On Thu, 25 Aug 2022 02:20:15 +0000 (UTC)
"Sam Birnbaum via groups.io" <w2jdb@...> wrote:

Yes, I know they can respond on another offset, but can they guess which one is open on my end and not causing QRM to someone else that they can not hear or if its that busy, how can they find an open slot to begin with?

If you use that reasoning you will never make any contacts because very frequency is busy somewhere.

As I said, I developed this method through experience.     

As have I, and also from reading relevant documentation. 95% of the time calling off frequency works better. Sometimes I need to dance around a bit to find a good spot but that is no different than what you described. The only use case where it doesn't work better if is the CQing op is using a narrow filter contrary to best practice as discussed in another thread or is simply not answering off frequency callers for whatever reason. I reckon that the astrophysicist and his team has more smarts than most of the rest of us put together, and then some, and have thought this through thoroughly.

73

-Jim
NU0C

Hi Jim,

And why would it not be? That is one of the features of FT/JT. You can continue calling if you are off his frequency and you have a better chance of being heard and next worked.Yes you can continue calling and possibly causing QRM to others that you can not hear but they can hear you. As far as a better chance of being heard and worked, that didn't seem to have been affected by my methodology.
In a perfect world you might be able to find a frequency slot that is clear both ways. With busy bands it is mostly impossible. I don't call CQ when I can't find an open offset in both time slots. If I am TX on odd, the even slot better be open so that I can hear a possible responding station. Yes, I know they can respond on another offset, but can they guess which one is open on my end and not causing QRM to someone else that they can not hear or if its that busy, how can they find an open slot to begin with?  
 It seems to me that you are thinking in terms of SSB/CW/etc. No I am not. I am quite aware of how FT8/FT4 operate vs SSB/CW and what the so called "Rules" are.

As I said, I developed this method through experience.     
73, Sam W2JDB


-----Original Message-----
From: Jim Shorney <jimNU0C@...>
To: main@WSJTX.groups.io
Sent: Wed, Aug 24, 2022 9:54 pm
Subject: Re: [WSJTX] Using WSJTX Waaterfall #FT8


Sam,

On Thu, 25 Aug 2022 00:52:12 +0000 (UTC)
"Sam Birnbaum via groups.io" <w2jdb@...> wrote:

One of the features of WSJT-X is that if you are within 25Hz of his TX offset and they start working another station, your attempt is terminated immediately. That was implemented by the WSJT-X developers specifically in order to reduce QRM and that function is defeated by the very practice of holding TX on a different offset.

And why would it not be? That is one of the features of FT/JT. You can continue calling if you are off his frequency and you have a better chance of being heard and next worked.
 

When you are replying to a CQer, you are always transmitting on the opposite cycle and hopefully, that station found that offset by listening for multiple RX cycle (I listen for at least 115 seconds) before selecting that offset. That means that its clear to them on Odd and Even.  I would never select an offset that is being used in either time slot. If the band is that crowded, where I can't find an open slot, I don't call CQ, I revert to the search and pounce mode of operation.  

In a perfect world you might be able to find a frequency slot that is clear both ways. With busy bands it is mostly impossible. In the real world you will always be QRMing someone, somewhere. But it also is not necessary. It seems to me that you are thinking in terms of SSB/CW/etc. This is strictly timed multiplexed resource sharing. The rules are different.

Since you felt the need to qualify yourself, I am also an experienced operator. My first JT65 contact was in June 2014 (FT8 came around in 2017) and I've been along for the ride ever since.

73

-Jim
NU0C

Sam,

On Thu, 25 Aug 2022 00:52:12 +0000 (UTC)
"Sam Birnbaum via groups.io" <w2jdb@...> wrote:

One of the features of WSJT-X is that if you are within 25Hz of his TX offset and they start working another station, your attempt is terminated immediately. That was implemented by the WSJT-X developers specifically in order to reduce QRM and that function is defeated by the very practice of holding TX on a different offset.

And why would it not be? That is one of the features of FT/JT. You can continue calling if you are off his frequency and you have a better chance of being heard and next worked.
 

