Weather fax on netbook + portable sony radio

[Pye_End]

Just tried to get Cullercoats on a NASA set and failed.

However, also tried Hamburg on 4583 (RTTY) and got lots of text, so will be worth having a go at this one as well. Worked well with Mscan Meteo Lite.

 

[timbartlett]

You might like this:
http://www.nws.noaa.gov/om/marine/rfax.pdf

Also, I have found that SSB comes through better if you tune the radio a couple of kHz lower than the nominal frequency of the station.

And for some reason, I get much better results on a very old (Windows 98!) IBM laptop than on any of my more recent machines.

 

[grumpy_o_g]

Blimey, good post Ed. I'll probably confuse more than clarify but I'll try and explain SSB very simply (the concept's dead simple but then lot's of people came along and made it much more complicated).

The carrier frequency is 4610. When you modulate that to put the information on top of the carrier you change that frequency. Actually it doesn't matter if it's frequency modulated (FM) or amplitude modulated (AM) the frequency still gets changed. If you change the peak voltage of the signal whilst keeping the number of cycles per second the same you actually change the instantaneous frequency as the only way to get to the higher peak is to increase the steepness of the slope up to the peak of that cycle. So, although the actual Hz stays the same on AM, the rate of change of voltage against time varies, just like a frequency modulated signal.

Until SSB a signal would vary equally either side of the carrier frequency. So 4610KHz would actually vary between 4607KHz and 4613KHz. Then some bright spark said "why do we need to send both sides of the modulated signal? Let's just send one side." The bit from 4607KHz to 4610KHz is one Side Band and the bit from 4610KHz to 4613KHz is the other. What that means is that, with SSB, the average frequency of each side band (USB or LSB) is offset slightly from the original carrier frequency, hence you need to tune slightly above or below to receive the appropriate side band.

In practice the carrier frequency is actually provided by your receiver - that frequency is compared to the signal received by the receiver and the difference is what it uses. As long as the receiver knows it's SSB it will know that it's looking for the difference above or below the carrier frequency as opposed to either side of it.

I'm hesitating to post a link to this thread because some of it's a little over-simplified (to be kind) but the explanation by KM3K about halfway down the first page is pretty much spot on I reckon.

Somebody will no doubt be along to tell me it's all changed since I learnt radio theory ..

 

[iainhu]

Somebody will no doubt be along to tell me it's all changed since I learnt radio theory ..

From what I've read-up on SSB, this seems a perfect explanation. Thanks. Makes complete sense

 

[iainhu]

Nothing but static

Attached is all I heard at 110UTC on 2616, 4608, 8038 (all Northwood) and 7878 (Germany). Taken outside so no building, only some trees and a few birds (which can at times be heard in the background). Maybe it’s just a duff set?

Tried a SW sweep (SSB off) and it could find talkSport on 1053 and 1089 so there is SW reception, but on a scan from 2000KHz up, the ‘first stop’ was at 9390, so going straight past the first three Northwood frequencies. This was an attempt to see if the radio could tune to anything on SSB. A similar scan with SSB-USB on gives a ‘first stop’ at 9940 although there’s no discernible signal on that frquency, just periodical crackling.

The nearest I’ve got to any SSB reception was very feint/poor Shannon Volmet on 5503/5 and RAF Volnet on 5450 I think, so maybe I need to return it and request another set.

 

[Pye_End]

The 4610 might be decodable with a bit of tuning. The 8040 sounds a good signal which really ought to be decodable. I'll see if I can put my netbook on the pc and see! So if you put these into SeaTTY in USB mode and ready to recieve fax rather than RTTY what does the graph look like? Maybe the program is not picking up the signal?

 

[iainhu]

Thanks for that 'analysis' At least it gives me some encourgaement that I'm getting near. I was expecting something more 'data like', but if this is sounding good I'll go and try some more.

I did put SeaTTY into manual fax receive and just got noise...nothing remotely like a chart and no 'twin peak' on the spctrum...but I will now go and try further armed with your reassurance this is sounding 'about right'. Many thanks for your help

 

[Conachair]

Attached is all I heard at 110UTC on 2616, 4608, 8038 (all Northwood) and 7878 (Germany). Taken outside so no building, only some trees and a few birds (which can at times be heard in the background). Maybe it’s just a duff set?

