Vario sound is broken #18

kedder · 2020-03-09T18:27:16Z

Vario sound is reported to be broken by several users:

variod sound output broken

The audio was odd. It may have been that things were just incorrectly configured, but my original image, the audio seems very normal compared to a vario. This seemed like it was modulated with a low frequency square wave. I.e. it would just cut in and out constantly.

This gives an idea how it sounds like: https://drive.google.com/file/d/1qQbzZW6G60sZAgtfYW7ZHiwC8yZA5LGY/view

CazYokoyama · 2020-03-09T18:34:08Z

I already reported how to fix, didn't I? -caz

…

On Mon, Mar 9, 2020 at 11:27 AM Andrey Lebedev ***@***.***> wrote: Vario sound is reported to be broken by several users: variod sound output broken The audio was odd. It may have been that things were just incorrectly configured, but my original image, the audio seems very normal compared to a vario. This seemed like it was modulated with a low frequency square wave. I.e. it would just cut in and out constantly. This gives an idea how it sounds like: https://drive.google.com/file/d/1qQbzZW6G60sZAgtfYW7ZHiwC8yZA5LGY/view — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#18?email_source=notifications&email_token=AAWALMS44RX4DSYPWFKQAY3RGUYILA5CNFSM4LEOZPJKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4ITVGIKQ>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAWALMVCMMQQNB5TZXTWJXDRGUYILANCNFSM4LEOZPJA> .

-- -caz, caz at caztech dot com, 503-six one zero - five six nine nine(m)

kedder · 2020-03-09T18:43:07Z

I already reported how to fix, didn't I?

Um, really? Where can I find it?

kedder · 2020-03-09T19:22:41Z

Some hints from @CazYokoyama (taken from #11):

For variod audio and XCSoar airspace warning,

Kill variod

vario data from 4353 instead of 4352, i.e. from sensord.

mihu-ov · 2020-03-09T20:44:46Z

Kill variod

vario data from 4353 instead of 4352, i.e. from sensord.

This is the configuration we had before variod with XCSoar providing audio vario. Back then the update rate was 2Hz which was too slow and really annoying. It is much better now but leaves us without volume control over stick remote / .xci file and without STF (speed to fly) audio.

kedder · 2020-03-11T21:47:27Z

The noise is indeed coming from variod. OTOH xcsoar own audio vario gives a nice clean sound. I'll experiment with @CazYokoyama approach, maybe we can increase the update rate to acceptable level (if that's the only issue with it).

CazYokoyama · 2020-03-11T23:59:57Z

MS5611 -> sensord -> tcp 4353 -> variod +-> tcp 4352 -> xcsoar +
| |
+----------------------------+-> speaker
As mihu said, the path, tcp 4352 -> xcsoar -> speaker is slow. I analyzed sensord, but not variod. To analyze/debug variod, there is no good debug environment. I wish I can run variod on raspbian.

CazYokoyama · 2020-03-12T03:55:49Z

Is anyone working for variod? I built debug environment on raspbian. If no one, I will.

kedder · 2020-03-12T07:00:11Z

@CazYokoyama please do!

BTW, I've found that recent gcc (9.2 in my case) cannot build variod anymore, but it is fixable with this patch:

diff --git a/Makefile b/Makefile
index 4a017e7..591d422 100644
--- a/Makefile
+++ b/Makefile
@@ -22,7 +22,7 @@ doc:
        doxygen
        
 variod: $(OBJ)
-       $(CXX) $(LIBS) -g -o $@ $^
+       $(CXX) -g -o $@ $^ $(LIBS)
        
 install: variod
        install -D variod $(BINDIR)/$(EXECUTABLE)

linuxianer99 · 2020-03-12T07:37:45Z

variod was invented by Andreas Wasserbauer. A friend of Michael Huber.
But he is also very short on time.
@CazYokoyama : If you can debug, feel free
@mihu-ov : Think this is ok, isn't it ?

linuxianer99 · 2020-03-12T07:38:57Z

One other thing:
variod was invented, as XCSoar vario was not very responsive in the past.
Has this changed in the current releases ??

