HFU HF Underground

Loggings => Utility => Topic started by: Josh on August 05, 2018, 2347 UTC

Title: 12786 FAX
Post by: Josh on August 05, 2018, 2347 UTC
12786 FAX 2242Z 05AUG18
NMC Pt Reyes CA.

https://nofile.io/f/BVgcbedow2t/20180805_224210_fax.bmp
Title: Re: 12786 FAX
Post by: fpeconsultant on August 06, 2018, 0016 UTC
Very nice.
Title: Re: 12786 FAX
Post by: Josh on August 06, 2018, 1749 UTC
Thanks, was one of the cleaner sigs for faxing heard here.
Title: Re: 12786 FAX
Post by: Josh on August 08, 2018, 1902 UTC
Another snatched from the aether;
https://nofile.io/f/QMjzUshMUd9/12788+1836Z++08AUG18.bmp
Title: Re: 12786 FAX
Post by: Josh on August 10, 2018, 2145 UTC
And another from yesterday.
https://nofile.io/f/ZFJ5m1wosvF/12788+fax.bmp
Title: Re: 12786 FAX
Post by: Josh on August 14, 2018, 1625 UTC
12786 J3C 1600Z 14AUG18
NMC Pt Reyes Ca, 4kw.

https://nofile.io/f/KmHZvABUuPS/12784+1610Z+14AUG18.bmp

https://nofile.io/f/SzwG8YKaMmJ/faxy.bmp

bonus footage from yesterday;
https://nofile.io/f/2HRDeCxVe5k/20180813_183815_fax.bmp
Title: Re: 12786 FAX
Post by: fpeconsultant on August 14, 2018, 1954 UTC
You're certainly on a roll.
Congrats
Title: Re: 12786 FAX
Post by: Josh on August 15, 2018, 0426 UTC
I'll do some more as I try to figure out wich decoder is best or most reliable and then get bored. One or more of the decoders here seems to have an issue where it syncs fine at start of fax and then somewhere along the way loses sync and starts faxing somewhere in the middle of the page. Trying to determine if this is a system timing issue external to the decoder, an artifact of how the decoder applies system time, or the signal itself is messed up. It could also be an artifact generated by the sdr or sdr app, or something that arises whenm the system becomes loaded with other cpu intensive tasks, when I log in here the pc is usually running at least one decoder and sdr app and opening webpages can briefly max out the cpus because of the security checks applied to each web page.

:D
Title: Re: 12786 FAX
Post by: fpeconsultant on August 15, 2018, 1408 UTC
The US based faxes were having difficulties not long ago - Chris was the first to post about it I think - right in the middle of a fax, it would shift slightly to one side and continue normally from there - the result being a fax with an offset somewhere along the line.  That could be happening again.

Here's the thread:
https://www.hfunderground.com/board/index.php/topic,34676.0.html (https://www.hfunderground.com/board/index.php/topic,34676.0.html)
Title: Re: 12786 FAX
Post by: Josh on August 15, 2018, 1814 UTC
Yep exactly what's been seen here. Also they've been having dropouts in the tx, wich simply adds noise to the image but doesn't hinder sync. I should start taking note of wich are the offenders and send them an email, but I can't be the only one noting this as multiple fishing and other vessels use these and have to see what I see and I figure they'd contact USCG and tell them about the service quality long before anyone else does.
Title: Re: 12786 FAX
Post by: fpeconsultant on August 15, 2018, 1832 UTC
Yes I agree with that.  I'll try to grab some this weekend - CA, LA, MA are usually strong here and HI at times as well.  I use an (old) Universal M8000 so screen shots are out but I will report back.
FPE
Title: Re: 12786 FAX
Post by: Josh on August 16, 2018, 0645 UTC
The 8000 was the bomb back in the day, my friend with the wall of jrc gear has one that even does the government modes denied to us commoners as he's a "station engineer" and conned universal or whoever sold it to him with some gov letterhead saying he was clean unlike the rest of us. I had a 600 and it made lovely faxes, wish I still had it - I've a fetish for old embassy/spy gear, but yes hard to go back to something that can't copy/paste. The only thing more revolutionary to hf radio than software decoders are sdr rigs far as I'm concerned.
Title: Re: 12786 FAX
Post by: ChrisSmolinski on August 16, 2018, 2114 UTC
Are the USCG faxes messed up again?

