doing up a new machine for the club to replace one of those Fusion machines.
I wrote this up with a audio sample on my blog, if it violates terms or everyone hates going there, I can copy paste it here
this artifact only occurs on receiver audio, not on audio from other sources
OS : Debian GNU/Linux 13 (trixie)
OS Kernel : 6.12.62+rpt-rpi-v8
Asterisk : 22.7.0+asl3-3.7.1-1.deb13
ASL [app_rpt] : 3.7.1
I have been hearing this a LOT lately and have been meaning to get a good audio recording but just haven't gotten around to it. I also haven't had time to dig into which sources it does and does not come from or whether it is on the receive or transmit side. I never used to hear this before ASL3.
None of what I just said is useful information, but I just wanted to agree with you. That sound drives me crazy.
Edit: just wanting to add that my QTH node is 99% just me listening to nodes I'm linked to, and not me talking. So, the raspberries I hear are either in the remotes RX or my local TX. I think there are some specific nodes that I do not hear this when they key up, so I feel like it's on the RX side, not the TX side, like you said.
Possibly the same as / related to:
I also just remembered that I can transmit the entire 15 minutes or so worth of ARN flawlessly, so the raspberries can't be my local TX.
This is interesting. I have never encountered this before on any of my systems. I will say those all exclusively use chan_simpleusb though. Which channel driver are you using?
Putting one thing out there in case this helps. If I look at K9WKJ’s audio closely I can see a few things that remind me of a different problem I’ve seen.
The audio has a 1 kHz tone and also something right around 108Hz (-30dB) which I’m assuming is the PL tone? Anyhow, point A in the diagram is where one of the raspberries starts. If you look at the general shape of the distorted audio you can see 20ms boundaries pretty clearly, which is the audio frame rate of AllStar. The shaded region below is 20ms.
If I look at the spectrum during the distorted periods (all about 300ms in length) I still see the 1 kHz + 108 Hz power, but there is also a phase change and some other power around 500Hz that gets added. Those periods of raspberry remind me a bit of interpolated audio. When you are dealing with pure tones (two of them in your case), the interpolation becomes much more noticeable.
I’m not that familiar with the app_rpt architecture, but does the receiver audio get fed through any kind of PLC function? Normally you think of PLC as part of the network audio path, but is there any chance that the radio audio goes through that stage too?
The one that I listen to is usbradio, but I'm getting several responses back from simpleusb users stating they hear this a lot on their nodes as well.
detailed description is in the blog post
Every time I have heard this, it’s always RX to the node, not TX from it, and I think almost always simpleusb, not usbradio.
Cool, so just based on user reports it definitely sounds like this audio stuttering artifact is only/mostly present in simpleUSBs RX chain.
Which makes sense why I hear it so much. >75% of all non-radioless nodes probably use simpleUSB.
Has anyone tried running two simpleusb nodes connected to a common node to see if the effect is random vs. related to something in the stream?
If both simpleusb nodes show the effect at the same time, that might be a clue.
PS: I'll let you know. 
opposite for me
I have 2 machines running USBradio and they both do it , at a lower volume and less often
this is the first simpleUSB setup to do it
I have another ASL3 running on a Tait TB9100 running a Masters Communications RA-DR1X
OS : Debian GNU/Linux 12 (bookworm)
OS Kernel : 6.6.62+rpt-rpi-v8
Asterisk : 20.10.0+asl3-3.1.0-1.deb12
ASL [app_rpt] : 3.1.0
which runs simpleUSB and it exhbits no artifacts at all
working my way up the chain with audio recordings/spectrograms
will post things when I get there
so far it seems if there is no modulation (no tone,no PL) it doesnt appear to make raspberries
but if PL is there it makes little berries
and 3Khz/1Khz it makes big berries
PL filter bug? I think it's worse the more modulation is present. It seems to follow plosive syllables based on a net I was listening to last night.
Put together some more images and audio of the raspberries.
I pulled this audio from the output of the soundcard and fed into Baudline to do the capture and imagery.
to get the clean audio I turned off the PL encode/decode on the repeater and ran carrier squelch
I can’t see the graph (totally blind op here) but if you didn’t do it during this test, it’d be interesting to do the same test with the PL filter disabled in simpleusb if you are using that, along with CSQ on the repeater. There is no way to completely disable the HPF in usbradio, far as I know.
I have to say I'm not hearing it. I listened on two nodes through about 1000 keyups. Nothing. My production system has been running 22.7.0+asl3-3.7.1-1.deb13 for over a month (since about beta), and I've seen no RX issues there (the RF link uses simpleusb).
I should add that I already know it's not usbradio, since I've been using that exclusively for testing coming app_rpt versions since before 3.7.1.
Forgive me if this is a naive question, but any chance this is just a cellphone beaconing nearby…? Sounds like some I’ve heard before….just sayin’….
Same here. Come to think of it, when I set my phone on a poorly shielded guitar amp...
I used the phone to ID orange drops that were installed backwards once…
