locked S/N reporting on different modes #modes #Icom


SA5NTK
 

Hello Everybody!

My first post here...
I am wondering if S/N (dB) reported by the program on different modes are comparable with each other.

Hypothetically, if my sending is received at a given station with S/N, e.g. -10 dB on WSPR, would a FT8 signal will have the same S/N with everything  else being equal, bar a few KHz difference in frequency?

Sorry if it was discussed already. I did search the forum, but couldn't find an answer.

73 SA5NTK /Notko


Bill Somerville
 

On 12/07/2021 10:18, SA5NTK wrote:
Hello Everybody!

My first post here...
I am wondering if S/N (dB) reported by the program on different modes are comparable with each other.

Hypothetically, if my sending is received at a given station with S/N, e.g. -10 dB on WSPR, would a FT8 signal will have the same S/N with everything  else being equal, bar a few KHz difference in frequency?

Sorry if it was discussed already. I did search the forum, but couldn't find an answer.

73 SA5NTK /Notko
Hi Notko,

the main variable in assessing SNR estimates is themeasured noise level, that is particularly challenging with modes that decode multiple signals from busy sub-bands. The noise estimation algorithms in WSJT-X reflect both the expected band occupancy and the evolution of noise estimating in the software. While SNR numbers are all intended to use the same signal power versus the noise power in a 2500 Hz bandwidth, the results are not going to be perfect. Other factors might include restricted receiver bandwidth, which should be avoided, many overlapping signals, and the fact that as the decoding sensitivity threshold is approached the error band of the SNR number approaches infinity. This last point is worth noting, often users make a big deal of a particularly low SNR number, whereas in fact it is most likely only reflecting this potentially large error band. Each mode enforces a clamp on the lowest SNR number reported, otherwise very large negative numbers that are theoretically impossible would be seen. The clamp value is chosen to be just below the theoretical maximum sensitivity for the mode.

73
Bill
G4WJS.


Bruce KX4AZ
 

On Mon, Jul 12, 2021 at 05:37 AM, Bill Somerville wrote:

.. The noise estimation algorithms in WSJT-X reflect
both the expected band occupancy and the evolution of noise estimatiin the software. ...

This is something I have always wondered about - does the noise estimation algorithm attempt to use an area of the audio spectrum where no signals were decoded?  But I guess a crowded band would make that nearly impossible.  In any event, it would be fascinating to hear more details about how the noise power is estimated to arrive at the S/N ratios.


Mike GM3PPE
 

I have been very active on 6 meters this summer and have noticed great variability in S/N reports sent and received. For example I worked a YV5 station twice in the same week. On the first occasion he sent -18 and received -05 from me, a difference of 13dB. Two days later he sent -15 but was +11 with me, a difference of 26dB. My setup was the same, and assuming his setup was unchanged, why the comparative difference?

This effect has been seen on numerous occasions, and by other stations in a similar rural location. Are more details on exactly how the S/N ratio is calculated available? I want to understand how the background noise power is estimated, without having to read the source code!

Mike GM3PPE


Bill Somerville
 

On 13/07/2021 01:59, Bruce KX4AZ wrote:
On Mon, Jul 12, 2021 at 05:37 AM, Bill Somerville wrote:

.. The noise estimation algorithms in WSJT-X reflect
both the expected band occupancy and the evolution of noise estimatiin the software. ...

This is something I have always wondered about - does the noise estimation algorithm attempt to use an area of the audio spectrum where no signals were decoded?  But I guess a crowded band would make that nearly impossible.  In any event, it would be fascinating to hear more details about how the noise power is estimated to arrive at the S/N ratios.

Hi Bruce,

I believe we use variations on two basic strategies:-

  1. measure the noise power in unused symbol tone bins and interpolate to a 2500 Hz bandwidth noise power,
  2. measure the noise power in a 2500 Hz bandwidth by using a baseline fitting algorithm.

(1) is relatively straightforward as we are already measuring signal power in all symbol tone bins as part of decoding, but is liable to inaccuracies when there are overlapping signals or other QRM.

(2) is a little more computationally complex but will perform well on a busy sub-band with many signals that may overlap, but requires that the receiver pass band is at least 2500 Hz wide.

