[sane-devel] Fwd: HP scanjet IIcx /t resolution

George Georgalis George Georgalis <georgalis@gmail.com>
Mon, 6 Sep 2004 21:00:50 -0400


Hi Peter.

Thanks. Right, the site I looked at before I bought the special cable
must have been wrong or I didn't read it right (remember 600 optical
and 1600 interpolated). Oh well 400dpi does seem correct per sites
today and is decent for a freebie; at least it's fast, and based on
the original retail price, I think I saved ~$1,100, heh :)

Oh and the part about going to one bit and not going back... I can't
reproduce it now, but at some point I noticed I could type a single
digit, but not drop down for color depth.
Another odd thing I noted is use of 'own color map' results in black
image, my video card may not be 100% Linux compatible though (nVidia).

I'm thinking it is scanning  "3x8 bits = 24 bits" but the viewer does
say "1 bit/color, 3 colors" cf http://galis.org/onebitcolor.png (209k,
it's a thumbnail from a contact sheet)

Anyway, thanks all for your help. I'm impressed with the low noise,
expert help on this list!

// George



On Mon, 06 Sep 2004 20:51:52 +0200, Peter Kirchgessner
<peter@kirchgessner.net> wrote:
> Hi,
>
> the HP SCL documentation says that the IIcx has a native resolution of
> 400 dpi and that it supports grayscale and color scans of 8 bits per
> channel. So scanning at more than 400 dpi does not give more details.
> When selecting to do color scans, it will scan with 3x8 bits = 24 bits
> which is about 16.777.216 possible colors. More colors are not supported.
>
> --Peter
>
> George Georgalis schrieb:
>
>
> > On Mon, 6 Sep 2004 15:22:42 +0100, Martin Collins <martin@mkcollins.org> wrote:
> >
> >>On Sat, 4 Sep 2004 18:21:11 -0400
> >>George Georgalis <georgalis@gmail.com> wrote:
> >>
> >>
> >>>It's spec says 600 dpi but somewhere between 300 and 400 dpi it
> >>>starts to do double passes (exactly when it starts depends on the
> >>>image). I think I understand what's happening, though I don't know
> >>>the vocabulary to explain it (software gets a higher resolution than
> >>>the optical quality requiring multiple passes).
> >>
> >>You are thinking of "software interpolation". That's not it though.
> >>I would call this behaviour "backtracking" and like Abel says it is
> >>probably due to a bottleneck in the data path.
> >>
> >>Are you using the SCSI card that came with the scanner? Does it have
> >>a serial-port style connector or a proper SCSI connector?
> >>If the former this is probably the problem. Getting a proper SCSI card
> >>will fix it. I recommend Iwill as good value for money.
> >
> >
> > Thanks for both responses -- makes sense, not  interpolation but,
> > channel bandwidth.
> > Seems I was scanning different widths when scanning different images.
> >
> > But I still find this odd as I have scsi-2 controller, (Buslogic
> > FlashPoint) connected to the older "scsi-1" scanner. It's unfortunate
> > and strange that the scanner is faster than the the bus on it...
> > apparently.
> >
> >
> > More importantly it I've not been able to set the bit depth to greater
> > than 8, I think the scanner is rated to 16. There have been times I've
> > adjusted xsane settings when the bit depth has gone to 1, and even
> > restoring the old settings will not make 8 bit depth available again,
> > I have to fix the setting and restart it.
> >
> > In the viewer console it says "1 bit/color, 3 colors"  ...I'm don't
> > think the image is more than 8 bit but I think is is that many. So I
> > seem stuck with 256 colors, what could be wrong? (BTW - this box has
> > 1Gb memory)
> >
> > // George
> >
> >
>
> --
> Peter Kirchgessner
> http://www.kirchgessner.net
> mailto:peter@kirchgessner.net
>
>


--
George Georgalis, systems architect, administrator Linux BSD IXOYE
http://galis.org/george/ cell:646-331-2027 mailto:george@galis.org