Re: [wsjt-devel] Exaggerated reports given to lower tone freqs

[Bill Somerville via wsjt-devel]

On 17/10/2021 21:38, Marco Calistri via wsjt-devel wrote:

Il 17/10/21 17:31, Bill Somerville via wsjt-devel ha scritto:

On 17/10/2021 21:11, Marco Calistri via wsjt-devel wrote:

Il 17/10/21 16:52, Kevin McQuiggin (SFU) ha scritto:

Hi Marco:

Did you follow the steps to record a reference spectrum first?  You 
have to record a minute or so of noise in order for the program to 
be able to compute gain figures which are then used to flatten 
frequency response.  I am not sure what would happen if you 
activated reference spectrum without recording it first.


Kevin

Sent from my iPad

On Oct 16, 2021, at 15:11, Marco Calistri via wsjt-devel 
 wrote:



Il 16/10/21 17:30, Kevin McQuiggin via wsjt-devel ha scritto:

Hi All:

There is a pretty good article on the why and how of reference 
spectrum use by Bob KA1GT at 
http://www.bobatkins.com/radio/WSJTX_flatten.html.  His site is 
targeted primarily at EME and UHF-and-above users, but the 
information applies to use of the reference spectrum at HF as well.


73,

Kevin



Hi Kevin,

I found this topic very interesting.

I've followed the step by step described in the article by KA1GT 
but when I click on Ref Spec my WSJT-X crashes!


I'm using WSJT-X 2.5.0 on Linux openSUSE Tumbleweed.

Regards,

Yes Kevin!

For this reason I went disappointed:

I followed the step by step from the URL you shared here regarding 
how to create a reference spectrum and then I checked that a new 
refspec.dat was produced inside the folder, but it seems that 
something in such file is not appreciated by WSJT-X and it keep 
crashing until I renamed such file to refspec.old.


marco@localhost:~> find   ~/.local/share/WSJT-X/ -name "refspec*"  -exec ls -la 
{} \;
-rw-r--r-- 1 marco users 204045 16 ott 19.35 
/home/marco/.local/share/WSJT-X/refspec.old

I don't know why, I know that in this way it works!


---
*73 de Marco, PY1ZRJ (former IK5BCU)*


Hi Marco,

send me (g4wjs  classdesign  com) your refspec.dat file that 
causes a crash please?


73
Bill
G4WJS.



Hi Bill!

I'll do it!
---
*73 de Marco, PY1ZRJ (former IK5BCU)*


Hi Marco,

your reference spectrum file works FB for me, I don't see any issues 
with it. Here is the plot of it from the Equalization Tools dialog:


73
Bill
G4WJS.
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Exaggerated reports given to lower tone freqs

[Marco Calistri via wsjt-devel]

  
Il 17/10/21 17:31, Bill Somerville via
  wsjt-devel ha scritto:


  
  On 17/10/2021 21:11, Marco Calistri
via wsjt-devel wrote:
  
  
Il 17/10/21 16:52, Kevin McQuiggin
  (SFU) ha scritto:

 Hi
  Marco:
  
  
  Did you follow the steps to record a reference spectrum
first?  You have to record a minute or so of noise in order
for the program to be able to compute gain figures which are
then used to flatten frequency response.  I am not sure what
would happen if you activated reference spectrum without
recording it first.  
  
  
  Kevin

Sent from my iPad

  On Oct 16, 2021, at 15:11, Marco Calistri via wsjt-devel
  
  wrote:
  


  
Il 16/10/21 17:30, Kevin
  McQuiggin via wsjt-devel ha scritto:


  Hi All:
  
  
  There is a pretty good article on the
why and how of reference spectrum use by Bob KA1GT
at http://www.bobatkins.com/radio/WSJTX_flatten.html.
 His site is targeted primarily at EME and
UHF-and-above users, but the information applies to
use of the reference spectrum at HF as well.
  
  
  73,
  
  
  Kevin
  
  
  

Hi Kevin,

I found this topic very interesting.

I've followed the step by step described in the article
by KA1GT but when I click on Ref Spec my WSJT-X crashes!

I'm using WSJT-X 2.5.0 on Linux openSUSE Tumbleweed.

Regards,
  

  

Yes Kevin!

For this reason I went disappointed:

I followed the step by step from the URL you shared here
regarding how to create a reference spectrum and then I checked
that a new refspec.dat was produced inside the folder, but it
seems that something in such file is not appreciated by WSJT-X
and it keep crashing until I renamed such file to refspec.old.

marco@localhost:~> find   ~/.local/share/WSJT-X/ -name "refspec*"  -exec ls -la {} \;
-rw-r--r-- 1 marco users 204045 16 ott 19.35 /home/marco/.local/share/WSJT-X/refspec.old


I don't know why, I know that in this way it works!


