sometimes, i get distorted audio locally on the repeat and also on simple announcements like the time announcement. If and when I get that distorted audio it kind of sounds as if the audio stream is sampled i 20ms increments and every other piece is left out. This happens only every now and then, while most of the time the audio is absolutely perfect.
As hardware is a nondescript 2GHz dual core, (all-star uses about 15-25% CPU), so it should not be a speed problem. Two URI’s are connected to this PC, one for the public all-star link, the other to a private GMRS node,
Tried two different radios, so it should not be the radio either, unless two unrelated GM300s have exactly the same problem (For the time being I’ll go under the assumption its not the radio) I don’t know yet if that ever happens on the TCP end, but I doubt it.
What is the TX voice setting on y our node? If its any where near
900 then it may be trying to drive your TX but is running out of
room.
Do you have TX boost turned on in USBradio.conf?
hdwtype=0
rxboost=0
txboost=1
If not turn it on then retune the TX voice. If you can TX voice near
500 (Mid scale) then you have a better chance of not having the
audio distort on some peaks.
The same would be true if your TX voice setting is near 100 to 200
and you are over driving the TX a little. In this case then turn off
TX boost in USBradio and retune.
Give this a try ans see if this helps.
Jon VA3RQ
···
On 9/10/2013 7:38 PM, Andreas
Pleschutznig wrote:
Good
sometimes, i get distorted audio locally on the repeat and
also on simple announcements like the time announcement. If and
when I get that distorted audio it kind of sounds as if the
audio stream is sampled i 20ms increments and every other piece
is left out. This happens only every now and then, while most of
the time the audio is absolutely perfect.
As hardware is a nondescript 2GHz dual core, (all-star uses
about 15-25% CPU), so it should not be a speed problem. Two
URI’s are connected to this PC, one for the public all-star
link, the other to a private GMRS node,
Tried two different radios, so it should not be the radio
either, unless two unrelated GM300s have exactly the same
problem (For the time being I’ll go under the assumption its not
the radio) I don’t know yet if that ever happens on the TCP end,
but I doubt it.
Does anyone have ANY idea what that could be? –
** Andreas
Pleschutznig, ** Cell:
(832) 633-7817, Sent: MacBook
KA5PLE, Allstar: 29841,
Echolonk: 884823
_______________________________________________
App_rpt-users mailing list
I too had a similar experience with occasional distortion. So I got test gear and checked the audio.
It was right on, but found the receiver was drifting off frequency from temp changes. It’s been pointed
out to me, a very large % of times it’s a hardware failure, as in my case. Test gear makes me happy J
Please; Let us know what you find out.
JK
···
From: Jon Rorke jrorke@cogeco.ca To:app_rpt-users@ohnosec.org Sent: Wednesday, September 11, 2013 7:51 AM Subject: Re: [App_rpt-users] Distorted audio in app_rpt/Allstar
I have seen this before.
What is the TX voice setting on y our node? If its any where near 900 then it may be trying to drive your TX but is running out of room.
Do you have TX boost turned on in USBradio.conf?
hdwtype=0
rxboost=0
txboost=1
If not turn it on then retune the TX voice. If you can TX voice near 500 (Mid scale) then you have a better chance of not having the audio distort on some peaks.
The same would be true if your TX voice setting is near 100 to 200 and you are over driving the TX a little. In this case then turn off TX boost in USBradio and retune.
Give this a try ans see if this helps.
Jon VA3RQ
On 9/10/2013 7:38 PM, Andreas Pleschutznig wrote:
Good
sometimes, i get distorted audio locally on the repeat and also on simple announcements like the time announcement. If and when I get that distorted audio it kind of sounds as if the audio stream is sampled i 20ms increments and every other piece is left out. This happens only every now and then, while most of the time the audio is absolutely perfect.
As hardware is a nondescript 2GHz dual core, (all-star uses about 15-25% CPU), so it should not be a speed problem. Two URI’s are connected to this PC, one for the public all-star link, the other to a private GMRS node,
Tried two different radios, so it should not be the radio either, unless two unrelated GM300s have exactly the same problem (For the time being I’ll go under the assumption its not the radio) I don’t know yet if that ever happens on the TCP end, but I doubt it.
I do not believe this is distorted audio because it is too low or too high. The audio has been measured in with the help of a service monitor and I just checked it. Both channels are set to values around 100. Also distorted because it is set too high or too low sounds different than what I am seeing here.