Tried a SW sweep (SSB off) and it could find talkSport on 1053 and 1089 so there is SW reception, but on a scan from 2000KHz up, the ‘first stop’ was at 9390, so going straight past the first three Northwood frequencies. This was an attempt to see if the radio could tune to anything on SSB. A similar scan with SSB-USB on gives a ‘first stop’ at 9940 although there’s no discernible signal on that frquency, just periodical crackling.

The nearest I’ve got to any SSB reception was very feint/poor Shannon Volmet on 5503/5 and RAF Volnet on 5450 I think, so maybe I need to return it and request another set.

That sounds like weatherfax, certainly on 7878. But not quite "right". it sounds like it's tuned a bit "high" This is a perfect signal http://www.youtube.com/watch?v=PwgUXKXbFQk

I use jvcomm with a degen 1103 reciever, never bothered too much with the exact frequncy but played around with the frequency and "fine" control until the 2 lines on the spectrum dislpay lined up with the black and white parts and the picture look OK. Not sure if seatty has a similar window, but jvcomm is free and might be worth a go.

 

[iainhu]

Many thanks for the pointer to the YouTube sample. Extremely helpful. Attached is what I recorded from 4608 just after 1300UTC. Sounds pretty similar - not as strong, but it would sound like I'm getting closer. Will try again at 1400.

I manually put SeaTTY into fax receive mode plus ran Mscan Meteo which was happy to run at the same time. Mscan Meteo liked to put big black horizonal lines through the fax and even printed a 'helpful' 'Please upgrade' message through the centre of the reception. A good trial version! Doesn’t really inspire me to pay for a copy

SeaTTY seemed to make a better job with more white than black, however, there was no real chart as the 'pattern' of black dots drifted towards the right across the lines. So am I decoding but in effect have my horizontal hold wrong? Interesting at 1320 the signal altered and the fax picture turned to mostly black with the odd white spec, pretty much reversing the previous pattern. End of transmission?

SeaTTY gives options of 60, 90, 120, 180, 240 and 360 lines per minute. It was running in 120. I tried 60 and 90 but couldn’t seem to get a better image…still squint lines. Is 120 (the SeaTTY default) the correct scan rate?

There is also a ‘Index of cooperation’ setting of either 576 (default) or 288. Any ideas on what this means or should be set to?

Will go and look for a copy of ‘jvcomm’ and hope to give it a trial at 1400UT

Many thanks again…

 

[Conachair]

Many thanks for the pointer to the YouTube sample. Extremely helpful. Attached is what I recorded from 4608 just after 1300UTC. Sounds pretty similar - not as strong, but it would sound like I'm getting closer. Will try again at 1400.

I manually put SeaTTY into fax receive mode plus ran Mscan Meteo which was happy to run at the same time. Mscan Meteo liked to put big black horizonal lines through the fax and even printed a 'helpful' 'Please upgrade' message through the centre of the reception. A good trial version! Doesn’t really inspire me to pay for a copy

SeaTTY seemed to make a better job with more white than black, however, there was no real chart as the 'pattern' of black dots drifted towards the right across the lines. So am I decoding but in effect have my horizontal hold wrong? Interesting at 1320 the signal altered and the fax picture turned to mostly black with the odd white spec, pretty much reversing the previous pattern. End of transmission?

SeaTTY gives options of 60, 90, 120, 180, 240 and 360 lines per minute. It was running in 120. I tried 60 and 90 but couldn’t seem to get a better image…still squint lines. Is 120 (the SeaTTY default) the correct scan rate?

There is also a ‘Index of cooperation’ setting of either 576 (default) or 288. Any ideas on what this means or should be set to?

Will go and look for a copy of ‘jvcomm’ and hope to give it a trial at 1400UT

Many thanks again…

My jvcomm settings are IOC - 576, LPM - 120. I'm getting a signal now on 8038.1, though it fades in and out a little and isn't a great image.
the attached file still sounds slightly different, the high clear tone on yours is a little higher than the one I'm getting. Might be worth tuning just 0.1hz higher, see if that helps.

You must be nearly there now though

 

[iainhu]

You must be nearly there now though

I sure hope so! BTW, no image attached to your post?

Okay, so I was in 120LPM and IOC or 576 so those are the correct settings. I was running both jvComm and SeaTTY together. The SeaTTY image again looked like it could be an image only with each line drifting to the right to look rather like some bad 1960s wallpaper Of the two the SeaTTY image looked a little better, but somehow vanished before I saved it…must have hit a key in coming indoors. Transmission seemed to end at 1417UT.