When you are replying to a CQer, you are always transmitting on the opposite cycle and hopefully, that station found that offset by listening for multiple RX cycle (I listen for at least 115 seconds) before selecting that offset. That means that its clear to them on Odd and Even.  I would never select an offset that is being used in either time slot. If the band is that crowded, where I can't find an open slot, I don't call CQ, I revert to the search and pounce mode of operation.  

In a perfect world you might be able to find a frequency slot that is clear both ways. With busy bands it is mostly impossible. In the real world you will always be QRMing someone, somewhere. But it also is not necessary. It seems to me that you are thinking in terms of SSB/CW/etc. This is strictly timed multiplexed resource sharing. The rules are different.

Since you felt the need to qualify yourself, I am also an experienced operator. My first JT65 contact was in June 2014 (FT8 came around in 2017) and I've been along for the ride ever since.

73

-Jim
NU0C

Hi Randy,
First of all, I am a very experienced operator as I have been using WSJT-X since FT8 first came out. I also wrote my own interface to WSJT-X when FT8 first came out.I developed this strategy based on the behavior of FT8 traffic and how I was or was not getting results. It is extremely infrequent that I have to shift my Tx offset when the original perceived SNR is better than -15 and under those conditions, better than 80% of my replies to a CQ are returned in the next RX cycle. One of the features of WSJT-X is that if you are within 25Hz of his TX offset and they start working another station, your attempt is terminated immediately. That was implemented by the WSJT-X developers specifically in order to reduce QRM and that function is defeated by the very practice of holding TX on a different offset. 
Back to my original statement about an open offset at the CQer station. That statement is still true. Mind you, if 2 or more stations arrive at precisely the same offset (frequency) with the exact same signal strength at the exact same time, that station will not be able to reply to anyone and will continue to send CQ (unless of course someone else on another offset also called that station). In that case, I would have moved my offset (after my second Tx with no reply) and then one of us would get the reply. I also watch the SNR and I may abort the attempt if I see the SNR value change dramatically while that the station is still sending CQ. If I don't see a continuation CQ calls, I abort the attempt after 2 more TX cycles.    
When you are replying to a CQer, you are always transmitting on the opposite cycle and hopefully, that station found that offset by listening for multiple RX cycle (I listen for at least 115 seconds) before selecting that offset. That means that its clear to them on Odd and Even.  I would never select an offset that is being used in either time slot. If the band is that crowded, where I can't find an open slot, I don't call CQ, I revert to the search and pounce mode of operation.  
73,  

Sam W2JDB

On Wed, Aug 24, 2022 at 04:33 PM, Sam Birnbaum wrote:

Well, based on my experience that has not been the case.What you may not be
aware of is the fact the TX offset you are holding may be causing QRM for
others that you do not hear.

But how can you know whether you are QRMing? If I can't know whether I'm QRMing on a frequency that I've been holding for a good while (and of course I can't disagree with you on that), I don't know how you can confidently state, "that has not been the case" for your method. Of course we have no guarantees of operating QRM-free, either on the receiving or the giving end. My strong hunch, though, is that you are not going to find experienced ops agreeing that your method is best.

So, for example, on a crowded band, I'm more often that not operating on a frequency whose slots opposite mine are being used by another station, and all you have to do is look at the waterfall to see that this is very common. A minority of "columns" are blank half the time when the band is crowded. So if you are indiscriminately jumping to the frequency of the station you're calling, the probability is fairly high that you are transmitting on top of somebody else. Sometimes both will be decoded by their targets and everyone is happy. Sometimes not. FT-world is certainly imperfect, and we have to make our best judgments about how to make the best of it. I hope that not many people will adopt your suggestion over the more conventional wisdom of holding your TX frequency as much as possible, and I imagine that a large majority would side against you on this. Thanks for interacting on the issue kindly!

Randy, WS4C

Hi Randy,
Well, based on my experience that has not been the case.What you may not be aware of is the fact the TX offset you are holding may be causing QRM for others that you do not hear.
73,

Sam W2JDB

On Wed, Aug 24, 2022 at 06:09 AM, Alessandro I2SVA wrote:

2.6.0 rc3 - Win10
Mode Q65:
Decode Fast does not persist. Close/reopen WSJT-X goes always to Decode Deep.

I cannot reproduce this problem.

