plustek: Test Version 0.42-0 (Parallelport and USB)
Subject:
Re: Test Version 0.42-0 (Parallelport and USB)
From:
Gene Heskett ####@####.####
Date:
18 Feb 2002 02:01:43 -0000
Message-Id: <20020218015656707.AAA323@mail.iolinc.net@there>
On Sunday 17 February 2002 05:38 pm, Jaeger, Gerhard wrote:
>Hi there,
>
>I've currently released the test-version 0.42-0. Main features
>are:
>- support for UMAX 3400 (thanx Allan)
>- downloadable Gamma tables for all supported devices
>- warmup time and lamp off timer for the USB devices
>- check plustek.conf for more options to set upon startup
>
>Please download and test it. Note: Use SANE-1.0.7 as base
> version! Also NOTE: Check your configuration file
> (plustek.conf)
>
>For the USB guys:
>Allan told me about two problems:
>1. The MinDpi is not used:
>This is true, but the major problem seems to be in the function
>usb_SetAsicDpiY() in plustek_usbscan.c. I think there should be
>an additional test against the MinDpi.y setting...
>The user is free to set 50dpi as minimum. If the minimun
> resolution of the scanner will be 100dpi, the reader function
> performs this scaling...
I'd noticed that 50 actually worked, but at that resolution, the
term butt ugly comes to mind. Like a 99% compressed jpeg.
> 2. The number of buttons has influence
> on the shading function: True, Plustek uses this to distinguish
> their models - I've added an additional MODEL_NOPLUSTEK branch
> to overcome this problem there, but I've no idea what influence
> this has on the other devices...
This epson has 4 buttons, of course non-functional, but how does
this effect the shading? Inquiring minds want to know... :-)
>Oh, Gene - I've also had the magenta in some of my pictures, but
> this went away after a certain warmup period...
I changed the effect by changing the scanner for a different one.
Warmup period allowed here didn't seem to have nearly as much
effect as a days layoff did on the first scan, with all scans
after that first one being very similar. IMO its either related
to the color of the end of the lamp, which I doubt, or its
related to the quality of the paint job on the calibration strip.
For this, I think the individual unit may have to have a 'fudge
file' per resolution and mode, to be added to the data sent back
during the calibration phase. Here, running on the paint, a
piece of 104% paper seems to be read at about 90%, which allows
the shading errors to be very easily seen. But if I calibrate on
the paper, then apparently we need to fudge the glass loss back
out as its then scaled up to where a 95% white is lost in the
white clipping. The shading looks great, till you realise the
highlight detail is also missing.
Maybe I'm too fussy. With half or more of a fairly long life
spent in my own color darkroom, even going so far as to mix my
own color paper developer for many years, ones eyes tend to get
pretty critical, and this is after all, only a 100 dollar
scanner. :-)
>So if this version is okay, it will go to the sane CVS. The I
> start and check the calibration stuff National uses for the
> reference code... We may have to use this one to make
> non-Plustek devices work...
>
>Okay that's it for now - thanks in advance for testing...
> Gerhard
--
Cheers, Gene
AMD K6-III@500mhz 320M
Athlon1600XP@1400mhz 512M
98.5+% setiathome rank, not too shabby for a hillbilly