You can do a simple experiment to determine which type is being used by reducing your receiver pass band while decoding a signal. Mode decoders that use (1) will give reasonably consistent SNR numbers so long as the wanted signal is within the restricted pass band, modes that use (2) will give greatly exaggerated SNR numbers as the receiver pass band is reduced below 2500 Hz.

73
Bill
G4WJS.


Bill Somerville
 

On 13/07/2021 09:00, Mike GM3PPE wrote:
I have been very active on 6 meters this summer and have noticed great variability in S/N reports sent and received. For example I worked a YV5 station twice in the same week. On the first occasion he sent -18 and received -05 from me, a difference of 13dB. Two days later he sent -15 but was +11 with me, a difference of 26dB. My setup was the same, and assuming his setup was unchanged, why the comparative difference?

This effect has been seen on numerous occasions, and by other stations in a similar rural location. Are more details on exactly how the S/N ratio is calculated available? I want to understand how the background noise power is estimated, without having to read the source code!

Mike GM3PPE
Hi Mike,

the noise component consists of local noise plus various other noises via one or more propagation paths from many directions, with Sporadic E propagation, which you are almost certainly receiving the YV5 station, the path loss is highly variable, both over the short term of a few seconds or minutes and from day to day. It is not reasonable to expect the noise and signal components to remain in the same ratio during different QSOs with the same station with such variable propagation dynamics.

It is interesting that you cite a YV station for your example, I have noticed that stations in Venezuela on 6m, among some others, often have some difficulty copying my signals when they are quite strong with me. My conclusion is that either they are running very high power or they suffer considerable levels of local QRN, my best guess is the latter. Perhaps there are YV stations on this list that might know about any specific issues with DX reception on 6m in their country?

73
Bill
G4WJS.


Jim Brown
 

On 7/13/2021 1:00 AM, Mike GM3PPE wrote:
This effect has been seen on numerous occasions, and by other stations in a similar rural location. Are more details on exactly how the S/N ratio is calculated available?
It is common for 6M propagation to be wildly variable over rather short periods of time. Last evening, I worked an HL2 station who I could decode for about 5 minutes, with the report (S/N) beginning around -22, peaking around -6, fading a minute or two later. On the next pass after our QSO completed, he was around -20, and on his next TX, he was gone.

73, Jim K9YC (CM87, S of San Francisco)


Tom V. Segalstad
 

Some comparisons for FT8 reports last Thursday evening (8th of July) on 50 MHz.

Here using 1 kW output power to a 7 element Yagi antenna, and a GaAsFET LNA RX preamplifier.

Listed below are the differences in dB S/N reports for some DX QSOs; positive numbers meaning I received stronger S/N reports than the DX station, and negative numbers meaning I received lower S/N reports, than given by me.

 

OD5KU        -18 dB

HI3MRV      -12 dB

HI8T             -11 dB

HI8RD                      -4 dB

4Z1TL                     +4 dB

4X1TI                     -11 dB

WP4G                     -18 dB

WP3R                       -9 dB

KP4EIT        -19 dB

CO7MS           0 dB

 

I can’t say what is the reason for these fairly large differences in S/N reports for DX QSOs.

But we see that the reports from Central America are all negative.

May be due to QRM from the many stations in USA?

But we in Europe also have many stations in Europe being copied at the same time.

So we wouldn’t think that the QRM level would be much different?

Reports for contacts within Europe were in most QSOs about equal, differing +/- 1 to 2 dB, and in rare cases up to +/- 5dB.

