On 2/17/2021 5:02 AM, Black Michael via wsjt-devel wrote:
I've also verified that reducing the volume level to eliminate the
effect does not reduce the power out either. My LP-700 wattmeter shows
the same wattage with 0 or -3.0dB
That could be caused by AGC within the rig. Or it could be in
.com]
Sent: 17. helmikuuta 2021 0:16
To: wsjt-devel@lists.sourceforge.net
Subject: Re: [wsjt-devel] Clipping
On 16/02/2021 22:10, Jim Brown wrote:
On 2/16/2021 2:01 PM, Bill Somerville wrote:
there are no analog audio stages involved. The trival distortion (lower than
-65dB
<mailto:g4...@classdesign.com>]
*Sent:* 17. helmikuuta 2021 0:16
*To:* wsjt-devel@lists.sourceforge.net
<mailto:wsjt-devel@lists.sourceforge.net>
*Subject:* Re: [wsjt-devel] Clipping
On 16/02/2021 22:10, Jim Brown wrote:
On 2/16/2021 2:01 PM, Bill Somerville wrote:
On 17/02/2021 11:08, Reino Talarmo wrote:
>From: Bill Somerville [mailto:g4...@classdesign.com]
Sent: 17. helmikuuta 2021 0:16
To: wsjt-devel@lists.sourceforge.net
Subject: Re: [wsjt-devel] Clipping
On 16/02/2021 22:10, Jim Brown wrote:
On 2/16/2021 2:01 PM, Bill Somervill
From: Bill Somerville [mailto:g4...@classdesign.com]
Sent: 17. helmikuuta 2021 14:01
To: wsjt-devel@lists.sourceforge.net
Subject: Re: [wsjt-devel] Clipping
On 17/02/2021 11:49, Bill Somerville wrote:
On 17/02/2021 11:08, Reino Talarmo wrote:
>From: Bill Somerville [mailto
Depending on the settings, the waterfall does a visual dynamic
compression, such that very weak signals are still visible next to very
strong signals. I.e. a -70 dB signal is well visible next to a -3 dB signal.
That's a feature, not a bug ;)
73
Stefan
On 17.02.21 12:08, Reino Talarmo wrote
On 17/02/2021 11:49, Bill Somerville wrote:
On 17/02/2021 11:08, Reino Talarmo wrote:
*>From:*Bill Somerville [mailto:g4...@classdesign.com]
*Sent:* 17. helmikuuta 2021 0:16
*To:* wsjt-devel@lists.sourceforge.net
*Subject:* Re: [wsjt-devel] Clipping
On 16/02/2021 22:10, Jim Brown wr
On 17/02/2021 11:08, Reino Talarmo wrote:
*>From:*Bill Somerville [mailto:g4...@classdesign.com]
*Sent:* 17. helmikuuta 2021 0:16
*To:* wsjt-devel@lists.sourceforge.net
*Subject:* Re: [wsjt-devel] Clipping
On 16/02/2021 22:10, Jim Brown wrote:
On 2/16/2021 2:01 PM, Bill Somerville wr
>From: Bill Somerville [mailto:g4...@classdesign.com]
Sent: 17. helmikuuta 2021 0:16
To: wsjt-devel@lists.sourceforge.net
Subject: Re: [wsjt-devel] Clipping
On 16/02/2021 22:10, Jim Brown wrote:
On 2/16/2021 2:01 PM, Bill Somerville wrote:
there are no analog audio stages involved.
On 16/02/2021 22:10, Jim Brown wrote:
On 2/16/2021 2:01 PM, Bill Somerville wrote:
there are no analog audio stages involved. The trival distortion
(lower than -65dB) is probably the result of being passed through an
unnecessary sample rate converter with a small defect within the MS
Windows a
On 2/16/2021 2:01 PM, Bill Somerville wrote:
there are no analog audio stages involved. The trival distortion (lower
than -65dB) is probably the result of being passed through an
unnecessary sample rate converter with a small defect within the MS
Windows audio sub-system. The issue can be demon
On 16/02/2021 20:53, Jim Brown wrote:
On 2/16/2021 12:40 PM, Black Michael via wsjt-devel wrote:
Also there with no transmitter involved. 2nd copy of WSJT-X
listening to the 1st WSJT_X transmit audio device.
But the audio stages of two computers. The measurements I cited were
at the audio ou
On 2/16/2021 12:40 PM, Black Michael via wsjt-devel wrote:
Also there with no transmitter involved. 2nd copy of WSJT-X listening
to the 1st WSJT_X transmit audio device.
But the audio stages of two computers. The measurements I cited were at
the audio output of a "good" computer generating a
Also there with no transmitter involved. 2nd copy of WSJT-X listening to the
1st WSJT_X transmit audio device.
But seems nobody cares about getting rid of unnecessary noise (albeit minimal
it's still noise).
On Tuesday, February 16, 2021, 02:25:25 PM CST, Jim Brown
wrote:
On 2/16/20
On 2/16/2021 5:56 AM, Black Michael via wsjt-devel wrote:
You're wrong -- it's being transmitted too.
I send that example a while ago.
