RX overrun error
Jan Karel
What RX overrun error 0 means? I am getting this error almost immediately when I hit D to proceed SSB. I tried different configuration, but I get it even with all default parameters. I also have one more error - RX A/D speed error. Both errors appear regardless sampling speed. I am using Lenovo W500 notebook, with Windows Vista, Core 2 Duo T9400 @2.53 GHz, 4 GB RAM, internal sound cards Conexant High Definition SmartAudio 221.
I had Windows XP on the notebook, but it behaves the very same? I also tried use external FireWire sound card Presonus FireBox and Virtual Audio Cable sound devices, same results.
When I tried these sound devices with WSJT, sampling correction factors were very close to 1 (1.0001, 1.0002).
Any recommendation/advise welcome.
vy 73 Jan
Leif Asbrink
> What RX overrun error 0 means? I am getting this error almost immediately
> when I hit D to proceed SSB. I tried different configuration, but I get it
> even with all default parameters.
Overrun errors is when Linrad does not fetch data from the soundcard in
time. Some data is then lost. There is a counter associated with overrun
errors and if you do not see the number increase it happened just once and
that would not affect performance although it is a problem that should be
reported on this list. Thank you:-)
> I also have one more error - RX A/D speed error.
That is most probably a serious error.
> Both errors appear regardless sampling speed. I am using Lenovo W500
> notebook, with Windows Vista, Core 2 Duo T9400 @2.53 GHz, 4 GB RAM, internal
> sound cards Conexant High Definition SmartAudio 221.
There are many soundcards - and many of them do not have correct drivers
for Vista. I have recently assembled a computer with two quad core Xeon
processors and I find that nearly nothing works properly under Vista,
and Windows 7. Not even the 64 bit XP is ok. The problems may be related
to the PCI bus and the chip-set. I have not investigated it in depth yet,
but it seems that the same soundcard and drive routine works ok on a
single-core machine.
> I had Windows XP on the notebook, but it behaves the very same? I also tried
> use external FireWire sound card Presonus FireBox and Virtual Audio Cable
> sound devices, same results.
RX A/D speed errors are reported when the input speed is wrong. Typically
far too low. In case you do not see overrun errors with an increasing number
the problem is that the soundcard is not serviced because the CPU(s) are busy.
> When I tried these sound devices with WSJT, sampling correction factors were
> very close to 1 (1.0001, 1.0002).
>
> Any recommendation/advise welcome.
Try Linrad in newcomer mode. That is without second fft and generally less
demanding for the computer. You might try higher priority for Linrad
and a lower value for "Max DMA rate"
Linrad-03.06 and earlier are very sensitive to actions causing other programs
to signal redraw events to Linrad. It is normal to see overrun errors as
well as speed errors if one moves around the windows of other programs over
the screen. Linrad-03.07 will ignore such redraw events and just update
the screen a couple of times per second when a burst of redraw events is
received (they may arrive at a rate of 1000 per second.)
Maybe someonme else can give you ideas about what is going wrong. Maybe
visual effects are enabled and implemented in software. If they use
hardware accelerators in the videocard they might be harmless, but
if they switch off interrupts for too long, the soundcard would loose
data. Linrad-03.07 might be less sensitive to visual effects. I will
upload it in the near future.
73
Leif / SM5BSZ
Jan Karel
Thanks. I continue to play with different settings for sound card and fount out some dependencies. When I set DMA rate bellow 93, I get sometimes only one RX overrun_error and RX A/D speed error with speed bellow 48000 - from about 40000 to 42500. When I set DMA rate 93 and above then there in no RX overrun_error and the RX A/D speed above 48000 - about 49000.
Only Linrad and anti-virus program is running, Linrad reports load bellow 2%, overall CPU load is bellow 3 %. I will continue with investigation and report any improvements.
Do you think that step down to Windows XP could help? I am decided to have separate PC for Linrad. I understood from your comment that single core could be better. Can anyone recommend some PC configuration that works well?
vy 73 Jan
Leif Asbrink
> Thanks. I continue to play with different settings for sound card and fount
> out some dependencies. When I set DMA rate bellow 93, I get sometimes only
> one RX overrun_error and RX A/D speed error with speed bellow 48000 - from
> about 40000 to 42500.
Presumably you would see an ugly spectrum if you would send a good
sinewave into the computer under these conditions. I would guess
that the soundcard runs at 48kHz but that sampling is not
continous. Each time a discontinuity occurs there would be
a wideband click that would easily be seen on the spectrum
and on the waterfall.
> When I set DMA rate 93 and above then there in no RX
> overrun_error and the RX A/D speed above 48000 - about 49000.