@mihu-ov : If so, we maybe can remove variod as XCSoar vario is good enough ?

mihu-ov · 2020-03-12T10:20:44Z

Andi has no github account yet but answered on the Redmine, I think he is OK with me copying his reply below. My personal preference (and I am no software developer) would be to keep audio vario in variod even if this causes an additional interface to XCSoar because I think

vario quality is the first priority
vario information (sensor data to audio) should be as quick as possible
in flight the audio vario from variod worked really very well (STF not yet)
variod could be useful with LK8000, Tophat or other software as well
further improvements are probably easier implemented and flight tested in a small app like variod than in a package like XCSoar

Hi caz, kedder

I don't have a github account yet, so let me just give you an update concerning variod this way.

First of all: It's of course ok to work on XCSoar's vario output, but it is very limited compared to variod. So it's probably quite some work to overcome it's current limitations.

Concerning the broken variod: I know about the issue and I am working on it by replacing the low-level alsa code with pulseaudio. I think that won't be too much work, and should also allow us to have XCSoar and variod output sound in parallel.
From my perspective this is the way to go.

cheers

Andi

kedder · 2020-03-12T11:07:13Z

That's awesome news, @mihu-ov!

kedder · 2020-03-22T21:11:47Z

Status update for anyone following this: Andreas Wasserbauer started working on porting variod to pulse audio. This will allow us to fix the sound output and also support mixing vario output with XCSoar own events (e.g. airspace warnings).

Andreas doesn't have github account and was posting updates on openvario redmine server (http://www.openvario.org/redmine/boards/3/topics/1773). Unfortunately redmine doesn't work anymore (Internal error) and we've lost the only communication link with him. @linuxianer99, can we get redmine back up at least until we figure this issue out?

mihu-ov · 2020-03-22T21:59:38Z

Andi is the Redmine admin, I have contacted him by mail as it is too late to call him.

kedder · 2020-03-24T21:49:26Z

I've included pulseaudio server into the image in my branch here: https://github.com/kedder/meta-openvario/tree/warrior-pulseaudio

We can use it as soon as variod is ported to pulseaudio.

kedder · 2020-04-01T18:46:23Z

So I was playing with variod, ported to pulsaudio, that Andreas sent me (pushed it to my fork, if anyone is interested).

And it seems like we were barking at the wrong tree this whole time.

Jittering sound is still there.

However, after closing investigation (I've stopped variod service and ran it in foreground with /opt/bin/variod -d -f), it became apparent that the jittering sound comes from this:

!! DEBUG LEVEL 1 !!
!! STAY in foreground !!
Socket created ...
=========================================================================
Runtime Configuration:
----------------------
Vario:
  Deadband Low:			-0.000000
  Deadband High:		0.000000
  Pulse Pause Length:		48000
  Pulse Pause Length Gain:	1.000000
  Base Frequency Positive:	350
  Base Frequency Negative:	350
Speed to fly:
  Deadband Low:			-2.500000
  Deadband High:		2.500000
  Pulse Pause Length:		48000
  Pulse Pause Length Gain:	0.200000
  Base Frequency Positive:	350
  Base Frequency Negative:	350
=========================================================================
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
Waiting for incoming connections...
Connection accepted
Connection dropped
...

My impression is that sensord is connecting and disconnecting from variod socket all the time and variod just making a short "beep" each time it does it.

Does it ring any bells?

kedder · 2020-04-01T18:58:43Z

Here's the output of sensord that is run in foreground:

# /opt/bin/sensord -f -d
!! STAY in g_foreground !!
!! DEBUG LEVEL 1 !!
Opened 24C16 on 0x50 failed !!!
No EEPROM found !!
=========================================================================
Runtime Configuration:
----------------------
Vario:
  Kalman Accel:	0.000000
Sensor TEK:
  Offset: 	0.000000
  Linearity: 	1.000000
Sensor STATIC:
  Offset: 	0.000000
  Linearity: 	1.000000
Sensor TOTAL:
  Offset: 	0.000000
  Linearity: 	1.000000
=========================================================================
Opened MS5611 on 0x76
Error writing to i2c slave (ms5611_reset)
Error writing to i2c slave (write cal reg)
Opened MS5611 on 0x77
Error writing to i2c slave (ms5611_reset)
Error writing to i2c slave (write cal reg)
Opened AMS5915 on 0x28
Opened ADS1110 on 0x48
Error writing to i2c slave (ms5611_start_temp)
Error writing to i2c slave(ms5611_read_temp)
Error writing to i2c slave: start conv: adr 76
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Error writing to i2c slave (ms5611_start_temp)
Error writing to i2c slave(ms5611_read_temp)
Error writing to i2c slave: start conv: adr 76
Error writing to i2c slave: start conv: adr 77
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Unable to read from slave
ams5915_calculate @ 0x28: Pressure: -6.248569 Temp: -50.000000
ads1110_calculate @ 0x48: Voltage: infV
Error writing to i2c slave (ms5611_start_temp)
Error writing to i2c slave (ms5611_start_temp)
Error writing to i2c slave(ms5611_read_temp)
Error writing to i2c slave(ms5611_read_temp)
Error writing to i2c slave: start conv: adr 76
Error writing to i2c slave: start conv: adr 77
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Unable to read from slave
ams5915_calculate @ 0x28: Pressure: -6.248569 Temp: -50.000000
ads1110_calculate @ 0x48: Voltage: infV
Error writing to i2c slave: start conv: adr 76
Error writing to i2c slave: start conv: adr 77
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Error writing to i2c slave: write Read result(ms5611_read_pressure)
Unable to read from slave

These errors are probably from the fact that I don't have any sensors connected.

linuxianer99 · 2020-04-02T07:06:59Z

Hmmm ... This is a good point !
This maybe the reason, not everybody has this problem ...
My openvario variod works fine here on my desk .. but i have sensors connected ..

I will hav a look at sensord tomorrow. Adding some detection at the start or something like this ...

mihu-ov · 2020-04-02T07:32:03Z

Unless for testing using variod without pressure sensors is pretty pointless, isn´t it?
If we have packages for pulseaudio and variod or an image for PQ070 display I can test (with sensorboard) in the evening.

kedder · 2020-04-02T08:18:48Z

I've uploaded packages to my PPA. If you want to try it out, you can do this:

echo src kedder http://openvario.lebedev.lt/opkg/armv7vet2hf-neon/ >> /etc/opkg/customfeeds.conf
opkg update
opkg install pulseaudio-server variod-testing
newgrp - audio

and restart the openvario. Next time it loads, you can run alsamixer, that should show the single volume bar - that's how you make sure pulseaudio is running.

I'll play with it a bit more today evening too.

mihu-ov · 2020-04-02T09:44:26Z

root@openvario-7-PQ070:~# alsamixer
ALSA lib ../../alsa-plugins-1.1.8/pulse/pulse.c:243:(pulse_connect) PulseAudio: Unable to connect: Access denied

cannot open mixer: Connection refused
root@openvario-7-PQ070:~#

kedder · 2020-04-02T09:49:38Z

@mihu-ov that means pulseaudio server is not running

kedder · 2020-04-02T10:20:31Z

@mihu-ov Oh, I remembered now. You have to do this:

newgrp - audio

and restart the device. This will be incorporated into the image with pulseaudio, but if you install server manually, you have to do that.

mihu-ov · 2020-04-02T10:21:21Z

Thank you kedder, variod is beeping on my desk now and alsamixer shows the single volume bar. I got a bit confused as there were actually two beep tones coming from the speaker but I had XCSoar audio vario turned on as well, so variod and XCSoar audio really go together now!
I need to start variod manually but this is something I may have broken on my OV.

Looks good! (And I need to go back to work but this was too exciting ;-)

kedder · 2020-04-02T10:24:03Z

Oh, that's a great news! I'll make a PR this evening.

And I need to go back to work but this was too exciting

Yeah, same here :)

mihu-ov · 2020-04-02T10:25:38Z

OK, after newgrp - audio autostart works again. Great news indeed!

kedder · 2020-04-02T19:40:38Z

Ok, so I learned something about the issue:

If I replace sensord with simple telnet localhost 4353, I get clear tone of about 700Hz out of my speaker. As soon as real sensord starts, sound gets corrupted. "Monitor" view in XCSoar devices screen shows a wall of repeated messages:

$POV,E,+99.00*09
$POV,V,+0.00*2A
$POV,P,+000.00.Q,+0.00*48

Again, I don't have any sensor hardware. I need someone with the hardware to check if it really works. I'll create couple of images for testing on real hardware.

kedder · 2020-04-02T20:23:15Z

@mihu-ov, please test this image with real sensors:

https://www.dropbox.com/s/emw5qi63ve73mfw/OpenVario-linux-openvario-image-testing-glibc-ipk-20093-openvario-7-PQ070.rootfs.pulseaudio.img.gz?dl=0

That's an image for PQ070 display (I'm assuming you've got that one) and pulseaudio-enabled variod.

