General foo2hiperc :  foo2zjs - foo2hp - foo2xqx - foo2lava - foo2qpdl - foo2oak
Support for linux printer drivers
The fastest message board... ever.
 
ghostscript 9.+ is broken w.r.t. color. Don't use
Posted by: elle2004 ()
Date: February 20, 2014 03:47PM

Hi,

In what way is the color broken in ghostscript 9.+

I'm using ghostscript version 9.10 F19 64bit and I get pale printouts
compare to Windows XP on same OKI C5200 printer.
In XP a pretty good printout but on linux a person with black hair
get a least 20 years older. A lot of colors are incorrect
compared with XP's printout.

Tested with different icc files in /usr/share/foo2hiperc/icm without
any noticeable difference....



Edited 1 time(s). Last edit at 02/21/2014 04:25AM by elle2004.

Options: ReplyQuote
Re: ghostscript 9.+ is broken w.r.t. color. Don't use
Posted by: rickrich ()
Date: February 21, 2014 06:04PM

GHOSTSCRIPT BUGS
----------------
    * ghostscript 8.64 is broken w.r.t. color.  Don't use!

    * ghostscript 9.00 is broken w.r.t. fonts in landscape mode.

    * ghostscript 9.+ is broken w.r.t. color. Don't use

    * Well Tempered Screening (WTS) works with ghostscript 8.54 thru 8.70+.
      So if you want BEST COLORS, you'll have to install gs 8.54 or better
      if you don't have it.

        $ gs --version
        8.15.2

        $ wget "http://mirror.cs.wisc.edu/pub/mirrors/ghost/GPL/gs871/ghostscript-8.71.tar.gz";
        $ tar zxf ghostscript-8.71.tar.gz
        $ cd ghostscript-8.71
        $ ./autogen.sh
        $ make
        $ su                            OR $ sudo cp bin/gs /usr/bin/gs.foo
        # cp bin/gs /usr/bin/gs.foo

        (Optional)
        # make install                  OR $ sudo make install

        $ gs.foo --version
        8.71

      If that doesn't compile (e.g. Ubuntu 11.10+), try this:
        $ cd ghostscript-8.71
        $ ./configure  --without-jbig2dec
        $ make CFLAGS+="-DHAVE_SYS_TIME_H=1"


Options: ReplyQuote
Re: ghostscript 9.+ is broken w.r.t. color. Don't use
Posted by: elle2004 ()
Date: February 23, 2014 07:20AM

Hi,

I prefer to keep current GS version, make life easier during yum updates.

I did some testings and what ever I did with different ICM profiles nothing happens.
Then I saw in page 30 http://www.ghostscript.com/doc/9.05/GS9_Color_Management.pdf that (CRDs) are ignored in version 9.05

With a little patching from /user/bin/foo2zjs-wrapper into foo2hiperc-wrapper

gs9)
# Use new icc method in gs 9.*
AIB=
BC=
if [ -r "$GAMMAFILE" ]; then
GAMMAFILE="-sOutputICCProfile=$GAMMAFILE"
elif [ -r "$SHARE/icm/$GAMMAFILE" ]; then
GAMMAFILE="-sOutputICCProfile=$SHARE/icm/$GAMMAFILE"
else
GAMMAFILE=
fi
;;

and

/usr/share/foomatic/db/source/opt/foo2hiperc-ICM.xml
<enum_vals>
<enum_val id="ev/foo2hiperc-ICM-none">
<ev_longname> <en>No ICM color correction</en> </ev_longname>
<!-- A multilingual <comments> block can appear here, too;
it should be treated as documentation for the user. -->
<ev_shortname> <en>none</en> </ev_shortname>
<ev_driverval>-Cgs9 -Gnone.icm</ev_driverval>
</enum_val>

<enum_val id="ev/foo2hiperc-ICM-301-600M">
<ev_longname> <en>OKC301H1.icm - 600dpi Multi C301dn</en> </ev_longname>
<ev_shortname> <en>OKC301H1</en> </ev_shortname>
<ev_driverval>-Cgs9 -GOKC301H1.icm</ev_driverval>
</enum_val>

All with -Cgs9....

Now colors are substantial better, not 100% as XP but going from 20% up to 80%, need some more work.

I stopped working on colors for a moment and focused on the compression.
A question.
How sure is it that OKI using jbig compression.
If it is jbig a reverse decompression from a XP to printer dump should work but i dosn't work.

Her is a session taken by Wireshark and run by hipercdecode -d XP4 XP4.raw


LPD session
\033%-12345X@PJL
@PJL JOB NAME = "WindowsNT"
@PJL RDYMSG DISPLAY = ""
@PJL OKIJOBACCOUNTJOB USERID="" JOBNAME="Untitled - Notepad"
...
@PJL ENTER LANGUAGE=HIPERC
RECTYPE 0 (len=52,0x34 cnt=1), Page 1
BLKNUM 0, nbie=1, pn=3 [black] uc=128,17, wid=4760 ud=0,10000000
BLKNUM 1 (len=20), uncompressed=0, bie:
DL = 0, D = 0, P = 1, - = 0, XY = 4760 x 6912
L0 = 256, MX = 16, MY = 0
Order = 0
Options = 72 LRLTWO TPBON
27 stripes, 0 layers, 1 planes
RECTYPE 1 (len=23,0x17 cnt=1)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=65,0x41 cnt=2)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=45), Data=6b 11 02 54 25 a7 45 ad 25 35 2b 13 15 43 34 cd a9 e3 07 7d ...
RECTYPE 1 (len=240,0xf0 cnt=3)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=220), Data=16 05 99 68 ea 70 b2 4c 94 06 ab 8b eb 1a 33 e0 98 3c 58 30 ...
RECTYPE 1 (len=23,0x17 cnt=4)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=5)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=6)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=7)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=8)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=9)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=10)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=11)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=12)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=13)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=14)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=15)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=16)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=17)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=18)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=19)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=20)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=21)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=22)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=23)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=23,0x17 cnt=24)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
RECTYPE 1 (len=279,0x117 cnt=25)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=259), Data=d7 5b 18 f1 62 50 e8 9a ab 39 e7 a9 75 4d 07 27 f3 9a fb 18 ...
Data=64 ea d8 6e 0f 72 98 f3 3d 59 4b 1b 2e fb fc 2a ae 74 ef 81 3f 35 4d f3 f1 eb fd 0e 93 9a ...
Warning: jbg_dec_in c=64 i=86 cnt=86 rc=110 (Input data stream contains invalid data)
RECTYPE 1 (len=23,0x17 cnt=26)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
Data=00 ff 02
Warning: jbg_dec_in c=0 i=0 cnt=0 rc=110 (Input data stream contains invalid data)
RECTYPE 1 (len=23,0x17 cnt=27)
BLKNUM 0 (len=4), plane=3, uc=0,0,0
BLKNUM 1 (len=3), Data=00 ff 02
Data=00 ff 02
Warning: jbg_dec_in c=0 i=0 cnt=0 rc=110 (Input data stream contains invalid data)
RECTYPE 255 (len=8,0x8 cnt=28)
\033%-12345X@PJL
@PJL EOJ NAME = "End "
\033%-12345X

As you see block 25, 26 and 27 got too much data.



Edited 1 time(s). Last edit at 02/23/2014 07:24AM by elle2004.

Options: ReplyQuote


Sorry, only registered users may post in this forum.
This forum powered by Phorum.