 

73 from Tom, LA4LN (and LC1V in contests)

 

 

Fra: Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 11.19
Til: main@WSJTX.groups.io
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

On 13/07/2021 09:00, Mike GM3PPE wrote:
> I have been very active on 6 meters this summer and have noticed great
> variability in S/N reports sent and received. For example I worked a
> YV5 station twice in the same week. On the first occasion he sent -18
> and received -05 from me, a difference of 13dB. Two days later he sent
> -15 but was +11 with me, a difference of 26dB. My setup was the same,
> and assuming his setup was unchanged, why the comparative difference?
>
> This effect has been seen on numerous occasions, and by other stations
> in a similar rural location. Are more details on exactly how the S/N
> ratio is calculated available? I want to understand how the background
> noise power is estimated, without having to read the source code!
>
> Mike GM3PPE

Hi Mike,

the noise component consists of local noise plus various other noises
via one or more propagation paths from many directions, with Sporadic E
propagation, which you are almost certainly receiving the YV5 station,
the path loss is highly variable, both over the short term of a few
seconds or minutes and from day to day. It is not reasonable to expect
the noise and signal components to remain in the same ratio during
different QSOs with the same station with such variable propagation
dynamics.

It is interesting that you cite a YV station for your example, I have
noticed that stations in Venezuela on 6m, among some others, often have
some difficulty copying my signals when they are quite strong with me.
My conclusion is that either they are running very high power or they
suffer considerable levels of local QRN, my best guess is the latter.
Perhaps there are YV stations on this list that might know about any
specific issues with DX reception on 6m in their country?

73
Bill
G4WJS.

 


--
Tom (LA4LN)


Bill Somerville
 

Hi Tom, and Jim,

I believe Mike in his OP was questioning the variations in SNR numbers between two stations on different occasions. For example I might have two QSOs with you and one day my SNR for you might be 20 dB different from yours to me, yet another day the difference might be less or more. One might expect the comparison of SNR numbers between two stations using unchanging setups to be consistent unless some form of one-way propagation enhancement were taking place, or more likely the noise environment at one end of the link has changed significantly. Hence my supposition that some stations may be located in regions where local noise is highly variable, for example heavy usage of solar panels with noisy switch-mode inverters, or air-conditioning units with poor noise suppression.

73
Bill
G4WJS.

On 13/07/2021 11:15, Tom V. Segalstad wrote:

Some comparisons for FT8 reports last Thursday evening (8th of July) on 50 MHz.

Here using 1 kW output power to a 7 element Yagi antenna, and a GaAsFET LNA RX preamplifier.

Listed below are the differences in dB S/N reports for some DX QSOs; positive numbers meaning I received stronger S/N reports than the DX station, and negative numbers meaning I received lower S/N reports, than given by me.

 

OD5KU        -18 dB

HI3MRV      -12 dB

HI8T             -11 dB

HI8RD                      -4 dB

4Z1TL                     +4 dB

4X1TI                     -11 dB

WP4G                     -18 dB

WP3R                       -9 dB

KP4EIT        -19 dB

CO7MS           0 dB

 

I can’t say what is the reason for these fairly large differences in S/N reports for DX QSOs.

But we see that the reports from Central America are all negative.

May be due to QRM from the many stations in USA?

But we in Europe also have many stations in Europe being copied at the same time.

So we wouldn’t think that the QRM level would be much different?

Reports for contacts within Europe were in most QSOs about equal, differing +/- 1 to 2 dB, and in rare cases up to +/- 5dB.

 

73 from Tom, LA4LN (and LC1V in contests)

 

 

Fra: Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 11.19
Til: main@WSJTX.groups.io
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

On 13/07/2021 09:00, Mike GM3PPE wrote:
> I have been very active on 6 meters this summer and have noticed great
> variability in S/N reports sent and received. For example I worked a
> YV5 station twice in the same week. On the first occasion he sent -18
> and received -05 from me, a difference of 13dB. Two days later he sent
> -15 but was +11 with me, a difference of 26dB. My setup was the same,
> and assuming his setup was unchanged, why the comparative difference?
>
> This effect has been seen on numerous occasions, and by other stations
> in a similar rural location. Are more details on exactly how the S/N
> ratio is calculated available? I want to understand how the background
> noise power is estimated, without having to read the source code!
>
> Mike GM3PPE

Hi Mike,

the noise component consists of local noise plus various other noises
via one or more propagation paths from many directions, with Sporadic E
propagation, which you are almost certainly receiving the YV5 station,
the path loss is highly variable, both over the short term of a few
seconds or minutes and from day to day. It is not reasonable to expect
the noise and signal components to remain in the same ratio during
different QSOs with the same station with such variable propagation
dynamics.