You should hear the nice vario once xcsoar starts, that should react to pressure changes on the sensors.

mihu-ov · 2020-04-02T21:11:13Z

You should hear the nice vario once xcsoar starts, that should react to pressure changes on the sensors.

Believe it or not, this is exactly what happens right after starting my OV with the new image without any tweaks or fiddling :-)

kedder · 2020-04-02T21:16:43Z

Wow :)

Please also check the responsiveness. How quickly does it react to pressure changes? That's something I'm not very confident about. Would it be possible to compare this aspect with the old image?

mihu-ov · 2020-04-02T21:36:44Z

I don´t think we can test responsiveness on pressure sensor data easily (even with an avionics test set). As we have changed variod only maybe we could generate a data file, replay it with sensord, record the speaker output and compare the two files?
Do we have to expect any additional delays due to pulseaudio or the new variod update method?

kedder · 2020-04-02T21:49:29Z

On old variod the vario tone was recalculated each 100ms. On the new variod the tone is calculated when pulseaudio sound buffer is exhausted. I'm not sure how big that buffer is and whether it will make any difference in terms of responsiveness. I don't expect hard numbers here, only if new version "feels laggy" or not.

kedder mentioned this issue Mar 11, 2020

Next preview Release #23

Closed

linuxianer99 added the bug Something isn't working label Mar 23, 2020

linuxianer99 added this to the For next release milestone Mar 23, 2020

This was referenced Apr 2, 2020

Port variod to pulseaudio Openvario/variod#1

Merged

Enable pulseaudio-based variod on testing image #45

Merged

mihu-ov mentioned this issue Apr 2, 2020

Enable USB drive automount #15

Merged

linuxianer99 closed this as completed in #45 Apr 10, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Vario sound is broken #18

Vario sound is broken #18

kedder commented Mar 9, 2020

CazYokoyama commented Mar 9, 2020 via email

kedder commented Mar 9, 2020

kedder commented Mar 9, 2020

mihu-ov commented Mar 9, 2020

kedder commented Mar 11, 2020 •

edited

Loading

CazYokoyama commented Mar 11, 2020

CazYokoyama commented Mar 12, 2020

kedder commented Mar 12, 2020

linuxianer99 commented Mar 12, 2020

linuxianer99 commented Mar 12, 2020

mihu-ov commented Mar 12, 2020

kedder commented Mar 12, 2020

kedder commented Mar 22, 2020 •

edited

Loading

mihu-ov commented Mar 22, 2020

kedder commented Mar 24, 2020

kedder commented Apr 1, 2020

kedder commented Apr 1, 2020

linuxianer99 commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020 •

edited

Loading

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

Vario sound is broken #18

Vario sound is broken #18

Comments

kedder commented Mar 9, 2020

CazYokoyama commented Mar 9, 2020 via email

kedder commented Mar 9, 2020

kedder commented Mar 9, 2020

mihu-ov commented Mar 9, 2020

kedder commented Mar 11, 2020 • edited Loading

CazYokoyama commented Mar 11, 2020

CazYokoyama commented Mar 12, 2020

kedder commented Mar 12, 2020

linuxianer99 commented Mar 12, 2020

linuxianer99 commented Mar 12, 2020

mihu-ov commented Mar 12, 2020

kedder commented Mar 12, 2020

kedder commented Mar 22, 2020 • edited Loading

mihu-ov commented Mar 22, 2020

kedder commented Mar 24, 2020

kedder commented Apr 1, 2020

kedder commented Apr 1, 2020

linuxianer99 commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020 • edited Loading

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

mihu-ov commented Apr 2, 2020

kedder commented Apr 2, 2020

kedder commented Mar 11, 2020 •

edited

Loading

kedder commented Mar 22, 2020 •

edited

Loading

kedder commented Apr 2, 2020 •

edited

Loading