As far as I can see, on program restart the selected decode level is properly restored in Q65, as in all other modes,

-- 73, Joe, K1JT

On Wed, Aug 24, 2022 at 02:03 PM, Sam Birnbaum wrote:

For those times that I choose to follow the search and pounce mode of
operation, I leave the "Hold TX Freq" checkbox unchecked. This allows WSJT-X
to move my Tx offset to the CQer offset. At this point, I only allow my
station to transmit just 2 times without a reply before I shift the Tx offset
25Hz up/down always towards the center of the bandpass. If after 3 more TX
cycles I do not receive a reply, I abandon the QSO attempt. This procedure
also allows WSJT-X to automatically terminate my attempt at the QSO if it
detects that the CQer is working another station. During this search and
pounce mode of operation I do not answer any station that is calling me.
That's just the way I choose to operate and it has served me well.

Others may already have replied to this, but I would humbly suggest that your procedure is not serving others well. The CQer's listening periods, on a busy band, have a probability of being occupied by other stations that I would guess with some confidence exceeds 50%. So by leaving "Hold TX" unchecked, you are running a strong likelihood of QRMing somebody else when you could be holding your own frequency with a much lower probability of QRM. The conventional wisdom of keeping "Hold TX" checked seems to me to be well founded. Just my two-cents' worth.

Randy, WS4C

On Wed, Aug 24, 2022 at 12:41 PM, Chuck Gelm wrote:

If you are calling CQ on 1000Hz and I am answering on 2000 Hz,
then your 250 HZ filter centered on 1000 Hz is degrading my signal.

Yes, but that's not the situation I'm talking about. I'm talking about trying to work a particular weak station and using my filter to increase my likelihood of decoding him. I'm not listening for other callers; just doing after one station on a known frequency. Known, that is, until for some reason he jumps to another spot!

Randy, WS4C

On Wed, Aug 24, 2022 at 12:34 PM, Chuck Gelm wrote:

Alas, no thanks, because the 'tail-ender' is likely still on the CQ'ers
frequency.

You lost me here. Why would his frequency matter, as long as he has his filter set wide enough to decode you on your previous frequency? If you decoded him when he was tail-ending (on top of the station you're working?), then it seems likely that you'll decode him again, regardless of what frequency he's on. I don't recall the original context anymore, but I think I was assuming that he's not on the CQ'ers frequency anyway. Probably not worth pursuing, so no need to reply as far as I'm concerned.

Randy, WS4C

If I remember correctly the sampling rate also needs to be 48K.
--
John P.
WA2FZW

Try this:

"If you go into System>Sound, then down to the Input section. Click the > symbol to the right of Microphone Audio.

On the next screen look for the Enhance Audio section and click on Advanced. Then disable the enhancements."

This is from
K7TMG
West Mountain Radio

It worked for me.

73
John
WV5JAW

Windows  audio should ideally be at 0dB (you can right-click the level in the sound manager to change it to dB).
Anything over 0dB is a digital multiplication which provides no benefit at all.
Analog adjustments to signal level are the only ones effective to increase signal level when Windows recording is at 0dB.
That means both RF Gain and output audio level from the rig are the two adjustments that do anything beneficial.If you have a sound card like a SignaLink or RigBlaster the analog adjustments on there work too.
Mike W9MDB

Hi,
My original response (Mon, Aug 22, 2022 8:14 pm) to this thread included this line:"The only guarantee that you can have for a clear offset at the intendent target is if you pick the same offset that they picked to transmit."
Following this logic and practicing what I preach, I have followed the following procedure starting with WSJT-X ver 1.8:If I am going to be calling CQ, I first listen for at least 115 seconds so that I can find a clear offset. After selecting that offset, I make sure the "Hold Tx Freq" checkbox is checked. I then proceed to call CQ. That allows me to periodically double click on a DX station (that I need/want) that pops up without losing or moving my Tx offset. For those times that I choose to follow the search and pounce mode of operation, I leave the "Hold TX Freq" checkbox unchecked. This allows WSJT-X to move my Tx offset to the CQer offset. At this point, I only allow my station to transmit just 2 times without a reply before I shift the Tx offset 25Hz up/down always towards the center of the bandpass. If after 3 more TX cycles I do not receive a reply, I abandon the QSO attempt. This procedure also allows WSJT-X to automatically terminate my attempt at the QSO if it detects that the CQer is working another station. During this search and pounce mode of operation I do not answer any station that is calling me.
That's just the way I choose to operate and it has served me well.
73,   
Sam W2JDB

I had the same issue. Enable the codec, but also check the level control in settings. Mine was so low that I didn't get enough audio into the transmitter to get the WSJT-X output.