It is interesting that you cite a YV station for your example, I have
noticed that stations in Venezuela on 6m, among some others, often have
some difficulty copying my signals when they are quite strong with me.
My conclusion is that either they are running very high power or they
suffer considerable levels of local QRN, my best guess is the latter.
Perhaps there are YV stations on this list that might know about any
specific issues with DX reception on 6m in their country?

73
Bill
G4WJS.

 


--
Tom (LA4LN)



Andy Talbot
 

The most effective noise estimation algorithm I've encountered uses statistics.  The power in each FFT bin is reordered by sorting into ascending power, then the bin at the lower quartile is taken.   Eg for a 4096 FFT bins the power in the 1024th bin is taken after reordering.   Add a 'fiddle factor' of about 5dB and in most scenarios including those with strong signals you end up with a close approximation of the true noise level.

This works because it doesn't take any account of the strength of a signal, just it's presence in one of the higher, resorted, bins.   SO there can be plenty of signals present, and they will all sit at the top of the list, contributing very little to teh value derived from the lower quartile. 

However it does mean you have to have a quick sorting routine for large FFT sizes.  But there are ways round that.as you're only interested in ONE value, not the contents of all the bins.   So the lower quartile sort can be found in real time rather than after a complete sort.

Andy


Tom V. Segalstad
 

Thanks for the clarification, Bill.

 

Let’s look at repeated QSOs on 6 meters, separated some days in time, but assuming same operating techniques (at least here). The problem is that we don’t want to make duplicate contacts, so the few contacts I show here have been made by me, when using different call signs. Here are the differences in dB reporting, for a previous QSO with some of the stations listed before:

 

OD5KU    -3 dB (-18 dB in a QSO before)

HI8T       -15 dB (-11 dB in a QSO before)

KP4EIT  -13 dB (-19 dB in a QSO before)

CO7MS     -9dB (0 dB in a QSO before)

 

The tendency is still, that the other DX stations are giving a lower dB S/N report.

 

73 from Tom, LA4LN / LC1V

 

Fra: Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 12.25
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

Hi Tom, and Jim,

 

I believe Mike in his OP was questioning the variations in SNR numbers between two stations on different occasions. For example I might have two QSOs with you and one day my SNR for you might be 20 dB different from yours to me, yet another day the difference might be less or more. One might expect the comparison of SNR numbers between two stations using unchanging setups to be consistent unless some form of one-way propagation enhancement were taking place, or more likely the noise environment at one end of the link has changed significantly. Hence my supposition that some stations may be located in regions where local noise is highly variable, for example heavy usage of solar panels with noisy switch-mode inverters, or air-conditioning units with poor noise suppression.

 

73
Bill
G4WJS.

 

On 13/07/2021 11:15, Tom V. Segalstad wrote:

Some comparisons for FT8 reports last Thursday evening (8th of July) on 50 MHz.

Here using 1 kW output power to a 7 element Yagi antenna, and a GaAsFET LNA RX preamplifier.

Listed below are the differences in dB S/N reports for some DX QSOs; positive numbers meaning I received stronger S/N reports than the DX station, and negative numbers meaning I received lower S/N reports, than given by me.

OD5KU        -18 dB

HI3MRV      -12 dB

HI8T             -11 dB

HI8RD            -4 dB

4Z1TL            +4 dB

4X1TI           -11 dB

WP4G          -18 dB

WP3R            -9 dB

KP4EIT        -19 dB

CO7MS           0 dB

 I can’t say what is the reason for these fairly large differences in S/N reports for DX QSOs.

But we see that the reports from Central America are all negative.

May be due to QRM from the many stations in USA?

But we in Europe also have many stations in Europe being copied at the same time.

So we wouldn’t think that the QRM level would be much different?

Reports for contacts within Europe were in most QSOs about equal, differing +/- 1 to 2 dB, and in rare cases up to +/- 5dB.

