear (delayed) what you actually sound like you have to disable PureSignal.

When it comes to adjusting the TX audio chain to achieve a particular "sound", the best approach to this problem is to create a tape loop sample of your unprocessed voice, feed that in via VAC, set the radio into dummy load at low power with PureSignal off, and then listen to MON and start moving audio controls around. This way you are not listening to yourself, you are listening to a recording, so no annoying delay.
sv1jso
Posts: 36
Joined: Wed May 30, 2018 7:06 pm

Re: VAC/audio problems on 200D

Postby sv1jso » Tue Jan 21, 2020 8:51 pm

w-u-2-o wrote:I'll add that I greatly prefer Voicemeeter Potato to Muzy VAC. It's easier to use, and provides better results. The only thing Voicemeeter won't do is transfer IQ data to CW Skimmer.


If you don't need the extra channel the banana is passing IQ just fine.

73
Mike SV1JSO
User avatar
w-u-2-o
Posts: 5540
Joined: Fri Mar 10, 2017 1:47 pm

Re: VAC/audio problems on 200D

Postby w-u-2-o » Tue Jan 21, 2020 9:58 pm

Good to know, Mike! I don't use Skimmer, so my information was second hand. Maybe you could post your Voicemeeter Skimmer setup in a separate topic for others?
W4WMT
Posts: 325
Joined: Sun Apr 09, 2017 10:12 pm

Re: VAC/audio problems on 200D

Postby W4WMT » Wed Jan 22, 2020 11:32 am

Hi All,

When you run VAC1(VAC2) with the RingBuffer Manual checkboxes unchecked (in the Buffer Latency group), here is what happens: PortAudio goes out to your sound driver and asks it to report the minimum latency it is capable of doing. Thetis gets that info from PortAudio and then tells the ChannelMaster to create its ringbuffers to hold as many samples as it takes to produce that much latency.

On busy computers, where Thetis is competing with other programs for cpu cycles, the Windows thread scheduler may not have enough time to service the relevant ChannelMaster threads quickly enough to accommodate the ringbuffer size described above. When that happens, your audio data is dumped on the floor :o

Often, the solution is to increase the Manual RingBuffer settings enough to give the Windows thread scheduler adequate time to do its job. A good place to start is a latency setting that is six or eight times the latency represented by your VAC1 "Buffer Size" setting. You'll have to do a little arithmetic to figure the multiple. It's best to make these adjustments with the "Force" boxes checked, so that you don't start chasing your own tail when the adaptive variable resampler tries to adapt to your changes.

A well designed sound driver for a high-performance audio API like WDM-KS or ASIO should not need a terribly large "Buffer Size" setting in VAC1. So don't be afraid to try a size like 256 or 128. Whatever the size, the ringbuffer has to be big enough to detain as many "VAC buffers" as it takes to give the busy Windows thread scheduler enough slack to do its job!

73, Bryan W4WMT
G3ZQH
Posts: 111
Joined: Mon Apr 10, 2017 5:21 pm

Re: VAC/audio problems on 200D

Postby G3ZQH » Wed Jan 22, 2020 12:17 pm

Hi All
Another update on my VAC stability after the suggested changes. Change to 1024 buffer size helped but I still had audio dropouts several times in a day. I tried ticking 'stereo' but this made no difference that I could tell. Next I upgraded to VAC 4.62 (thanks Richie for this suggestion) and with 1024 buffer setting this has been very stable for over 24 hours now. I'll leave it at this setting to assess long term stability. I have kept the WDM-KS driver but have not evaluated the other options with the new settings. Voicemeeter Banana is now installed and I'll also look at this later.

Thanks for all your inputs- I believe my problem is now mostly resolved.

Originally (several years ago...) I decided on the audio buffer size setting by monitoring RX for elimination of clicks and pops. I used the lowest setting which gave acceptable audio. Since then I have not changed it, so this re-evaluation was certainly long overdue.

Scott- your observations on TX audio monitoring means I can essentially not be too worried about trying to achieve 'real time' monitoring (old style). I recently changed my headset so had to make some adjustment to my TX audio hence the need to monitor. What you suggest using recording is good- but I managed to optimise settings on air with the help of another knowledgeable ham. so I now have 'DX' and 'ragchew' TX profile settings.

73 Dave G3ZQH
User avatar
W1AEX
Posts: 425
Joined: Sun Apr 09, 2017 6:17 pm
Location: Connecticut, USA
Contact:

Re: VAC/audio problems on 200D

Postby W1AEX » Wed Jan 22, 2020 3:29 pm

Nicely done Dave! Sounds like you are on your way with the digital software a little happier using different settings in the VAC setup. Also, no problem with the reference to the mythical v1.9 f/w for the Orion! I have done the same key mashing now and then with numerical file names! :)

Note to Bryan - Thanks for the detailed explanation of what's going on inside Thetis as it handles our VAC setups. Definitely good to know for anyone looking to tinker with whatever settings they have settled upon!

73, Rob W1AEX
"One thing I am certain of is that there is too much certainty in the world."

Return to “Digital ("Virtual") Audio”