UPGRADES TO SSTV/FAX480/WEFAX SYSTEM We have added a number of improvements to the soft- ware-based system described in reference 1. A brief outline follows with a much more extensive description being included in the instruction manual which is includ- ed in the new software. (1) Since the Wefax pictures take up a lot of memory space to store the whole thing, we have added the capa- bility to save a portion of the image in FAX480 format. This gives a "blow-up" of the area of interest, and provides another source of FAX480 files. Figure 1 shows a visual mode sample of this. Get out your magnifying glass and look at the surprizing amount of ground detail avail- able in these HF images when you have 64 shade process- ing. If the recipient of the image has our system, he can colorize the IR images to see cloud tops, ocean currents, etc. (2) Automatic recognition of the VIS codes has been added as a mode in the receive program. The recognized mode is printed out and the machine continues to copy in that mode. Manual choice of a mode is available while in Automatic if the VIS header got QRM'ed out. (3) Another automatic mode is identical to (2) above except it has an automatic SAVE added for every picture copied, with the file names being sequential letters of the alphabet. I use the Ramdisk as the recipient store for these files while in unattended operation, as I'm afraid I'll forget it's turned on and fill up my hard disk. The VIS recognition doesn't pick up very weak signals, so most of the images stored are not too noisy. (4) We added an Automatic mode to the SLIDESHO program. You set the dwell time you want for each slide, and when running, just hit A(uto) to start a continuous sequence of all the current directory files in the cate- gory chosen. The slide sequence continues indefinitely until you halt it. (5) KY1S added a feature to his copy of the transmit program (actually the View and Transmit program, VT.BAS) that automatically provides a listing of all the file names in the current directory which match the mode chosen (ie,Scottie 1, etc.) after you activate VT.BAS. You pick a file picture from this selection and it is displayed for you to be sure it is the one you wanted. This only took a few BASIC commands to implement, so I added it. Since not everyone might want this feature, a switch is provided in the CONFIGURATION list to turn it on or off. Dennis has provided a perfect example of how you can adapt the program to your own operating style using the BASIC format. (6) Since we found that 2-color files for the LABEL program were more available from the multitude of Paint programs in a PCX format, we added a PCXLABEL program. (7) Some people objected to the "tone tick" audio cue but were not satisfied with the loss of information when they turned it off with the program switch. I find this invaluable as a measure of the "heart-beat" of the system. Since some computers have a fairly loud speaker, you can combat the noise by inserting a resistor in series with the speaker lead to reduce the volume. Some- what reluctantly, I have added a visual screen cue that will light when the machine is ticking if you must turn off the speaker (lines 485-487 in RT.BAS). (8) Since the program uses 262,144 color processing and display, picture sources must support the high color resolution. Some had difficulty finding image format manipulation programs which offered an output in 24-bit TIFF format to match my TIFCONV program. Listing the people who have found solutions to this: KY1S reports using Tempra and Photo-Finish 3; N9ARX reports using Paint Shop Pro For Windows; WB7PAP reports using Graph- ics Workshop 7.0 (he found this on the same Steve Rimmer BBS listed in reference 1 as a source for video drivers: 416-729-4609). The last two of these are shareware pro- grams. These programs accept different color resolution files in GIF, PCX, etc. formats. The final SSTV files, of course, have no more resolution than the source files. (9) One unexpected result that has arisen as the system has been tried on a larger number of systems is a much wider dispersion of Time-of-Day clock frequencies in many of the newer systems (mostly 486 systems). I chose this source of system timing specifically because it was generally held to close tolerances. Since the Time gets updated from the battery-operated clock every time the computer reboots, this tolerance apparently was let slide due to competitive pressures. The line timing adjust- ments, LT, in some cases required a larger correction than expected, but most people handled that well. They report that you only need determine one mode's line timing experimentally, and all the others come right on using the LT formula. I have added another simple formula which makes the initial correction possible after copying only one picture to ease the work if you have a system which diverges from the mean significantly. A somewhat more serious problem has arisen in some of these systems, as the wider variation in frequency has moved the transmit header frequencies (ie, the Frame Sync) far enough to exceed the acquisition window of most systems with automatic VIS recognition. This can be adjusted out with a few judicious Pokes. To make this easy we have added a 3 position switch to the SYSTEM CONFIGURATION in VT.BAS which allows you to put the system in a Test mode so as to measure and adjust the header frequencies directly. In one extreme case, the frequencies were far enough to break out of SYNC. This adjustment should take care of such situations and also satisfy those who want their signals "right on" frequen- cy. (10) Some of my sailing friends who use laptops with only 16 gray shades displays wanted the improved process- ing on Wefax incorporated into those programs. Also some Wefax users prefer the 16 shade format since it uses only 1/2 the memory for storage. A collection of these pro- grams including automatic scheduling for maps has been included for those folks and others who only have 16 color display capability. (11) Another problem which has shown up in a few cases is noisy I/O cards. This shows up as noise in the picture even with very strong signals. KY1S reported this first and fixed it by replacing the offending card with an older type he had around. Others have used the same fix. I have no significant data on which brand cards are faulty and more importantly which ones aren't. If those who have encountered this problem and solved it will send me the information, I will include it in later revisions of the software for others to use. These cards are cheap ($10 to $15) so replacement shouldn't be burdensome. (12) Other information gleaned from helping others get the system working is included in the new instruction manual, TVINFO.TXT. The above additions are included in a new release available on the ARRL BBS (tel: 203-666-0578) as file VESTER-A.ZIP. By the time this gets published, the revi- sion may be beyond -A. Ben Vester, K3BC 4921 Bonnie Branch Road Ellicott City, MD 21043 Reference: (1)B. Vester, "An Inexpensive SSTV System", QST, Jan 1994, pp 27-29.