 73 from Tom, LA4LN (and LC1V in contests)

Fra: Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 11.19
Til: main@WSJTX.groups.io
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

On 13/07/2021 09:00, Mike GM3PPE wrote:
> I have been very active on 6 meters this summer and have noticed great
> variability in S/N reports sent and received. For example I worked a
> YV5 station twice in the same week. On the first occasion he sent -18
> and received -05 from me, a difference of 13dB. Two days later he sent
> -15 but was +11 with me, a difference of 26dB. My setup was the same,
> and assuming his setup was unchanged, why the comparative difference?
>
> This effect has been seen on numerous occasions, and by other stations
> in a similar rural location. Are more details on exactly how the S/N
> ratio is calculated available? I want to understand how the background
> noise power is estimated, without having to read the source code!
>
> Mike GM3PPE

Hi Mike,

the noise component consists of local noise plus various other noises
via one or more propagation paths from many directions, with Sporadic E
propagation, which you are almost certainly receiving the YV5 station,
the path loss is highly variable, both over the short term of a few
seconds or minutes and from day to day. It is not reasonable to expect
the noise and signal components to remain in the same ratio during
different QSOs with the same station with such variable propagation
dynamics.

It is interesting that you cite a YV station for your example, I have
noticed that stations in Venezuela on 6m, among some others, often have
some difficulty copying my signals when they are quite strong with me.
My conclusion is that either they are running very high power or they
suffer considerable levels of local QRN, my best guess is the latter.
Perhaps there are YV stations on this list that might know about any
specific issues with DX reception on 6m in their country?

73
Bill
G4WJS.

 


--
Tom (LA4LN)

 

 


--
Tom (LA4LN)


Sam Birnbaum <w2jdb@...>
 

Hi

I used my alltext.exe program to analyze the the snr of every decoded station and calculate vsriance, min, max and mean. Not only does it show that there is a large variance in the same short time period for any particular station but that the same station will have a different set of values the next day.

The program is free and available on programsbyw2jdb on groups.io

Please note it is a windows based executable.

73,

Sam W2JDB


On Tuesday, July 13, 2021, 07:18:18 AM EDT, Tom V. Segalstad <la4ln@...> wrote:


Thanks for the clarification, Bill.

 

Let’s look at repeated QSOs on 6 meters, separated some days in time, but assuming same operating techniques (at least here). The problem is that we don’t want to make duplicate contacts, so the few contacts I show here have been made by me, when using different call signs. Here are the differences in dB reporting, for a previous QSO with some of the stations listed before:

 

OD5KU    -3 dB (-18 dB in a QSO before)

HI8T       -15 dB (-11 dB in a QSO before)

KP4EIT  -13 dB (-19 dB in a QSO before)

CO7MS     -9dB (0 dB in a QSO before)

 

The tendency is still, that the other DX stations are giving a lower dB S/N report.

 

73 fromTom, LA4LN / LC1V

 

Fra: Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 12.25
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

Hi Tom, and Jim,

 

I believe Mike in his OP was questioning the variations in SNR numbers between two stations on different occasions. For example I might have two QSOs with you and one day my SNR for you might be 20 dB different from yours to me, yet another day the difference might be less or more. One might expect the comparison of SNR numbers between two stations using unchanging setups to be consistent unless some form of one-way propagation enhancement were taking place, or more likely the noise environment at one end of the link has changed significantly. Hence my supposition that some stations may be located in regions where local noise is highly variable, for example heavy usage of solar panels with noisy switch-mode inverters, or air-conditioning units with poor noise suppression.

 

73
Bill
G4WJS.

 

On 13/07/2021 11:15, Tom V. Segalstad wrote:

Some comparisons for FT8 reports last Thursday evening (8th of July) on 50 MHz.

Here using 1 kW output power to a 7 element Yagi antenna, and a GaAsFET LNA RX preamplifier.

Listed below are thedifferences in dB S/N reports for some DX QSOs; positive numbers meaning I received stronger S/N reports than the DX station, and negative numbers meaning I received lower S/N reports, than given by me.

OD5KU        -18 dB

HI3MRV      -12 dB

HI8T             -11 dB

HI8RD            -4 dB

4Z1TL            +4 dB

4X1TI           -11 dB

WP4G          -18 dB

WP3R            -9 dB

KP4EIT        -19 dB

CO7MS           0 dB

 I can’t say what is the reason for these fairly large differences in S/N reports for DX QSOs.