Interesting to note one or both of the packages are adjusting the mic input level. If I raised the level during transmission I could see the input volume slider notching down to near zero, and now the radio is disconnected and so there’s no input, the mic input level is on max.

Have tried 518 and in Navtex mode but no luck there.

Anyway, my magnificent chart is attached. This is from jvComm which is a lot blacker than the SeaTTY image (which I lost). Both though show the right to left drift.

So then…

1. I seem to be getting the right signal, even if not tuned quite right. I noticed tuning it to a higher pitch gave me more white in SeaTTY and looked more like a squint fax. Tuning to a lower sound gave me mostly black
2. I have my LPM and IOC correct
3. The mic input seems to be being adjusted by one or both of the packages. jvComm nicely shows a mic input level although it tended to be red more than green which is maybe a symptom. Maybe I should use the headphone out so that I can turn down the line input to the laptop?
4. While I can get mostly white with some black spots (in SeaTTY at least), in no way could they be considered a readable fax

So really, it would seem most of my variables are sorted…so am I down to tuning and maybe input level? Presumably you want to see the Audio level in jvComm in green up around the vertical bar about 2/3 of the way up the scale. Presumably red indicates the signal is too strong, hence the idea of switching from the line out to the headphone so I can adjust the output signal??

 

[iainhu]

1500 effort

My latest effort, which pretty much looks like previous efforts. This is from the 8038KHz 1500 transmssion. Interesting that around 1517UT the fax auto-stopped after what looks like an end black block and change in tranmission tone, suggesting things are partially there.

You'll see what I mean about the rightwards drift. It might be a fax image if each line were shifted back to the left.

SeaTTY reports is can't detect the LPM rate which it tries to do. The simultaneous jvComm image was much like the previous one

Tried different line/audio leads with no noticable effect, so while I might be getting nearer, I ain't there yet and am not quite sure what will be the magic to make it all work. Fine tuning, different leads, different software, same output

Sorry for poor quality of image...but not allowed to upload many more images

 

[Pye_End]

On my set-up, JVCOMM gives black images with white lines. No idea why, or whether this is normal. Mscan gives white images with black lines.

120 lines is correct. There should be a slant correction somewhere. In JVCOMM there is a FAX-tools button with slant correction on it. Havn't tried it, and may not work on a demo copy. Mscan it is under File - Configuration - Fax, at least on my version.

The images I have been pulling off this afternoon have not been clear. Recognisable, but a struggle to get much information from them. The other day they were as clear as internet images. The German RTTY has been fairly good though.

I think that you need to concentrate on fine tuning. Try narrow, and tweeking the tone, and then trying to get the biggest peaks, and make sure they are in between the markers (unless SeaTTY finds the peaks automatically). Also play with the volumes. Also try not de-tuning it 2 below and see if you get better response. I tried de-tuning mine and the picture was worse.

Mscan seems to erase images when the next one comes in, unless I am quick enough to save it first. Frustrating.

 

[iainhu]

Black with white lines makes sense in that what I see is a lot more black. No image, but white lines on black would therefore make a lot of sense.

Yes, 'Narrow' would be a smart move so thanks for the hint...will just have to keep trying I guess, but it does seem to be a bit of a black art getting it to all to work

 

[iainhu]

Oh, and was looking at slant, but in the SeaTTY example at least, you need a whole lot more of an image to adjust that the more 'random scattering' I'm seeing

 

[iainhu]

With a little help from your friends...

Well, as the song says, with a little help from your friends…not forgetting a bucket-load of encouragement

I am now getting charts…I still need to sort ‘horizontal rotation’ which I now at least understand, but I am most definitely getting there

The Eureka moment came by taking a bold stab at a slant correction. So using something very similar to the complete gibberish I posted before, I (in SeaTTY) drew a slant correction and started collecting again. What came out was a recognisable chart and so it was sampling adjustment (via the slant adjustment) that sorted things. Once I had something resembling a synoptic I redid my slant adjustment so as to ‘fine tune’ and carried on collecting. I also switched to ‘narrow’ which probably helped the quality, but the key was entering the slant adjustment into the ‘gibberish’. Clearly this is the purpose of the black boxes down the left-hand side of the charts as it was these I was able to make out from the ‘gibberish’.

