Some background:

At the time of preparing this page, most images have been obtained using my old hardware configuration: a RIG RX1 or RX2 together with the IF1 interface feeding into the bi-directional parallel interface in a 16 MHz AT286 computer. This latter is networked to other computers which are more suited for data workup, though the same software [JVFAX Ver. 7.0] is used in each.

 By today's standards, this is a very old hardware package indeed, and during coming months, it's due to be superceeded by a more up to date system using faster & more modern software packages- JVComm32, SatSignal, WXSat and WxtoImg. They all have their virtues, and the jury is still out on which one I'll eventually settle upon (indeed, others may well come along to change all that.)

To begin with, I thought I'd go right back to the beginning and show one of my early images, in fact the very first one that I archived. It dates from 12th. August, 1995 and shows a morning pass, southbound, from NOAA 12. It was producing fine imagery in those days (as it still is, long past its design lifetime):

NOAA 12 - Channel 2, A.M. 12th. August, 1995

It's pretty poor isn't it, especially when compared with the images on the index page ! Some of the defects are worth commenting on. The tilt and curve of the image are due to the use of the serial rather than the synchronous parallel input. It should be remembered that the RF carrier is at a nominal 137.5 MHz and the sub-carrier at 2400 Hz. As the satellite approaches, the received frequency will in fact appear to change due to the Doppler effect, being slightly higher on approach and slightly lower on recession. The receiving computer 'expects' the data to arrive at a constant rate, and starts each received line at the appropriate time. As the data actually arrives sometimes early, sometimes late, the displayed position of each line will vary due to this Doppler shift, giving the picture its 'bent' appearance. The same effect will be seen with any software that makes this kind of assumption. Moreover, the computer assumes that the correct timing can be obtained from the computer's internal clocks, but in general these will only be close to, but not exactly equal to the correct sampling rate. As a result, the picture may also be tilted as a whole one way or another by an amount depending on the degree of error. Software packages which work in this way normally have a feature built in which enables a correction to be applied to the computer clock so as to receive the picture without tilt - JVFAX as well as JVComm32, WxtoImg for example have this feature.

A perfectly straight picture can be obtained if the software package gets its timing data from the satellite signal itself. The later guise of JVFAX does just that by extracting a synchronising signal from the 2400 Hz subcarrier; the data is input through the parallel interface and each line is not started until the sync pulse is received from the interface. This completely gets round the Doppler shift problem, and can also be emulated in more recent packages where the decoding software derives its synchronisation from the digitised data (via the sound card) directly. Sometime one can decode synchronously or asynchronously, according to choice.

Some other problems are evident - dropouts or uneven areas of signal. These are due to deficiencies in the receiving antenna which was a 2 metre vertical intended for amateur VHF communications. In an ideal world, the receiving antenna would have an equal response in all directions from horizon to horizon. In practice, few antennas manage to come close to that, and many fall far short of the mark. A 2 metre vertical is pretty unsuitable in this case for several reasons - it's the wrong length to start with and has the wrong orientation, where horizontal would have been a bit better. The length means the impedance matching will be incorect but also that the response in different directions will vary significantly. This means that whilst a good signal will be received in some directions as the satellite passes over, in others it will fall away or even disappear - this is the cause of the dark bars at intervals. In these places the receiver AGC tries to compensate with a resulting increase in noise. There is also some computer noise from direct pick-up by the antenna as it was not far away - generally these should be as far away from each other as possible, as a precaution.

So much for that, lets look at a recent image - NOAA 17 from its morning pass on 21st. January, 2003.

NOAA 17 A.M. 21st. January, 2003

This image illustrates some other points, most obviously that it consists in fact of two images. That on the left is the Channel 2, visible/near-IR image whilst that on the right is the Channel 4, thermal IR image. (See the Sending link for more details about this.) I'd like to think that the image quality has come some way since the earlier example !   It also shows one of the useful features of JVFAX in that the Vis & IR images are acquired simultaneously, are identically cropped and displayed together for easy comparison. Other packages may treat the channels separately. This image was acquired using a turnstile antenna (see the Receiving link for more details.) The coverage is very much more even without dropouts. The antenna at around 5 metres above ground, is about 20 metres from the receiver/computer, has a low noise masthead pre-amplifier (Gain around 12 dB) and about 30 metres feeder. Actually it's not quite true to say there are no dropouts for indeed there is one which is usually observed just below the bottom of this picture and arises from a combination of properties of this particular antenna, its situation in relation to the roof of my house and the surrounding buildings. It is due to be replaced with a quadrifilar hybrid during the coming year when exterior conditions improve !   It's not seen here as I've edited it out, but it does presently limit my south looking coverage, though hopefully that will change in time.

The rest of this page consists of a couple of galleries which I've compiled from the archives, together with a selected picture - this is intended to be one of the more interesting pictures that I've had during the past week or so. So do come back regularly to have a look, and of course, I'd always value feedback.   

E-mail me !   



Back to Introduction

© Stuart Hill [Updated February 2003]