---
  73 de Marco, PY1ZRJ (former IK5BCU)
  
  Hi Marco,
  send me (g4wjs  classdesign  com) your
refspec.dat file that causes a crash please?
  73
Bill
G4WJS.
  
  

Hi Bill!

I'll do it!
---
  73 de Marco, PY1ZRJ (former IK5BCU)
  
  
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Exaggerated reports given to lower tone freqs

[Bill Somerville via wsjt-devel]

On 17/10/2021 21:11, Marco Calistri via wsjt-devel wrote:

Il 17/10/21 16:52, Kevin McQuiggin (SFU) ha scritto:

Hi Marco:

Did you follow the steps to record a reference spectrum first?  You 
have to record a minute or so of noise in order for the program to be 
able to compute gain figures which are then used to flatten frequency 
response.  I am not sure what would happen if you activated reference 
spectrum without recording it first.


Kevin

Sent from my iPad

On Oct 16, 2021, at 15:11, Marco Calistri via wsjt-devel 
 wrote:



Il 16/10/21 17:30, Kevin McQuiggin via wsjt-devel ha scritto:

Hi All:

There is a pretty good article on the why and how of reference 
spectrum use by Bob KA1GT at 
http://www.bobatkins.com/radio/WSJTX_flatten.html.  His site is 
targeted primarily at EME and UHF-and-above users, but the 
information applies to use of the reference spectrum at HF as well.


73,

Kevin



Hi Kevin,

I found this topic very interesting.

I've followed the step by step described in the article by KA1GT but 
when I click on Ref Spec my WSJT-X crashes!


I'm using WSJT-X 2.5.0 on Linux openSUSE Tumbleweed.

Regards,

Yes Kevin!

For this reason I went disappointed:

I followed the step by step from the URL you shared here regarding how 
to create a reference spectrum and then I checked that a new 
refspec.dat was produced inside the folder, but it seems that 
something in such file is not appreciated by WSJT-X and it keep 
crashing until I renamed such file to refspec.old.


marco@localhost:~> find   ~/.local/share/WSJT-X/ -name "refspec*"  -exec ls -la 
{} \;
-rw-r--r-- 1 marco users 204045 16 ott 19.35 
/home/marco/.local/share/WSJT-X/refspec.old

I don't know why, I know that in this way it works!


---
*73 de Marco, PY1ZRJ (former IK5BCU)*


Hi Marco,

send me (g4wjs  classdesign  com) your refspec.dat file that 
causes a crash please?


73
Bill
G4WJS.
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Exaggerated reports given to lower tone freqs

[Marco Calistri via wsjt-devel]

Il 17/10/21 16:52, Kevin McQuiggin (SFU) ha scritto:

Hi Marco:

Did you follow the steps to record a reference spectrum first?  You 
have to record a minute or so of noise in order for the program to be 
able to compute gain figures which are then used to flatten frequency 
response.  I am not sure what would happen if you activated reference 
spectrum without recording it first.


Kevin

Sent from my iPad

On Oct 16, 2021, at 15:11, Marco Calistri via wsjt-devel 
 wrote:



Il 16/10/21 17:30, Kevin McQuiggin via wsjt-devel ha scritto:

Hi All:

There is a pretty good article on the why and how of reference 
spectrum use by Bob KA1GT at 
http://www.bobatkins.com/radio/WSJTX_flatten.html.  His site is 
targeted primarily at EME and UHF-and-above users, but the 
information applies to use of the reference spectrum at HF as well.


73,

Kevin



Hi Kevin,

I found this topic very interesting.

I've followed the step by step described in the article by KA1GT but 
when I click on Ref Spec my WSJT-X crashes!


I'm using WSJT-X 2.5.0 on Linux openSUSE Tumbleweed.

Regards,

Yes Kevin!

For this reason I went disappointed:

I followed the step by step from the URL you shared here regarding how 
to create a reference spectrum and then I checked that a new refspec.dat 
was produced inside the folder, but it seems that something in such file 
is not appreciated by WSJT-X and it keep crashing until I renamed such 
file to refspec.old.


marco@localhost:~> find   ~/.local/share/WSJT-X/ -name "refspec*"  -exec ls -la 
{} \;
-rw-r--r-- 1 marco users 204045 16 ott 19.35 
/home/marco/.local/share/WSJT-X/refspec.old


I don't know why, I know that in this way it works!


---
*73 de Marco, PY1ZRJ (former IK5BCU)*
**___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Exaggerated reports given to lower tone freqs

[Kevin McQuiggin (SFU) via wsjt-devel]

Hi Marco:

Did you follow the steps to record a reference spectrum first?  You have to 
record a minute or so of noise in order for the program to be able to compute 
gain figures which are then used to flatten frequency response.  I am not sure 
what would happen if you activated reference spectrum without recording it 
first.  