But we see that the reports from Central America are all negative.

May be due to QRM from the many stations in USA?

But we in Europe also have many stations in Europe being copied at the same time.

So we wouldn’t think that the QRM level would be much different?

Reports for contacts within Europe were in most QSOs about equal, differing +/- 1 to 2 dB, and in rare cases up to +/- 5dB.

 73 fromTom, LA4LN (and LC1V in contests)

Fra:Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 11.19
Til: main@WSJTX.groups.io
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

On 13/07/2021 09:00, Mike GM3PPE wrote:
> I have been very active on 6 meters this summer and have noticed great
> variability in S/N reports sent and received. For example I worked a
> YV5 station twice in the same week. On the first occasion he sent -18
> and received -05 from me, a difference of 13dB. Two days later he sent
> -15 but was +11 with me, a difference of 26dB. My setup was the same,
> and assuming his setup was unchanged, why the comparative difference?
>
> This effect has been seen on numerous occasions, and by other stations
> in a similar rural location. Are more details on exactly how the S/N
> ratio is calculated available? I want to understand how the background
> noise power is estimated, without having to read the source code!
>
> Mike GM3PPE

Hi Mike,

the noise component consists of local noise plus various other noises
via one or more propagation paths from many directions, with Sporadic E
propagation, which you are almost certainly receiving the YV5 station,
the path loss is highly variable, both over the short term of a few
seconds or minutes and from day to day. It is not reasonable to expect
the noise and signal components to remain in the same ratio during
different QSOs with the same station with such variable propagation
dynamics.

It is interesting that you cite a YV station for your example, I have
noticed that stations in Venezuela on 6m, among some others, often have
some difficulty copying my signals when they are quite strong with me.
My conclusion is that either they are running very high power or they
suffer considerable levels of local QRN, my best guess is the latter.
Perhaps there are YV stations on this list that might know about any
specific issues with DX reception on 6m in their country?

73
Bill
G4WJS.

 


--
Tom (LA4LN)

 

 


--
Tom (LA4LN)




William Smith
 

Indeed, for instance, here's one chosen at random:

210713_110900     7.074 Rx FT8     -7  0.1 1190 CQ KB4LHP EL98
210713_110930     7.074 Rx FT8    -12  0.1 1189 CQ KB4LHP EL98

So within 30 seconds he dropped 5dB.  Had I been 30 seconds earlier or later, I could have reported him as either of those two numbers, and he would have reported me as plus or minus some other amount, so tomorrow, being off by 10dB wouldn't surprise me at all.

I have to admit I don't pay a lot of attention to the SNR other than to know that I can work -10 or stronger without any problems, -15 to -10 is worth a try, and -20 and under isn't worth trying.  But then I'm only putting 100W into a 40M halfwave dipole barely above the roofline of my house, so legal limit into a 'real' antenna would make those ' under -24' stations workable.  But resources (including time) aren't infinite, it's just a hobby, and there are plenty of stations I haven't worked.

Here's another example of a QSO:
210707_221430     7.074 Rx FT8     -4  0.1  434 N1JBJ VA3EBM +06
210707_221456     7.074 Tx FT8      0  0.0 2350 VA3EBM N1JBJ R-04
210707_221500     7.074 Rx FT8      1  0.1  434 N1JBJ VA3EBM +06
210707_221515     7.074 Tx FT8      0  0.0 2350 VA3EBM N1JBJ R+01

So he heard me at +6, I responded with -4, he obviously didn't hear me, I responded with +1 (and he appears to have stopped hearing me at all).

73, Willie N1JBJ


On Jul 13, 2021, at 8:26 AM, Sam Birnbaum via groups.io <w2jdb@...> wrote:

Hi

I used my alltext.exe program to analyze the the snr of every decoded station and calculate vsriance, min, max and mean. Not only does it show that there is a large variance in the same short time period for any particular station but that the same station will have a different set of values the next day.

The program is free and available on programsbyw2jdb on groups.io

Please note it is a windows based executable. 