There should be nothing wrong in not limiting the DMA rate.
The bandwidth you set will determine the buffer size for the
soundcard. A narrow bandwidth for fft1 or fft2 may cause the
same problem with the soundcard as a DMA rate limit.
The reported speed of 49 kHz seems odd. What happens when you
leave Linrad running for 5 minutes with 'T' pressed to display
the sampling rate? For this test, do not click any signal, just
monitor the input rate.
> Only Linrad and anti-virus program is running, Linrad reports load bellow
> 2%, overall CPU load is bellow 3 %. I will continue with investigation and
> report any improvements.
The problem can not be CPU overload. Something is wrong because the
operating system should give an error message if the device driver
can not use the larger buffers associated with lower DMA rates.
> Do you think that step down to Windows XP could help?
I have no idea. I have just recently become aware of this class
of problems because I have them on my new hardware. I have a feeling
that the problem is not the soundcard itself, but some other drivers
that are needed for USB, PCI or whatever.
> I am decided to have
> separate PC for Linrad. I understood from your comment that single core
> could be better. Can anyone recommend some PC configuration that works well?
I have never seen any problem of this kind in the Pentium II to Pentium IV
computers. They are old now, but for running at 48 kHz anything you can
find in the junk-yard should be fine:-)
I have not much experience with modern computers with modern operating
systems, but I have recently started to investigate what works (not much)
and what does not work (nearly everything) on my D5400XS Intel motherboard
with two quad core Xeon processors. Intel specifies that it should
work with both 32 and 64 bit versions of XP and Vista, but nothing else.
That is not quite correct, it works perfectly well with reasonably modern
Linux and also with Windows 7 but I can not find correctly working
drive routines for most of my (old) soundcards. It would have been
interesting to see how Windows 2000 would behave, but I have not been
able to install it on the D5400XS.
The problems I see with Delta 44 and Soundbalster Audigy are present
also when I run Winrad so the problems are not related to Linrad.
The Vista drivers seem to work with Vista on older computers having
a single CPU. I do not know whether the problems are because of the
PCI bus or whether it is due to multiple processors (or something
else.) Linrad works perfectly well under all operating systems that
I have been able to install when input is a file on the hard disk.
It is interesting to note that the CPU load for individual threads
differs significantly. The reason probably being poor usage of cache
in older operating systems.
Right now I am in the process of documenting the timing tests when
running Linrad-03.07 with very high CPU load from a 2 MHz recording.
With parameters that make the fft1 thread run at about 90% CPU usage
on most of the operating systems, the oldest Linux, Ubuntu 6.10
needs a little more than 100% so it has short gaps in the output
(overrun errors) while 32 bit XP uses 98% with no errors.
The 64 bit version of XP uses 92% and so does both versions of Vista
while the 64 bit version of Windows 7 uses 75%. The most recent
version of Linux uses 88% (64 bit Fedora 11).
I do not know whether the CPU load per thread that is reported to
Linrad is accurate under all operating systems. I will increase
the CPU load until processing fails to check. Under Ubuntu 6.10
it seems OK, but under Debian Lenny the timings for individual
threads seems to be incorrect.
Investigating (and documenting) the problems with the soundcards
will be the next project once 03-07 and associated timing tests are
finalized.
73
Leif / SM5BSZ
Jan Karel
Thanks for suggestions. I was not patient enough. Now, with some settings of DMA rate I am now able to run up to twice stereo 96 kHz channels from my external FireWire sound card. Values stabilize after about 20 seconds and are almost stable.
I have another problem. It is probably related to driver for LAN card. When I use N and T command to send out data to MAP65, sampling rate goes down and MAP65 cannot receive data. MAP65 is installed on an another PC. Checking network throughput; I assume that slightly more than 6 Mbps is requited for MAP65 (4x 96 kHz x 16 bits); I can see sometimes these 6 Mbps, but quite often less, even down to about 3 Mbps.
One more issue with MAP65. Just running MAP65 without any incoming data, it seems to work well. When I am sending data via LAN, any try for decode ends with crash of MAP65. It is the same when 52nd seconds start automatic decode. Program freeze and the only way is to close it from program manager. Sometimes it closes itself.
Changing program priority of both the Linrad and MAP65 has no effect. Overall CPU load is about 10 % for Linrad (Vista, dual 2 core @ 2.53 GHz, 4 GB) and about 50 % on the second PC (XP, P IV @ 3 GHz, 2 GB) where MAP65 is running.
One would expect that newer HW with high CPU power will work fine, it is not true at all. I tried to play with LAN card setting, buffer size etc. I am giving it one more week and then probably try to find some older PC for these application.