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
···
On 9/10/2013 7:38 PM, Andreas
Pleschutznig wrote:
Good
sometimes, i get distorted audio locally on the repeat and
also on simple announcements like the time announcement. If and
when I get that distorted audio it kind of sounds as if the
audio stream is sampled i 20ms increments and every other piece
is left out. This happens only every now and then, while most of
the time the audio is absolutely perfect.
As hardware is a nondescript 2GHz dual core, (all-star uses
about 15-25% CPU), so it should not be a speed problem. Two
URI’s are connected to this PC, one for the public all-star
link, the other to a private GMRS node,
Tried two different radios, so it should not be the radio
either, unless two unrelated GM300s have exactly the same
problem (For the time being I’ll go under the assumption its not
the radio) I don’t know yet if that ever happens on the TCP end,
but I doubt it.
Does anyone have ANY idea what that could be? –
** Andreas
Pleschutznig, ** Cell:
(832) 633-7817, Sent: MacBook
KA5PLE, Allstar: 29841,
Echolonk: 884823
_______________________________________________
App_rpt-users mailing list
I have tried that one already. I have exactly the same effect on two different radios, in two different bands. As much as I have to replace one of the radios because it is on its way out (slightly of center) in this case it does not seem to be a specific radio’s problem.
I too had a similar experience with occasional distortion. So I got test gear and checked the audio.
It was right on, but found the receiver was drifting off frequency from temp changes. It’s been pointed
out to me, a very large % of times it’s a hardware failure, as in my case. Test gear makes me happy J
Please; Let us know what you find out.
JK
From: Jon Rorke jrorke@cogeco.ca To:app_rpt-users@ohnosec.org Sent: Wednesday, September 11, 2013 7:51 AM Subject: Re: [App_rpt-users] Distorted audio in app_rpt/Allstar
I have seen this before.
What is the TX voice setting on y our node? If its any where near 900 then it may be trying to drive your TX but is running out of room.
Do you have TX boost turned on in USBradio.conf?
hdwtype=0
rxboost=0
txboost=1
If not turn it on then retune the TX voice. If you can TX voice near 500 (Mid scale) then you have a better chance of not having the audio distort on some peaks.
The same would be true if your TX voice setting is near 100 to 200 and you are over driving the TX a little. In this case then turn off TX boost in USBradio and retune.
Give this a try ans see if this helps.
Jon VA3RQ
On 9/10/2013 7:38 PM, Andreas Pleschutznig wrote:
Good
sometimes, i get distorted audio locally on the repeat and also on simple announcements like the time announcement. If and when I get that distorted audio it kind of sounds as if the audio stream is sampled i 20ms increments and every other piece is left out. This happens only every now and then, while most of the time the audio is absolutely perfect.
As hardware is a nondescript 2GHz dual core, (all-star uses about 15-25% CPU), so it should not be a speed problem. Two URI’s are connected to this PC, one for the public all-star link, the other to a private GMRS node,
Tried two different radios, so it should not be the radio either, unless two unrelated GM300s have exactly the same problem (For the time being I’ll go under the assumption its not the radio) I don’t know yet if that ever happens on the TCP end, but I doubt it.
Sounds to me like a computer with too little RAM or too little CPU
horsepower.
When you run the top command, is it using swap memory? If so,
that’s bad…and you need more RAM.
Kevin
···
On 9/11/2013 12:20 PM, Andreas
Pleschutznig wrote:
Best I can describe the effect is that the audio is time
sliced in 20ms packets and every other packet is missing, left
out. and it does not happen when audio is especially loud or
quiet. Sometimes it happens on the time announcements, which are
absolutely perfect 99% of the time.
I have that exact problem if I use DSP, moved to COS and PTT on DB25 and it stops. Did this in every case I tried, though the only platform I used was a D945 or D201 Intel board.
Subject: Re: [App_rpt-users] Distorted audio in app_rpt/Allstar
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
Sounds to me like a computer with too little RAM or too little CPU horsepower.
When you run the top command, is it using swap memory? If so, that’s bad…and you need more RAM.
DSP takes a significant amount of horsepower. Ya don’t get something for nothing; reading a hardware status bit or two is a lot less intensive at the expense of additional wiring and configuration statements in the config files than decoding the stream, deriving the signals, etc. any overhead in doing so is bound to have an effect on other streams going on at that time.
and we humans don’t process breaks in audio as we do with visual perceptions.
