rick@pcrat.uucp (Rick Richardson) (08/06/89)
In article <9634@b-tech.ann-arbor.mi.us> zeeff@b-tech.ann-arbor.mi.us (Jon Zeeff) writes: >I'm interested in getting a fax board for a PC (running dos or unix). >It's not going to save any money over a standalone machine, but I believe >that it will be more flexible. > >Does anyone have any good/bad experiences with fax boards and scanners. Is >anyone using the fax under unix stuff that Stuart Lynn has? Boards ------ I've dealt with both the Quadram JT Fax 9600 (PC Mag Editors choice, April 11, 1989), and the Brooktrout TR111. Both boards are high quality and a bit pricey. The Quadram failed one compatibility test to the Canon Faxphone 20, and the Brooktrout failed two tests - to the Canon Faxphone 20 and also to the Facsimile Pack. The Brooktrout beat the Quadram in transmission speed. These tests were performed by PC magazine using the vendor supplied DOS software. Your mileage may vary under a third party UNIX implementation. I've run the Quadram using VP/ix, and I've heard that the Brooktrout can also run under VP/ix, but haven't tried it yet. The biggest difference in the two boards is the hardware interface between the card and the PC. Brooktrout choose to use DMA, a couple of IO ports, and interrupts, while the Quadram has a 2K dual ported memory (locatable on a 2K boundary anywhere from C8000 to F0000) and no interrupts. Personally, I would have designed a board with (minimal wait state) dual port memory *and* an interrupt, giving the best bandwidth between the board and the disk as well as the instant notification that the board needs servicing. The Brooktrout board is also loaded with features that the Quadram doesn't have - some obviously there to support combined voice and FAX systems. The end result is that the Quadram needs a very simple UNIX device driver, while the Brooktrout device driver is an order of magnitude more complex. In our case, we supply and support the UNIX device driver for the Quadram. We will support *the interface to, but not the driver for* the Brooktrout board. All of the usual PC limitations will apply in choosing a FAX board: if you are short of interrupts: go for the Quadram. If you have trouble disabling the cache between C8000 and F0000, go for the Brooktrout. UNIX Packages ------------- I've seen three advertised, plus our own. There's Certifax, Bell Tech, Driver Design Labs, and our own JetFax (PC Research). I think that the Bell Tech package was withdrawn (temporarily?) from the market when Bell Tech was bought by Intel. The Driver Design Labs package is not scheduled for availability until November and the price hadn't been set, though they said the software would be $300-$800 (according to UNIX Today!). Certifax has been advertising for some time now in the UNIX magazines. Our package ships this week for ISC 386/ix 2.0 or UNIX System V Release 3.2 and the Quadram board at a cost of $249 for the software. Beyond the usual features of spooled scheduled transmission, group broadcast capability, ASCII or Group 3 file transmission or reception, up to 8 boards per computer, and printing to a laser printer, our own package adds the capability to send troff documents using up to 7 different supplied typefaces, previewing on the UNIX EGA compatibile console, automatic user defined cover sheet generation. Some packages also claim automatic routing to the recipient. These claims are more fantasy than reality, there being only a few ways to achieve this: sender does something "special" on the cover sheet to identify the desired receiver, sender sets the twenty character identification string to identify the receiver, or receiver has his own unique telephone "number". The first method is flawed, since there is no standard to go by in determining where or what is the "special" thing. The second method is flawed, since it is difficult or inconvenient to set the identification string on many FAX machines (and its usage is supposed to be to identify the sender!). The third method is practical, and can be achieved at several different cost levels. Justification ------------- There are really 4 reasons to choose a PC Fax board that I can see: - You don't want to wait in line for the regular machine - You send to groups of people frequently - You send mostly text files, or, with our package only, troff documents. These come out looking absolutely beautiful. - You don't have a copier available to convert the FAX machine's Dead Sea Scrolls to regular paper. Notice that the advantages almost all lie on the sending side. If you are only an occasional FAX sender, you are probably better off with a regular FAX machine. There is also a fairly major problem with receiving FAXes to a PC. Resolution. Faxes get sent at ~200 DPI, but laser printers are 300 DPI animals (the NeXT 400 DPI laser starts looking like a good move). You can print the FAX at 2/3 size (portrait or 2 on 1 landscape) and get the identical (but reduced) image. Or you can print at full size using a little image processing which results in an image that isn't identical to the original, but is the same size. Depending on the algorithm chosen, text can look pretty good, but images (especially ones with a lot of alternate dots on-off) can get ugly. In either case, the printing of the received FAX takes much longer than if it had come directly to a regular FAX machine. -Rick Richardson -- Rick Richardson | JetRoff "di"-troff to LaserJet Postprocessor|uunet!pcrat!dry2 PC Research,Inc.| Mail: uunet!pcrat!jetroff; For anon uucp do:|for Dhrystone 2 uunet!pcrat!rick| uucp jetroff!~jetuucp/file_list ~nuucp/. |submission forms. jetroff Wk2200-0300,Sa,Su ACU {2400,PEP} 12013898963 "" \d\r\d ogin: jetuucp