Josh - have you considered using imgur? You can then display your received fax images directly in a posting here.

Title: Re: 12786 FAX
Post by: Josh on August 17, 2018, 1947 UTC
I'll have to look at imgur and see. Here's sone great faxes from 12788 NMG New Orleans;
https://nofile.io/f/xeFEujBOGZa/perfect+fax.bmp

https://nofile.io/f/nx6snFDt7pr/broke+fax.bmp

And some decent ones from yesterday;
https://nofile.io/f/zXWorkOdBfa/bnzxcvmasdfj+qwer.bmp

https://nofile.io/f/nYhC6oYj3A9/dfjikl.bmp

I suspect the decoder has a issue perhaps as this is what results from checking autosave;
https://nofile.io/f/2HaDaZv3A1U/20180817_193843_fax.bmp

This one just came in perfectly, I was suspecting that by minimising and maximising the decoder app had somethiung to do with the issue so I reduced the decoder several times to task bar yet the image came out great;
https://nofile.io/f/yNhsamfvjbQ/nice+fax.bmp
 
Title: Re: 12786 FAX
Post by: Josh on August 17, 2018, 2016 UTC
This one turned out nice too;
https://nofile.io/f/qjYQm2Y0jd8/50+shades+of+greyscale.bmp

However it reveals a few issues. One is the slight timing issue as the image slants from top to bottom, usually this is not anywhere near as apparent as in this image. Another issue is the decoder didn't clear screen when it was given the start/sync signal so it applied the new image to the old. The slant could come from tx, rx, decoder, or sound card, while the reset signal going missing could be either the decoder's broke or it never got one. All this trouble for something I'll never really use other than for a few minutes diversion, isn't America great!?!
Title: Re: 12786 FAX
Post by: Josh on August 18, 2018, 1930 UTC
A few clean copies today from 12789.9;
https://nofile.io/f/VnCzRTvD2Wx/good.bmp

https://nofile.io/f/zJKP9ag2UEd/good+one+also.bmp

And then the tx just dropped out before the fax was finished;
https://nofile.io/f/aWbxGB5xdhz/dropout+at+1918Z+18AUG18.bmp

The first fax started out well then I noted it needed right justified so I used the decoder's phasing control and after saving the whole fax I noted that the fax was broken, possibly when I adjusted the phase of the image.



Title: Re: 12786 FAX
Post by: fpeconsultant on August 18, 2018, 2048 UTC
I can report faxes normal as follows:
New Orleans 2000z 12788khz
Boston 2015z 12748.1khz
Pt. Reyes 2030z 12784.1khz

Honolulu not heard at all on 16133.1khz 2030z.

All 8/18

FPE
Title: Re: 12786 FAX
Post by: Josh on August 18, 2018, 2101 UTC
I'm starting to think the decoder or the pc has timing issues, perhaps the decoder doesn't like hpet rather than rtc or similar.

Disabled hpet in bios and os and the issue remains, with multiple decoders.
I wonder if this is related to cpu multiplexing where the individual cpu cores are slowed or sped up according to load? Argh.
Title: Re: 12786 FAX
Post by: ChrisSmolinski on August 19, 2018, 1230 UTC
Decoding issues are typically due to sound card problems, either dropping sampled data (if you get skips/shifts in the fax image) or large sampling rate errors (lots of slant). CPU cores and other things like that aren't a factor. Once the sound has been digitized, nothing needs to be done real time.
Title: Re: 12786 FAX
Post by: Josh on August 19, 2018, 1932 UTC
Disabled hpet and set system to prioritise foreground rather than background apps and still messed up faxes.