Kevin

Sent from my iPad

> On Oct 16, 2021, at 15:11, Marco Calistri via wsjt-devel 
>  wrote:
> 
> Il 16/10/21 17:30, Kevin McQuiggin via wsjt-devel ha scritto:
>> Hi All:
>> 
>> There is a pretty good article on the why and how of reference spectrum use 
>> by Bob KA1GT at http://www.bobatkins.com/radio/WSJTX_flatten.html.  His site 
>> is targeted primarily at EME and UHF-and-above users, but the information 
>> applies to use of the reference spectrum at HF as well.
>> 
>> 73,
>> 
>> Kevin
>> 
>> 
> Hi Kevin,
> 
> I found this topic very interesting.
> 
> I've followed the step by step described in the article by KA1GT but when I 
> click on Ref Spec my WSJT-X crashes!
> 
> I'm using WSJT-X 2.5.0 on Linux openSUSE Tumbleweed.
> 
> Regards,
> ---
> 73 de Marco, PY1ZRJ (former IK5BCU)
> 
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Some minor issues.

[Richard Shaw via wsjt-devel]

On Sun, Oct 17, 2021 at 9:08 AM Marco Calistri via wsjt-devel <
wsjt-devel@lists.sourceforge.net> wrote:

> Il 17/10/21 03:47, jarmo via wsjt-devel ha scritto:
>
> Wed, 29 Sep 2021 16:15:00 + (UTC)
> Black Michael via wsjt-devel  
> 
> kirjoitti:
>
> Hi Mike
>
> Compiled latest git version 4.4 and no cahnge with TS-590SG.
> We can run wsjtx (2.5.0) when radio set as TS-590SG, 9600 rest
> as default. Audio goes out.
> But again, if I set there /usr/local/bin/rigctld -m 2037 -r /dev/radio
> -s 9600, wsjt setting radio set as Hamlib Net Rigctl, audio does not
> go out, just tiny sound goes, but no matter how much turn output +
> it stays kind of "muted"
>
>
I was having the same issue using hamlib's flrig interface, however, I
updated both hamlib and wsjtx to the latest release and have new behavior:

1. When in WSJT-X settings "Test CAT" works fine as well as "Test PTT".
2. When trying a QSO or "Tune", I get rapid relay clicking and WSJT-X loses
connection to hamlib randomly after < 1 to a couple of seconds. Clicking
retry always works until the next TX attempt.

Changing to the FT-991 hamlib interface everything works as expected, but
of course I lose my Flrig interface.

Thanks,
Richard
KF5OIM
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Wsjtx crash

[Kari Sillanmäki via wsjt-devel]

Hi Marco,

On 17.10.2021 17.17, Marco Calistri via wsjt-devel wrote:



Hi Kari!

I generated the refspec.dat by following the instructions and it has 
been produced correctly but this doesn't avoid WSJT-X 2.5.0 Linux to 
crash!


-rw-r--r-- 1 marco users 204045 16 ott 19.35 
/home/marco/.local/share/WSJT-X/refspec.dat


That seems to be the right location and the file size is same as mine.

Try as I might, I could not make WSJT-X 2.5.0 to crash using ref-spec.

I tried in 2 systems ( XUbuntu 18.04.6 LTS and Raspbian GNU/Linux 10 
(buster) ). On both WSJT-X was compiled locally from source.


I tested with refspec.dat after measuring ref-spec, with empty ( length 
0 bytes ) file and with nonexistent refspec.dat  file.


No crashes in any situation...

Do you get an error message when WSJT-X crashes?

What happens if you delete refspec.dat ( or rename it to something else) 
? Does WSJT-X start then?


73's de Kari, oh2gqc


___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Rig LF Rolloff

[Virginia Greene via wsjt-devel]

Mike,

Your solution is a great possibility.

I’ve worked out a solution that’s fine for me.  I’ve set my radio up so that 
the frequency response at the low end starts falling off around 200 Hz.  And, 
it’s easy to switch between modes.  Problem solved.  For my end.

The problem is that I have no control over any Foxes.  Plus, it’s to every 
Hound’s benefit to get this right.  It would be wonderful if there was a fix 
that could be universally applied without relying on users to measure their 
radio’s frequency response and being able to apply whatever fix works for them.