Barry Priddy - K5VIP

On 8/23/22 18:46, Randy, WS4C wrote:

That's interesting.*How narrow is narrow?* Are we assuming digital or analog here? I've experimented with my 250Hz analog filter with no apparent degrading of decode reliability

If you are calling CQ on 1000Hz and I am answering on 2000 Hz,
then your 250 HZ filter centered on 1000 Hz is degrading my signal.
:-|
In regard to FT8, I regard a narrow filter is one that rejects signals anywhere between 200 to 3000 Hz.


Chuck

I want to apologize a bit in advance as I have age related digital issues.
In Windows 11 I see the USB codec for speakers. Under General/Audio it says Allow apps and Windows to use this device for audio. I got the USB codec for the speakers and Mic changed from Don't Allow to Allow. Enhanced Audio is off as well. I rebooted WSJT-x and there is no audio showing on the vertical bar graph WSJT-x. I changed the Allow for the USB codec back to Don't Allow and I still don't see any audio.

Any help here is appreciated.

Cheers,

Marv
W0PSY

On 8/23/22 17:52, Randy, WS4C wrote:

In that case, moving back to your original frequency--if it remains clear--would likely allow you to make the QSO with the tail-ender, assuming that he does not have his filter set tight.

Alas, no thanks, because the 'tail-ender' is likely still on the CQ'ers frequency.
:-|

73, Chuck

On 8/23/22 17:07, David Herring wrote:

Following on with Chuck’s message, I would recommend that after having worked the CQ-er on their frequency, moving back to the frequency you were on originally, or some other clear frequency, before continuing on with other QSOs.

+1

In support of answering a CQ-er on their frequency when they don’t hear you on yours, there is some logic to that. I suppose some folks may be using filters, and I would say that’s probably not the best choice, but another reason why they don’t respond when you call them from the frequency you are using is that there may be another station calling on that same frequency as you are currently on, maybe you can’t see them, but they may be loud at the other end. When faced with what new frequency to try responding on, there is some logic to calling on the CQ-er’s frequency because, one would hope, they chose a clear frequency to start calling CQ on.

As I mentioned: "often they answer in their next time slot and give me a 'R+NN'."
i.e. Not a R-24 or R-1, but a R+value indicating that they received sufficient SNR to be decoded.
Alas, perhaps the CQ'er is using 'Decode Fast' which does not re-scan the recorded audio file after subtracting first pass decoded stations.

In either case, moving back to another clear frequency after the QSO is done and before working anyone else seems proper.

+1

73, Chuck

On Wed, Aug 24, 2022 at 02:16 AM, Jim Brown wrote:

In the presence of a very strong signal, the solution is to increase the gain
between the receiver and the computer so that the green bar is as close as
possible to 80 dB without turning red (indicating digital clip). What this
does is put that very big signal at the very top of the dynamic range of the
A/D, so that the weaker signals we want to work are not below the cutoff of
the A/D.

Without knowing why, this is the practice I've developed. The explanation is helpful.

Another poster mentioned adjusting this gain by the Windows input level control. I have been making the adjustment with my rig's RF gain control, with my Windows input set at a level that allows me to run the radio's RF gain control in a range of about 45%-70% of full bore, depending on a variety of factors related to noise and signal levels. The RF gain is more efficient than the Windows control for frequent adjustments. Do I seem to be thinking straight here, or am I likely missing some advantage that I could get by using the Windows control instead? My limited audio training has taught me the rule of thumb to get needed gain as early in the signal chain as possible (while maintaining adequate headroom at that stage), for optimal S/N ratio, so that's the way I've been thinking, though I admit to cheating on the rule a bit by often lowering the RF gain to around 50% just to avoid fiddling with a software control.

Randy, WS4C