73 Jan
Joe Taylor
I know of no reason that Linrad and MAP65 should not be able
to run on the computers you are using -- or even both
programs on one PC. MAP65 runs fine on multi-core or
multi-CPU machines. It was developed on a dual-Xeon machine
and in my shack it always runs on a "Core 2 Duo" machine.
Many people are running both Linrad and MAP65 on one PC
(usually configured with two monitors). This works fine, also.
> I have another problem. It is probably related to driver for LAN card. When
> I use N and T command to send out data to MAP65, sampling rate goes down and
> MAP65 cannot receive data. MAP65 is installed on an another PC. Checking
> network throughput; I assume that slightly more than 6 Mbps is requited for
> MAP65 (4x 96 kHz x 16 bits); I can see sometimes these 6 Mbps, but quite
> often less, even down to about 3 Mbps.
The crucial point is not what Windows Task Manager or some
other utility says about your network throughput. It is
what MAP65 says. The status-bar indicator on MAP65 should
never say "No Rx data" (with red background) during an Rx
minute, and the displayed counter of dropped packets should
read 0% or very close to 0%. If these things are not true,
you have a problem with getting data into MAP65.
> One more issue with MAP65. Just running MAP65 without any incoming data, it
> seems to work well. When I am sending data via LAN, any try for decode ends
> with crash of MAP65. It is the same when 52nd seconds start automatic
> decode. Program freeze and the only way is to close it from program manager.
> Sometimes it closes itself.
Very clearly you have a problem. None of these things
happen in a normally functioning Linrad-MAP65 system.
When you are receiving data from Linrad, is the "Rx noise"
reading somewhere close to 0 dB?
One of the reasons I now advise people to try Linrad and
MAP65 on one computer is that it eliminates the network as a
possible source of trouble. Modern machines have plenty of
memory and CPU power for both programs.
> Changing program priority of both the Linrad and MAP65 has no effect.
> Overall CPU load is about 10 % for Linrad (Vista, dual 2 core @ 2.53 GHz, 4
> GB) and about 50 % on the second PC (XP, P IV @ 3 GHz, 2 GB) where MAP65 is
> running.
I very much doubt that program priority is your problem. I
run everything at "normal" priority. If your system runs
many other programs (most of them, probably, started
automatically at bootup, without your knowledge), it's hard
to say what problems might arise. For precisely this
reason, I do not allow such things as virus-scanning
programs to run in my machines (unless I explicitly ask them
to run).
> One would expect that newer HW with high CPU power will work fine, it is not
> true at all. I tried to play with LAN card setting, buffer size etc. I am
> giving it one more week and then probably try to find some older PC for
> these application.
I have found no problems with very modern, powerful PCs. If
you have Linrad configured so that it runs properly and
sends data to MAP65, your system should be perfectly able to
do what's wanted.
-- 73, Joe, K1JT
Jan Karel
Thanks for your reply. You ask if RX level that MAP65 shows is close to 0, no it is not. It is very high, about 65 dB.
Is this the problem?
73 Jan
Joe Taylor
> Thanks for your reply. You ask if RX level that MAP65 shows is close to 0,
> no it is not. It is very high, about 65 dB.
>
> Is this the problem?
Well, it is certainly *a* problem. Hard to say, yet,
whether it's *the* problem.
I suspect it means that you have not chosen a fixed-point
FFT for the second FFT. Off the top of my head, I forget
the version numbers that Leif uses here; but whatever
version you are now using -- I think probably a
floating-point one -- change to another.
When I'm home again I will check the FFT versions I'm
actually using.
-- Joe
Jan Karel
I will try to change fft2 to MMX and see if I can get RX levels close to 0.
73 Jan
Jan Karel
I am sorry for bandwidth, but I am starting to be desperate.
I changed the fft to MMX (16 bits) and I can set the RX level from about -14 to +18 dB by volume control of a sound card. Unfortunately the problem persists. Red No Rx data is irregularly flashing in about 1 second interval. Meantime there is no dropped packets. It realy looks like network problem. I also tried single PC configuration - local loop 127.0.0.1, but it looks very similar. Windows reports CPU load bellow 30 % when both Linrad and MAP65 is running on the same configuration.
Is there a big difference in LAN load (I/O requests) between MAP65-IQ and MAP65? I am asking because MAP65-IQ works fine with default par files. Can you send me your par files to get some ideas if I haven't mess anything in Linrad?