Clarke

> On Oct 17, 2021, at 10:17 AM, Black Michael via wsjt-devel 
>  wrote:
> 
> On more potential solution -- In fox mode run pseudo-split and drop the 
> frequency 500Hz automaticallyi.e. Fake-It mode for both RX & TX.
> 
> Mike W9MDB
> 
> 
> On Sunday, October 17, 2021, 09:11:52 AM CDT, Virginia Greene via wsjt-devel 
>  wrote:
> 
> 
> Bill,
> 
> This actually kept me up last night!
> 
> For “Normal” FT8 operation, your solution makes 100% perfect sense.  The only 
> downside I can see is that you lose that 200 Hz or 300 Hz at the top end of 
> the band.  That is a very fair trade-off.
> 
> But, in DXpedition mode with a Fox and Hounds, it gets murkier.
> 
> First, consider the Hound end.  Let’s assume that the Hound knows that 
> his/her radio rolls off at the low end and really should be offset by 200 Hz, 
> so s/he shifts the VFO by 200 Hz.  If the Fox is using the default TX 
> frequency of 300 Hz, that fools the Hound’s system into believing that the 
> Fox is now at 500 Hz, where the Hound’s receiver hears well.  That all should 
> work well.
> 
> But, imagine that the Fox has five streams running.  The lowest would appear 
> at 500 Hz for the Hound, with subsequent ones beginning at 560 Hz, 620 Hz, 
> 680 Hz, and 740 Hz.
> 
> If the Hound does not get through in three tries or the Fox fades so that 
> Hound doesn't receive a transmission from the Hound, won’t the Hound get 
> moved to the “repechage” band to get another couple minutes of attempts to 
> send the signal report?  When the Fox is at 300 Hz and up, that repechage 
> band is 300 Hz above the Fox TX frequencies, where the Fox can still hear 
> callers.  BUT, if the Fox is at 500 Hz TX or above, won’t the Hound get moved 
> 300 Hz *below* the Fox TX frequencies?  As in, perhaps, 0 Hz relative to the 
> bottom of FT8 channel?  In this case, the Fox would never be able to hear the 
> Hound there.
> 
> So, when a Hound offsets her/his frequency by 200 Hz to accommodate the 
> receiver low frequency roll-off, s/he essentially eliminates use of the 
> repechage band.  That may be the price a Hound pays for not having a properly 
> working FT8 radio.  That and remembering to call above 1200 Hz.
> 
> Now, think about when a Fox is using a radio where the receive passband rolls 
> off below 400 or so Hz, as is all too often the case.  Just look at the most 
> popular radios used on DXpeditions.
> 
> With a single transmit stream, the Hounds will move to 300 Hz to send their 
> reports.  Unless the Hound is at least as many dB stronger than the filter 
> roll-off, the Hound might get lost in the noise when s/he responds.  That 
> leads to multiple decode cycles being used, tying up the Fox.  The rate 
> drops, not only due to the immediate need for multiple decode cycles that 
> otherwise might not be necessary, but also because the Hound likely will 
> start the same dance again to initiate a contact.  That presumes that the Fox 
> does not log the contact while the Hound is in the repechage zone.  (My 
> observation is that most Hounds just go off and call again.  Why?  Dunno.  I 
> also notice that very few repechage contacts get made for whatever reason.)
> 
> I’ve observed the above many, many times.  Even just yesterday.
> 
> It seems to me that there’s a few solutions for this.
> 
> One is that all Hounds should be armed for bear with large antennas and big 
> power.  That way, they’ll overcome the Fox’s receiver filter roll-off.  Of 
> course, that precludes weaker signal contacts for tough propagation paths 
> where the Hound only merits a -18 signal report from the Fox to begin with.  
> Plus, it’s kind of against the spirit of FT8 being a weak signal mode usable 
> by stations with more moderate power.
> 
> A second would be for Foxes to all use radios that receive well to below 300 
> Hz.  Problem solved for the Fox end, which really affects all the Hounds 
> calling, too.  That still puts the burden on all the Hounds to have receivers 
> that are flat to below 300 Hz or shift their VFO by 200 Hz and sacrifice the 
> ability to use the repechage band .  Maybe that’s just tough on the Hounds 
> who don’t know enough to do that.
> 
> A third solution is for the Fox to move his or her TX frequency to 700 Hz.  
> With five streams, the topmost one would begin at 940 Hz - still good.  
> Essentially every Hound would have no problem, even if their receiver 

Re: [wsjt-devel] Wsjtx crash

[Marco Calistri via wsjt-devel]

  
Il 17/10/21 09:49, Reino Talarmo via
  wsjt-devel ha scritto:


  
  
  
  
I face the same crash on
linux by activating ref spec in the waterfall and enable
monitor in the main window, then I would like to resolve
this since it seems that enabling reference spectrum could
provide a better decoding(?).

   


  BTW also if decoding wouldn't
  enhanced I dislike the fact of program crashes..
  Hi Marco,
  Most probably you would not see any
  enhancement on narrow bandwidth modes like FT8 and FT4.
  Some EME modes could gain the last missing part of a dB.
  73, Reino OH3mA

  
  

Ok Reino!

I don't work EME, BTW I desire to understand the cause of program
crash when selecting Ref Spec.


