Graeme, I don’t suppose you have a little utility that will allow me to transpose the order of the data in my 2310 patch ti3 file? From: tony22p [mailto:tony22p@xxxxxxxxxxx] Sent: Thursday, June 30, 2011 4:45 PM To: 'argyllcms@xxxxxxxxxxxxx' Subject: RE: [argyllcms] Re: New user - chartread error Tsong-Min, you are RIGHT! I was assuming that the reading direction was the same as the text direction at the bottom of the chart. When I read it left to right it worked fine. THANK YOU! Graeme, may I suggest you add a reading direction arrow or something on the test charts? It is not intuitively obvious. From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx] On Behalf Of ??? (Tsong-Min Wu) Sent: Thursday, June 30, 2011 10:01 AM To: argyllcms@xxxxxxxxxxxxx Subject: [argyllcms] Re: New user - chartread error I suspect that you have read the test chart upside down. You should start with the end with printed alphabet. I have made that error when I first read the chart. Tsong-Min From: tony22p <mailto:tony22p@xxxxxxxxxxx> Sent: Thursday, June 30, 2011 8:19 PM To: argyllcms@xxxxxxxxxxxxx Subject: [argyllcms] Re: New user - chartread error Hmm, my Canon Pro9000 MarkII is supposed to be able to print via a 16 bit data path. They supply a XPS printer driver which runs in Windows 7 (I’m in Win7 x64), which supports this. This is my default driver for this printer and one which I’ve used with Photoshop for a while. I opened the test charts in PS without converting to a working space, and it opened as a 16 bit image. Printing of course was done with no color management. I’ve done this before with other print profiling solutions and have not run into the problems I’m having now. However, one can always learn something by trying. I will try an 8 bit target to see what happens. From: argyllcms-bounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx] On Behalf Of Nikolay Pokhilchenko Sent: Thursday, June 30, 2011 2:24 AM To: argyllcms@xxxxxxxxxxxxx Subject: [argyllcms] Re: New user - chartread error Graeme, here were my original generation commands: targen -v -d2 -G -f2310 -g256 -s256 name printtarg -v -ii1 -T360 -p203x254 name May be Your printing software can'd deal with 16-bit tiffs? And there was patch miss-coloring when you've printed the charts? Are the chart images on display looks like the printed images? Can You view the chart images through another image viewing program for reference? If Your program haven't correct support of 16-bit images, it can display and print these files equally wrong. I'd recommend You to re-generate the charts with 8-bit tiffs: printtarg -v -ii1 -t360 -p203x254 -Q8 name -----Original Message----- From: wlmailhtml:sentmsg?compose <wlmailhtml:sentmsg?compose&To=argyllcms%2dbounce@xxxxxxxxxxxxx> &To=argyllcms%2dbounce@xxxxxxxxxxxxx [mailto:argyllcms-bounce@xxxxxxxxxxxxx] On Behalf Of Graeme Gill Sent: Wednesday, June 29, 2011 10:36 PM To: wlmailhtml:sentmsg?compose <wlmailhtml:sentmsg?compose&To=argyllcms@xxxxxxxxxxxxx> &To=argyllcms@xxxxxxxxxxxxx Subject: [argyllcms] Re: New user - chartread error tony22p wrote: > Alastair, I tried re-reading the first row (A) after having recreated the > ti2 file from the seed number. I got the same error. It thinks I'm reading > row CD and not row A. Note that such warnings are not 100% reliable, particularly if you did not use a realistic profile as an argument to targen -c. But it is suspicious if every read gives this warning, and if the expected values differ wildly from the values you are reading. Very high self-fit delta E's in colprof would confirm a mis-read chart. Graeme Gill.