That doesn't mean it's necessarily coming from the code -- the
distortion could be generated anywhere in the analog chain, from the
analog side of the D/A of
Mike,
reducing the output to eliminate a smaller than -65 dB source of noise
in the transmitted signal on Windows only is dancing on the head of a
pin. If you are concerned then feel free to reduce the Pwr slider a tad.
I have no wish to invoke the Qt volume control on all platforms, or even
You're wrong -- it's being transmitted too.I send that example a while ago.
Mike W9MDB
On Tuesday, February 16, 2021, 02:44:26 AM CST, Reino Talarmo
wrote:
>Patch to fix the problem
index 8e442d05f..81a978852 100644
--- a/widgets/mainwindow.cpp
+++ b/widgets/mainwindow.cpp
@@ -1
Well I guess can wait for Microsoft to fix it...lemme' hold my breath...
Problem verified on Windows 10 with VoiceMeter VAC, and MuzychenkoProblem
verified on another operator's IC7300 system Microsoft USB audio driver.So
that's 3 different audio drivers which indicates it's an OS problem to me.U
Hi Mike,
For further identification of the source of the issue, you could try to
generate a 100% modulated sine wave in audacity and play it back.
"Generate -> Tone -> Amplitude 1".
That will contain no clipping or distortion.
You can also use audacity to search for clipping "Analyze -> Find
Mike,
we have verified the output from WSJT-X is clean, this is not an issue
that needs fixing in WSJT-X. Note also that on Linux and macOS no such
distortion is present.
73
Bill
G4WJS.
On 16/02/2021 04:18, Black Michael via wsjt-devel wrote:
Patch to fix the problem
index 8e442d05f..8
No clipping found in wsjtx-2.4.0-rc1 with slider at 0 dB on linux,
feedback using pulseaudio null sink.
Unwanted tones are >70 dB reduced (tune function).
73
Stefan
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourcefor
>Patch to fix the problem
index 8e442d05f..81a978852 100644
--- a/widgets/mainwindow.cpp
+++ b/widgets/mainwindow.cpp
@@ -1288,6 +1288,7 @@ void MainWindow::readSettings()
// setup initial value of tx attenuator
m_block_pwr_tooltip = true;
ui->outAttenuation->setValue (m_settings->val
Patch to fix the problem
index 8e442d05f..81a978852 100644
--- a/widgets/mainwindow.cpp
+++ b/widgets/mainwindow.cpp
@@ -1288,6 +1288,7 @@ void MainWindow::readSettings()
// setup initial value of tx attenuator
m_block_pwr_tooltip = true;
ui->outAttenuation->setValue (m_settings->valu
.
73, Reino OH3mA
From: Bill Somerville [mailto:g4...@classdesign.com]
Sent: 15. helmikuuta 2021 15:32
To: wsjt-devel@lists.sourceforge.net
Subject: Re: [wsjt-devel] Clipping
Mike,
the "nastiness" you refer to is more than 65 dB down. As far as we have been
able to determine it
Mike,
the "nastiness" you refer to is more than 65 dB down. As far as we have
been able to determine it is due to either Windows or VAC. My money is
on Windows since VB Cable exhibits the same behaviour, although that is
not conclusive. Note that 65 dB down is below the distortion added by
al
It's not just the waterfallthat nastiness gets transmitted too.
Mike W9MDB
On Monday, February 15, 2021, 02:53:44 AM CST, Reino Talarmo
wrote:
Hi Mike and Bill,
I looked those *.wav files using v2.4.0-rc1 and noted interesting presentation
difference, when I had Flatten activa
"40 dB down is pretty good for most consumer systems, but could produce spurs
from it to someone who hears us very loud . If, for example, someone is 40 dB
over S9, harmonics that are 40 dB over S9 would be S9! Sort of negates the
advantage of some WSJT-X modes if the signal you're trying to cop
Several observations. First, MANY years ago thoughtful engineers
listened with headphones to the distortion output of instrumentation
devices (like legacy distortion analyzers) as "ear training" to help us
recognize a particular form of distortion. An very smart former Ampex
engineer designed a
Hi Mike and Bill,
I looked those *.wav files using v2.4.0-rc1 and noted interesting presentation
difference, when I had Flatten activated or non-activated. Especially the
waterfall display changed a lot. It was difficult to see those multiple
sidebands in non-flatten display totally independent
Since I've duplicated this problem on two computers with two different rigs and
two different audio chains (one USB and one VAC) this is fairly obvious a
WSJT-X problem.
Anything that adds noise to the bands is undesirable. Especially when it can
be cured by a small software change.
Mike W9MDB
"Do you still see harmonics if you drop WSJT_X Pwr level to -0.4 or so?"
I ran two tests with 2.4.0-rc1. One with Pwr at 0 dB and one with Pwr at about
-1.4 dB. The 0 dB test was extra. All other versions were tested with approx
-1.4 dB setting. (Why can't I read the setting by mousing over
I used Spectrum Lab to examine my TS-590S TX monitor audio for WSJT-X versions
1.9.1, 2.00, 2.2.1, and 2.3.0. All versions exhibit audio harmonics but each
is about 40dB below the wanted modulation.