---
  73 de Marco, PY1ZRJ (former IK5BCU)
  
  
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Rig LF Rolloff

[Black Michael via wsjt-devel]

On more potential solution -- In fox mode run pseudo-split and drop the 
frequency 500Hz automaticallyi.e. Fake-It mode for both RX & TX.
Mike W9MDB 

On Sunday, October 17, 2021, 09:11:52 AM CDT, Virginia Greene via 
wsjt-devel  wrote:  
 
 Bill,
This actually kept me up last night!
For “Normal” FT8 operation, your solution makes 100% perfect sense.  The only 
downside I can see is that you lose that 200 Hz or 300 Hz at the top end of the 
band.  That is a very fair trade-off.
But, in DXpedition mode with a Fox and Hounds, it gets murkier.
First, consider the Hound end.  Let’s assume that the Hound knows that his/her 
radio rolls off at the low end and really should be offset by 200 Hz, so s/he 
shifts the VFO by 200 Hz.  If the Fox is using the default TX frequency of 300 
Hz, that fools the Hound’s system into believing that the Fox is now at 500 Hz, 
where the Hound’s receiver hears well.  That all should work well.
But, imagine that the Fox has five streams running.  The lowest would appear at 
500 Hz for the Hound, with subsequent ones beginning at 560 Hz, 620 Hz, 680 Hz, 
and 740 Hz.
If the Hound does not get through in three tries or the Fox fades so that Hound 
doesn't receive a transmission from the Hound, won’t the Hound get moved to the 
“repechage” band to get another couple minutes of attempts to send the signal 
report?  When the Fox is at 300 Hz and up, that repechage band is 300 Hz above 
the Fox TX frequencies, where the Fox can still hear callers.  BUT, if the Fox 
is at 500 Hz TX or above, won’t the Hound get moved 300 Hz *below* the Fox TX 
frequencies?  As in, perhaps, 0 Hz relative to the bottom of FT8 channel?  In 
this case, the Fox would never be able to hear the Hound there.
So, when a Hound offsets her/his frequency by 200 Hz to accommodate the 
receiver low frequency roll-off, s/he essentially eliminates use of the 
repechage band.  That may be the price a Hound pays for not having a properly 
working FT8 radio.  That and remembering to call above 1200 Hz.
Now, think about when a Fox is using a radio where the receive passband rolls 
off below 400 or so Hz, as is all too often the case.  Just look at the most 
popular radios used on DXpeditions.
With a single transmit stream, the Hounds will move to 300 Hz to send their 
reports.  Unless the Hound is at least as many dB stronger than the filter 
roll-off, the Hound might get lost in the noise when s/he responds.  That leads 
to multiple decode cycles being used, tying up the Fox.  The rate drops, not 
only due to the immediate need for multiple decode cycles that otherwise might 
not be necessary, but also because the Hound likely will start the same dance 
again to initiate a contact.  That presumes that the Fox does not log the 
contact while the Hound is in the repechage zone.  (My observation is that most 
Hounds just go off and call again.  Why?  Dunno.  I also notice that very few 
repechage contacts get made for whatever reason.)
I’ve observed the above many, many times.  Even just yesterday.
It seems to me that there’s a few solutions for this.
One is that all Hounds should be armed for bear with large antennas and big 
power.  That way, they’ll overcome the Fox’s receiver filter roll-off.  Of 
course, that precludes weaker signal contacts for tough propagation paths where 
the Hound only merits a -18 signal report from the Fox to begin with.  Plus, 
it’s kind of against the spirit of FT8 being a weak signal mode usable by 
stations with more moderate power.
A second would be for Foxes to all use radios that receive well to below 300 
Hz.  Problem solved for the Fox end, which really affects all the Hounds 
calling, too.  That still puts the burden on all the Hounds to have receivers 
that are flat to below 300 Hz or shift their VFO by 200 Hz and sacrifice the 
ability to use the repechage band .  Maybe that’s just tough on the Hounds who 
don’t know enough to do that.
A third solution is for the Fox to move his or her TX frequency to 700 Hz.  
With five streams, the topmost one would begin at 940 Hz - still good.  
Essentially every Hound would have no problem, even if their receiver rolled 
off below 400 Hz.  Any Hound that gets moved to the repechage band would get 
placed 300 Hz below the Fox TX frequency, which means they’d start at 400 Hz - 
pretty much within almost all Fox receivers, even with low frequency roll-off.  
(Isn’t that how it works?)
The problem with all of these is that it requires cooperation from everybody to 
make it work.  People would need to read the manual and know where their 
receiver rolls off.  That’s easy enough to discover, but how many FT8 users 
actually do?  Then, they’d need to adjust, repair, or replace their radio to 
have no meaningful low frequency receiver roll-off.
My own solution, if I was king of the WSJT universe, would be to make the 
default FT8 Fox TX frequency 700 Hz.  Almost every Fox just uses the default 
frequency.  Why should they know any different?  

Re: [wsjt-devel] Wsjtx crash

[Marco Calistri via wsjt-devel]

  
Il 17/10/21 09:58, Kari Sillanmäki via
  wsjt-devel ha scritto:

Hi
  Marco,
  
  
  The refspec.dat file is in the log directory, File => Open log
  directory
  
  
  73's de Kari, oh2gqc
  
  
  kOn 17.10.2021 15.33, Marco Calistri via wsjt-devel wrote:
  
  Jarmo,


Renamed in what?


