Severly distorted audio

[JJ4SDR]

While I was on 20 meters the other day, having a comfortable rag-chew with an old friend who really liked the quality of my transmission and audio, things went south all of a sudden. I should have kept my cans on so that I would have noticed the severe distortion immediately rather than subjecting my friend to it .

When I donned the cans the audio I was hearing (MON ON) I heard was really bad, as if the buffer sizes (USB/Audio interface and VAC1) had changed in mid-stream, while I was talking. I could still understand what I was saying but it sounded like an ugly "buzzing" sound had got mixed into the audio with my voice sitting in the background. Perhaps that was RF bleeding into the audio although I have not had that issue for a very long time but did, in the past (while putting out 600W from another amplifier). Snap-on ferrites fixed those issues. I wish I had made a recording to include here.

I did put out almost 1.5kW out, using my new SS PA. This SS PA I have had on the air for about 3 weeks almost daily now. I switched off the PA and the distortion was still there.

I usually don't believe in coincidences, but before I got on the air that day, it seemed like my PC didn't recognize the external HDMI monitor (never happened before) so the monitor powered off automatically (that's the way I have set it up). Something odd going on. I decided to exit from Thetis, power cycled the radio and the PC and got back on the air again to break in on my friend who was still chewing the rag and he told me I sounded good again.

Any ideas?

Juha

[w-u-2-o]

This is a well known problem with both PowerSDR and Thetis VAC operation.

Periodically, and for reasons unknown, the VAC audio channel will become unstable. This is normally, but not always, indicated by a rapid accumulation of under and/or overflow counts.

This condition exhibits itself as distorted audio, more typically on the transit channel but sometimes on the receive channel. The effect has often been described as sounding like "Mr. Roboto".

Avoidance:

Use more conservative settings at the expense of latency.

Make a test transmission with MON activated to gauge VAC performance before starting operations.

Keep an eye on under/overflow counters; the red/green box-like indicators on the main console UI under the VAC1 and VAC2 buttons can be used for this.

Corrective action:

If this problem occurs dramatic actions such as quitting and restarting the software or rebooting the PC are unnecessary. VAC can be reset by cycling the state of one of the Ringbuffer manual checkboxes (off-on-off or on-off-on, as appropriate). It make take several resets to clear up the problem.

[JJ4SDR]

I appreciate it Scott!

Incidentally, I am wondering why the Buffer size change (on the VAC1 form) doesn't save although I apply the buffer size after changing it to a more conservative value? I also exported the database (I don't want to do many of those) and then imported it but I still see the ole buffer size on the VAC1 form.

Juha

[w-u-2-o]

VAC parameters are saved as part of your transmit profile. Set them, save your TX profile, then you'll be all set.

[JJ4SDR]

Got it! You see, I am making things too complicated again

[JJ4SDR]

Scott,

Audio got distorted again today, about 1 week from its previous occurrence. I had been rag-chewing for quite a while without issues (I didn't really pay attention to the over/underflow indicator). Did change the buffer size to 256 (it was 128 last week) but perhaps that isn't "conservative" enough still. My 1 year old PC has a 10th generation i7 multi-core CPU, 32 GB RAM and NVMe 512 GB SSD so I am not sure why a 256 buffer size may still not be good?

This time the RF in the audio was much less severe than it was last week so I guess the small change helped and I need to tweak it again. I am not quite certain why my PC wouldn't cut the mustard though. Thetis and my DAW/VST Host are the major CPU hogs outside of what Windows 10 uses. I haven't wanted to upgrade to Windows 11 (my PC can run it) as I am not sure that such upgrade would really bring benefits that I (or ANAN) can't live without !

Juha
NI2M

[w-u-2-o]

You are not doing anything wrong, and there is nothing wrong with your PC. The way the VAC audio interface is architected makes it very fragile compared to other examples of the same thing. There is little you can do about that. Adding a DAW into the mix makes things more complex as well.

For instance: I use an expensive ASIO interface that is locked to a GPSDO reference on an i7 PC. I also use a DAW (Reaper). And Voicemeeter as a patch panel. I have to use a buffer size of 512 to make it as reliable as I like. It still goes wonky every once in a while.

Your choices are a) live with an occasional VAC reset being necessary at the current buffer setting or b) increase the buffer setting and see if resets become less of a problem (but they'll probably never go away completely).

[JJ4SDR]

Scott,

I suppose that there are 2 other "choices:

c) not use VAC at all, only audio processing tools inside the ANAN (Thetis)
d) only use external audio processing hardware and tools inside ANAN

Juha

[w-u-2-o]

Scott,

I suppose that there are 2 other "choices:

c) not use VAC at all, only audio processing tools inside the ANAN (Thetis)
d) only use external audio processing hardware and tools inside ANAN

Juha

To put a finer point on it: use the mic input and headphone speaker output built into the ANAN hardware. I don't use the DAW except for routing so I'm not losing any processing by deleting it.

For some, including myself, it's not an option because the ANAN is far away from the actual operating location. There is also added complexity if you want to use a professional microphone.

At any rate it's a trade-off only you can make.