73,

Sam W2JDB


On Tuesday, July 13, 2021, 07:18:18 AM EDT, Tom V. Segalstad <la4ln@...> wrote: 


Thanks for the clarification, Bill.

 

Let’s look at repeated QSOs on 6 meters, separated some days in time, but assuming same operating techniques (at least here). The problem is that we don’t want to make duplicate contacts, so the few contacts I show here have been made by me, when using different call signs. Here are the differences in dB reporting, for a previous QSO with some of the stations listed before:

 

OD5KU    -3 dB (-18 dB in a QSO before)

HI8T       -15 dB (-11 dB in a QSO before)

KP4EIT  -13 dB (-19 dB in a QSO before)

CO7MS     -9dB (0 dB in a QSO before)

 

The tendency is still, that the other DX stations are giving a lower dB S/N report.

 

73 fromTom, LA4LN / LC1V

 

Fra: Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 12.25
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

Hi Tom, and Jim,

 

I believe Mike in his OP was questioning the variations in SNR numbers between two stations on different occasions. For example I might have two QSOs with you and one day my SNR for you might be 20 dB different from yours to me, yet another day the difference might be less or more. One might expect the comparison of SNR numbers between two stations using unchanging setups to be consistent unless some form of one-way propagation enhancement were taking place, or more likely the noise environment at one end of the link has changed significantly. Hence my supposition that some stations may be located in regions where local noise is highly variable, for example heavy usage of solar panels with noisy switch-mode inverters, or air-conditioning units with poor noise suppression.

 

73
Bill
G4WJS.

 

On 13/07/2021 11:15, Tom V. Segalstad wrote:

Some comparisons for FT8 reports last Thursday evening (8th of July) on 50 MHz.

Here using 1 kW output power to a 7 element Yagi antenna, and a GaAsFET LNA RX preamplifier.

Listed below are thedifferences in dB S/N reports for some DX QSOs; positive numbers meaning I received stronger S/N reports than the DX station, and negative numbers meaning I received lower S/N reports, than given by me.

OD5KU        -18 dB

HI3MRV      -12 dB

HI8T             -11 dB

HI8RD            -4 dB

4Z1TL            +4 dB

4X1TI           -11 dB

WP4G          -18 dB

WP3R            -9 dB

KP4EIT        -19 dB

CO7MS           0 dB

 I can’t say what is the reason for these fairly large differences in S/N reports for DX QSOs.

But we see that the reports from Central America are all negative.

May be due to QRM from the many stations in USA?

But we in Europe also have many stations in Europe being copied at the same time.

So we wouldn’t think that the QRM level would be much different?

Reports for contacts within Europe were in most QSOs about equal, differing +/- 1 to 2 dB, and in rare cases up to +/- 5dB.

 73 fromTom, LA4LN (and LC1V in contests)

Fra:Bill Somerville
Sendt: tirsdag 13. juli 2021 kl. 11.19
Til: main@WSJTX.groups.io
Emne: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes

 

On 13/07/2021 09:00, Mike GM3PPE wrote:
> I have been very active on 6 meters this summer and have noticed great 
> variability in S/N reports sent and received. For example I worked a 
> YV5 station twice in the same week. On the first occasion he sent -18 
> and received -05 from me, a difference of 13dB. Two days later he sent 
> -15 but was +11 with me, a difference of 26dB. My setup was the same, 
> and assuming his setup was unchanged, why the comparative difference?
>
> This effect has been seen on numerous occasions, and by other stations 
> in a similar rural location. Are more details on exactly how the S/N 
> ratio is calculated available? I want to understand how the background 
> noise power is estimated, without having to read the source code!
>
> Mike GM3PPE 

Hi Mike,

the noise component consists of local noise plus various other noises 
via one or more propagation paths from many directions, with Sporadic E 
propagation, which you are almost certainly receiving the YV5 station, 
the path loss is highly variable, both over the short term of a few 
seconds or minutes and from day to day. It is not reasonable to expect 
the noise and signal components to remain in the same ratio during 
different QSOs with the same station with such variable propagation 
dynamics.