I face the same crash on linux by activating ref spec in the
waterfall and enable monitor in the main window, then I would
like to resolve this since it seems that enabling reference
spectrum could provide a better decoding(?).


BTW also if decoding wouldn't enhanced I dislike the fact of
program crashes..


Many thanks!


PY1ZRJ


  

Hi Kari!

I generated the refspec.dat by following the instructions and it has
been produced correctly but this doesn't avoid WSJT-X 2.5.0 Linux to
crash!

-rw-r--r-- 1 marco users 204045 16 ott 19.35 /home/marco/.local/share/WSJT-X/refspec.dat
---
  73 de Marco, PY1ZRJ (former IK5BCU)
  
  
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Rig LF Rolloff

[Virginia Greene via wsjt-devel]

Bill,

This actually kept me up last night!

For “Normal” FT8 operation, your solution makes 100% perfect sense.  The only 
downside I can see is that you lose that 200 Hz or 300 Hz at the top end of the 
band.  That is a very fair trade-off.

But, in DXpedition mode with a Fox and Hounds, it gets murkier.

First, consider the Hound end.  Let’s assume that the Hound knows that his/her 
radio rolls off at the low end and really should be offset by 200 Hz, so s/he 
shifts the VFO by 200 Hz.  If the Fox is using the default TX frequency of 300 
Hz, that fools the Hound’s system into believing that the Fox is now at 500 Hz, 
where the Hound’s receiver hears well.  That all should work well.

But, imagine that the Fox has five streams running.  The lowest would appear at 
500 Hz for the Hound, with subsequent ones beginning at 560 Hz, 620 Hz, 680 Hz, 
and 740 Hz.

If the Hound does not get through in three tries or the Fox fades so that Hound 
doesn't receive a transmission from the Hound, won’t the Hound get moved to the 
“repechage” band to get another couple minutes of attempts to send the signal 
report?  When the Fox is at 300 Hz and up, that repechage band is 300 Hz above 
the Fox TX frequencies, where the Fox can still hear callers.  BUT, if the Fox 
is at 500 Hz TX or above, won’t the Hound get moved 300 Hz *below* the Fox TX 
frequencies?  As in, perhaps, 0 Hz relative to the bottom of FT8 channel?  In 
this case, the Fox would never be able to hear the Hound there.

So, when a Hound offsets her/his frequency by 200 Hz to accommodate the 
receiver low frequency roll-off, s/he essentially eliminates use of the 
repechage band.  That may be the price a Hound pays for not having a properly 
working FT8 radio.  That and remembering to call above 1200 Hz.

Now, think about when a Fox is using a radio where the receive passband rolls 
off below 400 or so Hz, as is all too often the case.  Just look at the most 
popular radios used on DXpeditions.

With a single transmit stream, the Hounds will move to 300 Hz to send their 
reports.  Unless the Hound is at least as many dB stronger than the filter 
roll-off, the Hound might get lost in the noise when s/he responds.  That leads 
to multiple decode cycles being used, tying up the Fox.  The rate drops, not 
only due to the immediate need for multiple decode cycles that otherwise might 
not be necessary, but also because the Hound likely will start the same dance 
again to initiate a contact.  That presumes that the Fox does not log the 
contact while the Hound is in the repechage zone.  (My observation is that most 
Hounds just go off and call again.  Why?  Dunno.  I also notice that very few 
repechage contacts get made for whatever reason.)

I’ve observed the above many, many times.  Even just yesterday.

It seems to me that there’s a few solutions for this.

One is that all Hounds should be armed for bear with large antennas and big 
power.  That way, they’ll overcome the Fox’s receiver filter roll-off.  Of 
course, that precludes weaker signal contacts for tough propagation paths where 
the Hound only merits a -18 signal report from the Fox to begin with.  Plus, 
it’s kind of against the spirit of FT8 being a weak signal mode usable by 
stations with more moderate power.

A second would be for Foxes to all use radios that receive well to below 300 
Hz.  Problem solved for the Fox end, which really affects all the Hounds 
calling, too.  That still puts the burden on all the Hounds to have receivers 
that are flat to below 300 Hz or shift their VFO by 200 Hz and sacrifice the 
ability to use the repechage band .  Maybe that’s just tough on the Hounds who 
don’t know enough to do that.

A third solution is for the Fox to move his or her TX frequency to 700 Hz.  
With five streams, the topmost one would begin at 940 Hz - still good.  
Essentially every Hound would have no problem, even if their receiver rolled 
off below 400 Hz.  Any Hound that gets moved to the repechage band would get 
placed 300 Hz below the Fox TX frequency, which means they’d start at 400 Hz - 
pretty much within almost all Fox receivers, even with low frequency roll-off.  
(Isn’t that how it works?)

