![]() |
Lee Howard <faxguy@deanox.com> writes: > >> The fax recipient of the fax associated with the following log claimed that > >> the faxed page(s) were all black. I tried the same fax three times, and > >> each time he said the pages were black. I ended up faxing successfully > >> with a Canon fax machine manually. > >> > >> Faxing to this person isn't critical to my business, but if this log helps > >> indicate a protocol bug in HylaFAX... > > > >I don't think that's protocol bug. > > Why would it work fine with the manual faxing on the Canon and not via > HylaFAX, then? > > .... > >.. and now we see why. It seems to be that *very* rare case, when RTN means > >sudden line quality degradation and is not related to any protocol > >bugs. Look, we were able to train at 14400 for the first time, but now only > >training at 7200 has succeeded! Surely something horrible has happened to > >the line in the meantime... > > Why would it be consistently bad? Would the line quality suddenly become > degraded when using HylaFAX or my modem but not when using the manual fax? > I just tried doing the same fax with 'ModemRate: 7200', and the log looks > better (no 4-minute sending), but I can't say whether or not the page is > all black. So you always have got the same log? Then it really looks like a bug :-( Could you post such log when the bit rate is limited by 7200 cps? I only ask you to collect the log for absolutely the same source document. You may even intercept intermediate TIFF (see below) and then send it directly. > >The second page was above 200 Kbytes! No surprise, that it required more > >than 4 minutes to send. But Hylafax itself has nothing to do with it. The > >possible reason is that you are trying to send a grayscale image (not a > >text). Ghostscript may rasterise it with inacceptable quality (looking > >"black" on the other side), and moreover such images are hard to compress > >with T.4 algorithms used for fax transmission. That's why we had so much > >data to send. > > Hmm. I guess I could try sending an ASCII page to them for faxing, but > this is the first time this has happened, and I *always* use the same fax > client which *always* submits the fax in the same TIFF format to HylaFAX. > Meaning, there's never a variance in the type of image that I'm submitting > to HylaFAX (and in this case the content of the image is very typical of my > normal faxes that I send), so why would it differ now? Can you catch the TIFF that Ghostscript is created from your PS document? It resides in docq directory with a name like doc1956.ps;70 until the fax is in the send queue. Copy it to the safe place and rename to test.tif; then use your favourite TIFF viewer to see what is really inside. What is its size? Hope to hear from you soon, Dmitry ____________________ HylaFAX(tm) Users Mailing List _______________________ To unsub: mail -s unsubscribe hylafax-users-request@hylafax.org < /dev/null