Will it help if I send you the file and messages that windows show when program crashed? It reports APPCRASH event in the module Audio.pyd
73 Jan
Jan Karel
I have to apology for not fully correct information in my previous e-mail. When I started to try MAP65-IQ I used my old laptop, AMD Turion, 512 MB RAM and XP. MAP65-IQ works well, no problem. Then I tried MAP65 and I have problem I described earlier. I notice that 512 MB RAM is not sufficient for MAP65 and I tried my new notebook with Intel Dual 2 Core, 4 GB and Vista. I started again with MAP65-IQ and it looks fine so I went to MAP65. I tried today again MAP65-IQ on this notebook and there is problem with MAP65-IQ as well. MAP65-IQ receives data from Linrad with no problem, Drop 0.00%, but the very first decode somehow freeze the main window, the clock stops and no other decode. I can open for example options etc. It only stops clock, Moon and Sun data and no more decode lines in In the configuration window are the following messages:
******************************************************************
MAP65-IQ Version 0.9 r1116, by K1JT
Revision date: 2009-04-10 09:28:35 -0400 (Fri, 10 Apr 2009)
Run date: Thu Aug 27 21:06:14 2009 UTC
Using Linrad for input, PortAudio for output.
Will accept unicast data from Linrad.
Audio Output Device Name
Device Channels
----------------------------------------------------------
4 2 Microsoft Sound Mapper - Output
5 2 Speakers (Conexant High Definit
6 2 SPDIF Interface (Conexant High
7 2 Line 1 (Virtual Audio Cable)
8 2 Line 2 (Virtual Audio Cable)
Default Output: 4
Requested Output: 8
Opening device 8 for output.
Audio output stream running normally.
******************************************************************
Exception in Tkinter callback
Traceback (most recent call last):
File "C:\k1jt\svn\wsjt\map65iq\buildmap65\out1.pyz/Tkinter", line 1345, in __c
all__
File "C:\k1jt\svn\wsjt\map65iq\buildmap65\out1.pyz/Tkinter", line 456, in call
it
File "<string>", line 1178, in update
File "C:\k1jt\svn\wsjt\map65iq\buildmap65\out1.pyz/Tkinter", line 1139, in con
figure
File "C:\k1jt\svn\wsjt\map65iq\buildmap65\out1.pyz/Tkinter", line 1130, in _co
nfigure
TclError: invalid command name ".38032848.38033008.38033168"
Waterfall continues and every minutes show new lines, clock on waterfall windows shows correct time.
I checked it again on old AMD Turion notebook and MAP65-IQ still works well. All description of problems with MAP65 were correct.
vy 73 Jan
Leif Asbrink
> I am sorry for bandwidth, but I am starting to be desperate.
Network problems can occur when something "slow" is connected to
the ethernet port - or any other port of the computer running Linrad.
Try to disconnect everything. Then install a loopback device as
described here:
http://www.sm5bsz.com/linuxdsp/run/loopback.htm
Maybe you will find that Linrad and MAP65 will run fine in the
same computer.
73
Leif / SM5BSZ
Joe Taylor
I am running out of ideas that might help you debug your
system, from a distance. If a setup that works fine on your
old laptop fails on the new one, there is clearly a problem
in the way something is configured on the new one. I have
no experience with Vista, so I can't help there. (I might
add that I've intentionally stayed away from Vista because
I consider it a "bloated" O/S, laden with eye-catching
frills that have nothing to do with efficiency or
performance.)
Be that as it may, I do believe that Vista can be configured
so that it will do what you want. I don't know whether Leif
has tried running Linrad and MAP65 on his new 8-CPU machine,
under Vista? If so, maybe he can tell us about it.
-- 73, Joe, K1JT
Jan Karel
thanks for all advices and recommendations. I dedicate the following weekend for more tests. I found out one more dependency for MAP65-IQ on new PC. All work well after reboot. Once I switch the application off and would like to start it again, it behaves as I described yesterday.
I'll start with Leif's recommendation for MS Loop back interface, I've already installed it. What is some how strange for me is that Task Manger shows this interface as only 10 MB and doesn't show any traffic on it. Also netstat -n -a doesn't show any listening on the 50000 ports. Next I try to run two Linrads on the same PC with loop back connection to find out if the problem is on TX or RX of LAN.
I don't know how many members of the list have already tested Linrad/MAP65 with success. I guess that the main problem could be between keyboard and chair as none reported similar issues.
I understood that the traffic from Linrad to MAP65 is UDP. What is the network load for 2x I/Q @ 96 KHz input? Is there any input buffer at MAP65?
Anyway I am happy that I found the way how to run MAP65-IQ at least.
I have the same feeling about Vista as you Joe. When I purchased my new notebook I decided to give Vista try. Fortunately my license allows me to upgrade from Vista to XP. So my last try would be new installation of XP on the new notebook and give it one more try.
VY 73 Jan