The problem with all of these is that it requires cooperation from everybody to 
make it work.  People would need to read the manual and know where their 
receiver rolls off.  That’s easy enough to discover, but how many FT8 users 
actually do?  Then, they’d need to adjust, repair, or replace their radio to 
have no meaningful low frequency receiver roll-off.

My own solution, if I was king of the WSJT universe, would be to make the 
default FT8 Fox TX frequency 700 Hz.  Almost every Fox just uses the default 
frequency.  Why should they know any different?  They’re DXpeditioners, after 
all, focused on going to some rare place with all the associated challenges and 
working as many stations as they can.  They presume that 300 Hz is optimum - 
why else would it be chosen?  

There is nothing 

Re: [wsjt-devel] Some minor issues.

[Marco Calistri via wsjt-devel]

  
Il 17/10/21 03:47, jarmo via wsjt-devel
  ha scritto:


  Wed, 29 Sep 2021 16:15:00 + (UTC)
Black Michael via wsjt-devel 
kirjoitti:

Hi Mike

Compiled latest git version 4.4 and no cahnge with TS-590SG.
We can run wsjtx (2.5.0) when radio set as TS-590SG, 9600 rest
as default. Audio goes out.
But again, if I set there /usr/local/bin/rigctld -m 2037 -r /dev/radio
-s 9600, wsjt setting radio set as Hamlib Net Rigctl, audio does not
go out, just tiny sound goes, but no matter how much turn output +
it stays kind of "muted"

Jarmo, OH1MRR

  
Also...please the the git clone so you can get any patches I put
in.The tar file only gets updated overnight.

 

On Wednesday, September 29, 2021, 11:03:57 AM CDT, jarmo via
wsjt-devel  wrote: 
 Wed, 29 Sep 2021 12:05:31 -0300
Marco Calistri via wsjt-devel 
kirjoitti:



  Il 29/09/21 11:07, Black Michael ha scritto:  

  
Can you please try the latest hamlib and see if it's fixed now?  

  


Unfortunately not working. Compiled
hamlib-4.4~git-dd1376be-20210929.tar.gz and same behaviour.
If try to use NetHamlib rictl, audio does not go into radio.
Cat control works ok.
But if chosen Kenwood ts-590SG in WSJTX radio preferences, audio works
ok...

Jarmo




  

This exactly what happens with me using YAESU FT-100.

The only way to go is by selecting Data/Pkt and not USB in
  WSJT-X.

Once again I wan to to repeat that some months ago - supposedly with
different Hamlib coding - I could use USB.


---
  73 de Marco, PY1ZRJ (former IK5BCU)
  
  
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Wsjtx crash

[Kari Sillanmäki via wsjt-devel]

Hi Marco,

The refspec.dat file is in the log directory, File => Open log directory

73's de Kari, oh2gqc

kOn 17.10.2021 15.33, Marco Calistri via wsjt-devel wrote:

Jarmo,

Renamed in what?

I face the same crash on linux by activating ref spec in the waterfall 
and enable monitor in the main window, then I would like to resolve this 
since it seems that enabling reference spectrum could provide a better 
decoding(?).


BTW also if decoding wouldn't enhanced I dislike the fact of program 
crashes..


Many thanks!

PY1ZRJ

Scarica Outlook per Android 


*Da:* jarmo via wsjt-devel 
*Inviato:* domenica 17 ottobre 2021, 02:53
*A:* wsjt-devel@lists.sourceforge.net
*Cc:* jarmo
*Oggetto:* Re: [wsjt-devel] Wsjtx crash

Sun, 17 Oct 2021 04:45:47 + (UTC)
Black Michael via wsjt-devel 
kirjoitti:


Probably in the same directory as the wsjtx executable.
Mike W9MDB


Tnx, I try look...
By the way, renamed refspec.dat and started wsjtx,
now works ok.

Jarmo, oh1mrr


___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel 





___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel






___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Wsjtx crash

[Reino Talarmo via wsjt-devel]

I face the same crash on linux by activating ref spec in the waterfall and
enable monitor in the main window, then I would like to resolve this since
it seems that enabling reference spectrum could provide a better
decoding(?).

 

BTW also if decoding wouldn't enhanced I dislike the fact of program
crashes..

Hi Marco,

Most probably you would not see any enhancement on narrow bandwidth modes
like FT8 and FT4. Some EME modes could gain the last missing part of a dB.

73, Reino OH3mA

___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Wsjtx crash

[Marco Calistri via wsjt-devel]

Jarmo,

Renamed in what?

I face the same crash on linux by activating ref spec in the waterfall and 
enable monitor in the main window, then I would like to resolve this since it 
seems that enabling reference spectrum could provide a better decoding(?).

BTW also if decoding wouldn't enhanced I dislike the fact of program crashes..

Many thanks!