There is a very noticeable difference in the primary modulation signal between
WSJT-X versions
The WAV files I just sent when loaded back into WSJT-X shows the noise in the
waterfall.
Mike
On Sunday, February 14, 2021, 03:14:10 PM CST, Bill Somerville
wrote:
Mike,
here's a spectrum plot of that .WAV file selected from just after the start
of the main signal up to just be
Mike,
here's a spectrum plot of that .WAV file selected from just after the
start of the main signal up to just before the end of the main signal. I
have zoomed in as far as Audacity allows on the minimum amplitude bottom
of the spectrum. It looks perfectly clean to me!
73
Bill
G4WJS.
On 14
Same FT8 mode.
Do you still see harmonics if you drop WSJT_X Pwr level to -0.4 or so?
Mike W9MDB
On Sunday, February 14, 2021, 02:54:19 PM CST, Andy Durbin
wrote:
"I used Spectrum Lab to examine my TS-590S TX monitor audio for WSJT-X
versions 1.9.1, 2.00, 2.2.1, and 2.3.0. All ve
Link to wavhttps://www.dropbox.com/s/cex6bfqfqzan65j/clipping.wav?dl=0
You should be able to duplicate this problem...
What I notice looking at this plot is the bottom side of the sine wave has more
points close to the edge then the upper side.
Mike W9MDB
On Sunday, February 14, 2021, 01
"I used Spectrum Lab to examine my TS-590S TX monitor audio for WSJT-X versions
1.9.1, 2.00, 2.2.1, and 2.3.0. All versions exhibit audio harmonics but each
is about 40dB below the wanted modulation."
Same test, same results, with 2.4.0-rc1. All tests with FT8. Was the reported
problem with
Hi Mike,
OK, let's reset here. Which mode are you seeing this with? Please use
Audacity to record a 16-bit 48,000 Hz .WAV file being generated by
WSJT-X using VAC. Also check VAC is using 48,000 16-bit as well. Make
sure you have the latest version of VAC (4.65.0). We can start by
analysing t
Sorry to confuse
The receive was showing the effect of the bad Tx signal that has artifacts.
I did another test on Tx to show you it's occurring there.
It's the Tx being clippedJTDX does something like 95% or 90% of max value
and does not have the problem.
BTW...the windows playback level
On 14/02/2021 18:22, Black Michael via wsjt-devel wrote:
I just confirmed this same problem on an IC-7300 via the USB audio.
You can see the hash noise on the waterfall.
Seems we need to reduce the maximum amplitude WSJT-X generates.
Adjusting the Windows playback level does not affect the beha
I just confirmed this same problem on an IC-7300 via the USB audio.You can see
the hash noise on the waterfall.
Seems we need to reduce the maximum amplitude WSJT-X generates.
Adjusting the Windows playback level does not affect the behavior.
Mike W9MDB
But it's showing up on transmit tooat a different level.
At -0.2 and below it's OK. It's way down but if everybody is doing this on the
band it will add to all the noise.
At -0.1 problems.
It looks like it's clipping the bottom of the sine wave
On Saturday, February 13, 2021, 08:40
Hi Mike,
another data point would be to check a saved .WAV file from WSJT-X for a
period when you see this distortion on the waterfall. If the 12,000 Hz
.WAV data is clean then the issue must be in the waterfall processing.
73
Bill
G4WJS.
On 13/02/2021 14:37, Bill Somerville wrote:
Hi Mike,
Hi Mike,
that could be a minor rounding error in one of the filters in WSJT-X,
not sure there's anything to worry about as any input signal that is
that close to 0dB is bound to clip at some point anyway, i.e. input
levels should not be anywhere near that high.
73
Bill
G4WJS.
On 13/02/2021
This is with 2.4.0-rc1 by the way...
Audacity shows a perfectly clean signalas does FLDigi. Only WSJT-X and
JTDX show problems with the signal. Common factor is Qt sound input.
Mike
On Saturday, February 13, 2021, 08:13:45 AM CST, Bill Somerville
wrote:
Hi Mike,
two things
Hi Mike,
two things to check.
1) make sure that either volume controls are disabled in the VAC control
panel, or that the master level for the VAC device you are using is set
no higher than 0 dB.
2) record the Tx audio coming out of the VAC device with Audacity or
similar and check the wave
I'll add one more data point. Looking at the same signal in FLDigi I don't see
any problem...nice and clean at 0dB attenuation.
Only problem is in WSJT-X and JTDX.
Mike W9MDB
On Saturday, February 13, 2021, 07:53:25 AM CST, Black Michael
wrote:
If you have a virtual audio cable you
If you have a virtual audio cable you can do the same test.
This has nothing to do with the rig -- it's either Virtual Audio Cable (which I
doubt) or the Qt sound process. I've seen similar behavior from the Qt sound
device before.
I can run this test with the rig turned off and get the same res
Hi Mike,
It seems that you TX is very sensitive to a small overdrive! Which is rig you
are using and how you have set it at 0 dB level? It looks like a digital
processing goes wild due to a misunderstood sign bit or something similar that
generates a huge step in analog value due to misbehaving
49 matches
Mail list logo