···
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
4GB should be enough me more and 2x2GHz CPUs should be more than enough for asterisk. AT least it was in the past.
···
On 9/11/2013 12:20 PM, Andreas
Pleschutznig wrote:
Best I can describe the effect is that the audio is time
sliced in 20ms packets and every other packet is missing, left
out. and it does not happen when audio is especially loud or
quiet. Sometimes it happens on the time announcements, which are
absolutely perfect 99% of the time.
the problem I have is not limited to a repeat. it does happen when on simple time announcements. At the time when it happens the system is at less than 35% CPU, so I kind of doubt this is a problem in the CPU being too slow.
···
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
when it does that the system is at less than 35% CPU and still has about 3GB of RAM open (this is on a 4GB system)
···
On 9/11/2013 12:20 PM, Andreas
Pleschutznig wrote:
Best I can describe the effect is that the audio is time
sliced in 20ms packets and every other packet is missing, left
out. and it does not happen when audio is especially loud or
quiet. Sometimes it happens on the time announcements, which are
absolutely perfect 99% of the time.
There is GREAT truth to this statement.
Motorola, back before the era the MICOR radio set was being
developed, invested great time and money in this very issue.
It is my understanding, from talking to the inventor and several
engineers that worked at Motorola during the MICOR era that the
circuitry, which ultimately resulted in the famous M6709/M7716
squelch chip, was originally designed for receivers used in NASA
space missions (moon). Obviously, it is crucial the circuit opened
squelch under the slightest amount of quieting. In addition, it
was discovered that the human brain had a difficult time
understanding words that are broken by silence; it is much easier if
the unintelligent space is filled with noise. For this reason, what
became known as the MICOR squelch was a circuit that accomplished
this task, as well as having the ability to close quickly if the
recovered audio was considerably noiseless. Seasoned operators can put together enough recovered audio to make
sense out of a broken transmission, and people didn’t have to listen
to a squelch burst if the recovered transmission was fairly quiet.
This circuitry and theory can be reviewed in detail by going to the
USPTO and do a patent number search;
3628058
Kevin - WJ8G
···
On 9/11/2013 1:56 PM, Bryan D. Boyle
wrote:
and we humans don't process breaks in audio as we do with
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
Sounds to me like a computer with too little RAM or too little CPU horsepower.
When you run the top command, is it using swap memory? If so, that’s bad…and you need more RAM.
Hi Kevin, the internet connection if anything is too fast as well. Besides, even if I had no internet connection that should noyt have any impact on time announcements, right?
···
On 9/11/2013 2:00 PM, Andreas
Pleschutznig wrote:
4GB should be enough me more and 2x2GHz CPUs should be more than
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
Sounds to me like a computer with too little RAM or too little CPU horsepower.
When you run the top command, is it using swap memory? If so, that’s bad…and you need more RAM.
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
Sounds to me like a computer with too little RAM or too little CPU horsepower.
When you run the top command, is it using swap memory? If so, that’s bad…and you need more RAM.
It never fails to amaze me how AMD CPU's get blamed for various
issues--and the actual CPU's are almost never at fault! And, yes, AMD and
Intel have both had an occasional "oops" impacting a specific CPU series.
In my experience, most of these "blame" issues arise from really
poorly designed/buggy motherboards. Unfortunately, in practice, more of
these bad boards seem to be found in low-end AMD boxes, where system
builders are cutting every cost corner.
With a good quality MB, AMD and Intel CPU's both work great and exhibit
similar reliability/stability....And, in case you're wondering, I run a
Co-Lo center and have a few racks full of AMD boards! I measure up-times
in years.
73, David KB4FXC
Huh? Now you see me stumped. I can see some badly designed math library
having a problem with that, but what does that have to do with the
processor. I am not discounting the statement, just trying to gather
information, and the statement does not just come through the filter I
have in my brain as to makes sense.