PY1ZRJ

Scarica Outlook per Android


Da: jarmo via wsjt-devel 
Inviato: domenica 17 ottobre 2021, 02:53
A: wsjt-devel@lists.sourceforge.net
Cc: jarmo
Oggetto: Re: [wsjt-devel] Wsjtx crash

Sun, 17 Oct 2021 04:45:47 + (UTC)
Black Michael via wsjt-devel 
kirjoitti:

> Probably in the same directory as the wsjtx executable.
> Mike W9MDB

Tnx, I try look...
By the way, renamed refspec.dat and started wsjtx,
now works ok.

Jarmo, oh1mrr


___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Wsjtx crash

[jarmo via wsjt-devel]

Sun, 17 Oct 2021 11:31:05 +0100
Bill Somerville via wsjt-devel 
kirjoitti:


> Hi Jarmo,
> 
> the button to stop collecting reference spectrum data is the one 
> labelled "Stop" on the main window.
> 
> 73
> Bill
> G4WJS.

Hi Bill

How obvious, but never got idea to use it :)
Propably READING could help, not microcrab style
next next next :)

Jarmo, oh1mrr


___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Wsjtx crash

[Bill Somerville via wsjt-devel]

On 17/10/2021 05:32, jarmo via wsjt-devel wrote:

Started to play with ref.spec, started measure, never got
stop measuring, or didn't find stop measuring button.
Wsjtx 2.5.0 self compiled, Fedora 34, latest hamlib 4.4 git version.
Now try start wsjtx, it starts for few seconds and then crash, get
message, if I translate it right from Fin to Eng "memory overflow"
(created core-file).
Where this core-file is created, can't find it?

Jarmo


Hi Jarmo,

the button to stop collecting reference spectrum data is the one 
labelled "Stop" on the main window.


73
Bill
G4WJS.



___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Grid Location info does not attach on outgoing messages.

[Bill Somerville via wsjt-devel]

On 17/10/2021 04:31, William Spencer via wsjt-devel wrote:


On all modes my grid locator info does not appear in the message 
lines.  I have checked the settings and the info is in the correct 
box.  This used to work for a long time, but I clicked the wrong box 
and now it doesn’t happen.  I clicked on a command that said ‘reset’ 
thinking it would just reset com with the radio, but it seemed more 
like a ‘reset to default’ action as all my settings were gone.  After 
reentering all the settings, all works except this.


I decided today that since I was on ver. 2.3 and ver. 2.5 is out that 
I would just uninstall 2.3 and install 2.5.  Apparently, some lib file 
must not have gotten uninstalled as when I opened 2.5 all my settings 
were there as well as the problem!


The hardware is a P400 and I’m not sure what the OS is, but I think it 
is Buster.


My question for the group is that if anyone knows what folder the 
settings data is in and is it easily editable?  If not, how can I be 
assured I have gotten all the old folders and data deleted before I 
start another clean install?


73, Bill KY4O


Bill,

check the callsign you have entered in "Settings->General", the symptoms 
you describe would occur if you have used a zero rather than an Oh as 
the last character of your callsign.


73
Bill
G4WJS.
___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel

Re: [wsjt-devel] Some minor issues.

[jarmo via wsjt-devel]

Wed, 29 Sep 2021 16:15:00 + (UTC)
Black Michael via wsjt-devel 
kirjoitti:

Hi Mike

Compiled latest git version 4.4 and no cahnge with TS-590SG.
We can run wsjtx (2.5.0) when radio set as TS-590SG, 9600 rest
as default. Audio goes out.
But again, if I set there /usr/local/bin/rigctld -m 2037 -r /dev/radio
-s 9600, wsjt setting radio set as Hamlib Net Rigctl, audio does not
go out, just tiny sound goes, but no matter how much turn output +
it stays kind of "muted"

Jarmo, OH1MRR
> Also...please the the git clone so you can get any patches I put
> in.The tar file only gets updated overnight.
> 
>  
> 
> On Wednesday, September 29, 2021, 11:03:57 AM CDT, jarmo via
> wsjt-devel  wrote: 
>  Wed, 29 Sep 2021 12:05:31 -0300
> Marco Calistri via wsjt-devel 
> kirjoitti:
> 
> > Il 29/09/21 11:07, Black Michael ha scritto:  
> > > Can you please try the latest hamlib and see if it's fixed now?  
> 
> Unfortunately not working. Compiled
> hamlib-4.4~git-dd1376be-20210929.tar.gz and same behaviour.
> If try to use NetHamlib rictl, audio does not go into radio.
> Cat control works ok.
> But if chosen Kenwood ts-590SG in WSJTX radio preferences, audio works
> ok...
> 
> Jarmo
> 
> 
> ___
> wsjt-devel mailing list
> wsjt-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/wsjt-devel
>   



___
wsjt-devel mailing list
wsjt-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/wsjt-devel