It is interesting that you cite a YV station for your example, I have 
noticed that stations in Venezuela on 6m, among some others, often have 
some difficulty copying my signals when they are quite strong with me. 
My conclusion is that either they are running very high power or they 
suffer considerable levels of local QRN, my best guess is the latter. 
Perhaps there are YV stations on this list that might know about any 
specific issues with DX reception on 6m in their country?

73
Bill
G4WJS.

 


-- 
Tom (LA4LN) 

 
 

-- 
Tom (LA4LN)








Larry Banks
 

Hi Mike,
 
This is the “magic” of six meters!

73 -- Larry -- W1DYJ

 
Sent: Tuesday, July 13, 2021 4:00
Subject: Re: [WSJTX] S/N reporting on different modes #modes #ic-7600 #modes
 
I have been very active on 6 meters this summer and have noticed great variability in S/N reports sent and received. For example I worked a YV5 station twice in the same week. On the first occasion he sent -18 and received -05 from me, a difference of 13dB. Two days later he sent -15 but was +11 with me, a difference of 26dB. My setup was the same, and assuming his setup was unchanged, why the comparative difference?

This effect has been seen on numerous occasions, and by other stations in a similar rural location. Are more details on exactly how the S/N ratio is calculated available? I want to understand how the background noise power is estimated, without having to read the source code!

Mike GM3PPE






Les Listwa
 

I agree with W1DYJ, its the fun and the magic  of 6M  and  specificaly  multihop sporadic E.   Another fun experiement on 6M I enjoyed trying was using a 2nd remote reciever  60 miles from the home station using the same antenna ( Delta Loop) and except for local ground wave stations, it received completely differnt DX stations and I watched some fade in one receiver and increase in the other vary rapidly,  which just shows the varaibility of Sporadic E propagation.    

On SSB, the changes were less pronounced as a change of an S unit or more is hardly noticable.  On FT-8 where we are working so close to noise level limits,  the equivelent  change of an S unit or two is many dbs and when listening to FT-8 signals at  close to noise leve limits and below,   a small 3 to  6 db fluctation which might hardly be noticed on a strong SSB signal, can make or break a contact.

6M + Sporadic E + FT8 =  Fun

73
Les
W2LPL

73
Les
W2LPL


Tony Collett
 

From Bill
"some stations may be located in regions where local noise is highly variable, for example heavy usage of solar panels with noisy switch-mode inverters,"

Being in the (un)fortunate position living next door to one of these installations I can confirm this is true!
My noise floor can vary by something like 15dB depending if Sun is on the panels or not. Even cloud cover can change that level within seconds.
So I might be copying a 6M FT8 signal at +something one period that almost disappears in my noise the next.

I can guarantee that beaming towards most of Europe I know I transmit far better than I can hear and Scandinavia seems to be blessed with extremely low levels of noise so the difference working LA/OH/SM is even more marked.
Beam to USA and situation is often reversed and again as Bill commented YV stations on all bands 80M up to 6M seem to suffer with poor reception.

Not sure I feel the need to statistically analyse it though as none of us know what power or receive conditions are being used (I vary anything between 20W and 400W depending on need!).

73
Tony G4NBS


Hank Pfizenmayer
 

Look at Bill N6MW paper on FT8  and SNR

 https://n6mw.jimdofree.com/

Hank K7HP 


Jim Brown
 

On 7/13/2021 2:19 AM, Bill Somerville wrote:
I have noticed that stations in Venezuela on 6m, among some others, often have some difficulty copying my signals when they are quite strong with me. My conclusion is that either they are running very high power or they suffer considerable levels of local QRN, my best guess is the latter.
I run 1kW on 6M, worst case noise level is S3 on a well calibrated S-meter, best case is S1. At least half of the signal reports I receive are 10 dB or more worse than those I give, which leads me to believe that IF the other station is using WSJT-X, he's got serious RX noise.

Because FTDX has the edge over WSJT-X on the weakest signals, I've been using FTDX for FT8 this season. I find that their noise estimation algorithm is really poor -- I rarely see it give reports over 0 dB, even on the very strongest locals that WSJT-X would give +35. PSKReporter will identify which software is feeding them each spot.

73, Jim K9YC