···
On Wed, 11 Sep 2013, Andreas Pleschutznig wrote:
As far as the CPU goes, yes I do have an AMD chip:
[root@s02 proc]# cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 20
model : 1
model name : AMD E-350 Processor
stepping : 0
cpu MHz : 1597.148
cache size : 512 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
apicid : 0
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 6
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc nonstop_tsc pni ssse3 cx16 popcnt lahf_lm cmp_legacy svm extapic cr8legacy abm sse4a misalignsse 3dnowprefetch ibs skinit wdt ts ttp tm stc 100mhzsteps hwpstate [8]
bogomips : 3194.29
processor : 1
vendor_id : AuthenticAMD
cpu family : 20
model : 1
model name : AMD E-350 Processor
stepping : 0
cpu MHz : 1597.148
cache size : 512 KB
physical id : 0
siblings : 2
core id : 1
cpu cores : 2
apicid : 1
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 6
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc nonstop_tsc pni ssse3 cx16 popcnt lahf_lm cmp_legacy svm extapic cr8legacy abm sse4a misalignsse 3dnowprefetch ibs skinit wdt ts ttp tm stc 100mhzsteps hwpstate [8]
bogomips : 3194.88
On Sep 11, 2013, at 4:01 PM, Corey Dean <n3fe@repeater.net> wrote:
> Any chance you are using an AMD processor? If you tx tone and your voice are more than 999 combined, that could also be the problem.
>
> Corey N3FE
>
> From: app_rpt-users-bounces@ohnosec.org [mailto:app_rpt-users-bounces@ohnosec.org] On Behalf Of Kevin Custer
> Sent: Wednesday, September 11, 2013 12:45 PM
> To: Andreas Pleschutznig
> Cc: app_rpt-users@ohnosec.org
> Subject: Re: [App_rpt-users] Distorted audio in app_rpt/Allstar
>
> On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
>
>
> Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
>
> Sounds to me like a computer with too little RAM or too little CPU horsepower.
>
> When you run the top command, is it using swap memory? If so, that's bad...and you need more RAM.
>
> Kevin
>
>
> --
> This message was scanned and is believed to be clean.
> Click here to report this message as spam.
I had sound related problems with a AMD processor as well many years ago. Finally gave up and used an Intel based and all the sound problems went away.
I have nothing but problems with an amd processor that I was using as my first node. A lot of it was audio related as well. I can’t remember if it was Steve or Jim that ask me the very same question. I ended up trying an intel CPU on a pretty Low end machine and all my problems went away. I have built many nodes since for both myself and others and have never had that problem show up again.
Corey n3fe
Sent from my iPhone
On Sep 11, 2013, at 6:39 PM, “Andreas Pleschutznig” andy@ple.org wrote:
Huh? Now you see me stumped. I can see some badly designed math library having a problem with that, but what does that have to do with the processor. I am not discounting the statement, just trying to gather information, and the statement does not just come through the filter I have in my brain as to makes sense.
As far as the CPU goes, yes I do have an AMD chip:
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
Sounds to me like a computer with too little RAM or too little CPU horsepower.
When you run the top command, is it using swap memory? If so, that's bad...and you need more RAM.
It never fails to amaze me how AMD CPU’s get blamed for various
issues–and the actual CPU’s are almost never at fault! And, yes, AMD and
Intel have both had an occasional “oops” impacting a specific CPU series.
In my experience, most of these “blame” issues arise from
really
poorly designed/buggy motherboards. Unfortunately, in practice, more of
these bad boards seem to be found in low-end AMD boxes, where system
builders are cutting every cost corner.
With a good quality MB, AMD and Intel CPU’s both work great and exhibit
similar reliability/stability…And, in case you’re wondering, I run a
Co-Lo center and have a few racks full of AMD boards! I measure up-times
in years.
73, David KB4FXC
On Wed, 11 Sep 2013, Andreas Pleschutznig wrote:
Huh? Now you see me stumped. I can see some badly designed math library
having a problem with that, but what does that have to do with the
processor. I am not discounting the statement, just trying to gather
information, and the statement does not just come through the filter I
have in my brain as to makes sense.
As far as the CPU goes, yes I do have an AMD chip:
[root@s02 proc]# cat /proc/cpuinfo
processor : 0
vendor_id : AuthenticAMD
cpu family : 20
model : 1
model name : AMD E-350 Processor
stepping : 0
cpu MHz : 1597.148
cache size : 512 KB
physical id : 0
siblings : 2
core id : 0
cpu cores : 2
apicid : 0
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
On 9/11/2013 12:20 PM, Andreas Pleschutznig wrote:
Best I can describe the effect is that the audio is time sliced in 20ms packets and every other packet is missing, left out. and it does not happen when audio is especially loud or quiet. Sometimes it happens on the time announcements, which are absolutely perfect 99% of the time.
Sounds to me like a computer with too little RAM or too little CPU horsepower.
When you run the top command, is it using swap memory? If so, that’s bad…and you need more RAM.
Kevin
–
This message was scanned and is believed to be clean.
Click here to report this message as spam.