I then set about ‘slant correcting’ Mscan Meteo and jvComm and had all three collecting at the same time. An interesting experiment. Mscan claimed to do slant/frequency adjustment automatically, but entering a manual slant adjustment (the most awkward of the three) suddenly started giving results.

SeaTTY was very good at just keeping going. Having manually started it, it just kept on recording. jvComm and Mscan liked to stop and so I have less output from them. Mscan also likes to save its fax images in BMP format which is a uncompressed bitmap image, e.g. it takes huge amounts of disk. SeaTTY and jvComm both save in PNG which I a compressed bitmap format and so will ‘eat’ less disk space. Mscan has an option to save in GIF format although as that is horizontally compressed I’m not sure it’s the most efficient compression, but I digress…

So to the results… Well, YBW won’t let me upload much more here, so instead there are some links to the images as collected, e.g. not resized down to meet upload requirements of the forum. The exception is the Mscan image which I’ve saved to jpeg simply because the small section was 5.5Mb.

So for SeaTTY look at: http://www.hunneybell.org/hf-fax/SeaTTY-111016-173039.PNG
For Mscan Meteo look at: http://www.hunneybell.org/hf-fax/mscan20111016-1.jpg
And for jvComm look at: http://www.hunneybell.org/hf-fax/jvComm-HF-Fax_20111016_1643.PNG

The SeaTTY image is by far the most complete and gets better as you go down as I fine-tuned. This is because SeaTTY just kept going having been manually started. I also slant-correct SeaTTY first. Note the horizontal shift still needs to be sorted. You may also need to click on the image in your browser as it’s a long image and so the browser may shrink it to fit onscreen – it’s not really that shrunk and ‘bitty’

The small section of chart collected by Mscan is probably the clearest although ruined by having ‘UNREGISTERED’ plastered over it. A shame you can trial it first to see how it works without this kind of thing.

jvComm is the worst (most grainy) of the three, but is black/grey lines on white, not white on black. There is an ‘Invert picture’ option in ‘SSTV Picture Processing’ (top right button, a little sun-like symbol) – although I can’t find the ‘Picture processing’ option no matter what is says on page 40 of 91 of the English documentation. However what I can find is that in File:Configuration:FAX mode editor there is a ‘Phase mode’ option of ‘normal’ or ‘inverted’. So Pye_End, you might want to see if your config is set for ‘inverted’ and so possibly cure your white line on black inverted image?

So I will need to do some more experimenting. I need to get my horizontal offset sorted in each, but I am getting there. I’d be very interested to know how the images compare to ones other people have been receiving today, as it will give a clue as to the effectiveness of the receiver.

So thank to everyone for help, advice and most importantly, encouragement. I will do some more ‘fiddling’ and hope to report back with some better comparisons of SeaTTY, Mscan Meteo an jvComm, meanwhile enjoy the images, all from 8038KHz this late afternoon

 

[iainhu]

Harwich or Walton Back Waters

Pye_End, should I presume you're not too far down the road from me and are based in or around Harwich or the Walton Back Waters, hence your 'nom de plume'?

 

[Pye_End]

Well done. Looks like some good results.

This is one I pulled off on the 7th (before I straightened it!):

I was getting similar quality for several days. However, although this is not from today, it is a bit more like what I was getting (but not quite so poor):

Please let us know how you get on with faxes over the coming days - the receiver looks good value very compact. Good value if that is the quality of image that you can expect. Interesting to hear what the different softwares are capable of as well. Will be interested in how you get on with RTTY and NAVTEX.

In the link that Tim posted earlier there is a comprehensive list of the different faxes on offer. However, from the titles it is not clear what they all are. It would be useful to know which are the most relevant - eg current pressure/front analysis, and forecasts.

 

[Pye_End]

Pye_End, should I presume you're not too far down the road from me and are based in or around Harwich or the Walton Back Waters, hence your 'nom de plume'?

Afraid not. Boat is currently Medway. My username was named after the buoy to which you refer for a couple of reasons.

 

[iainhu]

Afraid not. Boat is currently Medway. My username was named after the buoy to which you refer for a couple of reasons.

The Medway? Well I sail from Chatham marina fairly regularly and know the river and from my YM, many of its 'unavigable' creeks, better than I might care to