First whole copy of the day;
https://nofile.io/f/2OEhSVKzq9B/1.bmp
Looks fine.

Second faulted about in the middle of the fax as you can see, then the tx dropped completely;
https://nofile.io/f/DjwJxyVTZzB/2.bmp

Mebbe it's the tx site having issues.

Title: Re: 12786 FAX
Post by: Josh on August 19, 2018, 1951 UTC
Letting it run on NMF Boston 12750 as New Orleans is ota - the tx that had issues, so far perfect copy barring fade.
Title: Re: 12786 FAX
Post by: ChrisSmolinski on August 20, 2018, 1354 UTC
Disabled hpet and set system to prioritise foreground rather than background apps and still messed up faxes.

First whole copy of the day;
https://nofile.io/f/2OEhSVKzq9B/1.bmp
Looks fine.

Second faulted about in the middle of the fax as you can see, then the tx dropped completely;
https://nofile.io/f/DjwJxyVTZzB/2.bmp

Mebbe it's the tx site having issues.

I can't see the images. Maybe try imgur and put the image URL in tags in your message? It's really painless  :)
Title: Re: 12786 FAX
Post by: Josh on August 20, 2018, 1535 UTC
But I have to make an acct to use imgur!
Title: Re: 12786 FAX
Post by: Josh on August 25, 2018, 1945 UTC
A fax battle! See if you can detect where the fm fax demodulator switched from one cochannel fax to the other in the included image!

https://nofile.io/f/rsKrBT02Ukc/fax+battle.bmp

The battle still rages, dunno if this is due to two separate tx sites being received, two feeds getting smeared together into one, or wtf is going on here.

Title: Re: 12786 FAX
Post by: Josh on September 05, 2018, 1803 UTC
Am trying some fax reception with the fractional output resampler locked in SDRuno. Perhaps this will reduce untoward happenstances. First fax was all over the place, after activating resampler lock it's been stable. Should also consider the tcxo takes some time to stabilise as each app controlling the sdr takes control of the sdr, I often switch between HDSDR and SDRuno and each has a different mechanism to handle correcting the master oscillator, both apps provide for automagically setting the reference oscillator to a known standard over the air, I try for the highest wwv signal I can get at the time.

Second fax progressing and it's aligned with the first one, this may be the solution but means fax may only be best received with SDRuno as HDSDR offers no resampler lock.
Title: Re: 12786 FAX
Post by: Josh on September 06, 2018, 1843 UTC
On the fractional reticulated sampler platter, don't bother.
Title: Re: 12786 FAX
Post by: Josh on January 24, 2019, 2347 UTC
Was trying to see if the haphazard fax reception that has happened to me and others bored enough to decode faxes was due to the sdr or to the pc or what, and used the Icom Pro2 to receive the fax, ruling out the sdr and/or usb connection as only audio was fed to the pc via a transformer isolated cable. I copied a few faxes, enough to feel confident that the sdr/usb connection may have been part of the synch issues, and became certain another aspect was revealed.

See this link for a pic of the results;
http://s000.tinyupload.com/index.php?file_id=32389565419432764128

In the image we see the results from K500 and WCODE both decoding the same input. In the middle of the K500 fax session you see where I changed mode from fm greyscale to fm black and white and much cleaner results are noted. The only time you want greyscale demod is when you are decoding  press fax with black and white pics, or decoding a satellite weather image. The main problem I was trying to resolve is displayed where the fm black and white takes over from the fm greyscale, K500 does not switch modes gracefully, and as you can see synch was lost and restarted elsewhere on the screen.

What this tells us is do not change anything while a desired fax is being decoded or you might cause the decoder to lose synch.

The very weak fax was copied on 17MHz band and was emitted from Pt Reyes California.