[sane-devel] Re : Re: LiDE 90 half ccd

[Guillaume Gastebois]

Hello,

Sorry not to answer quickly I have had lots of work.
Nothing very new since months....

I just do some windows snoops in 1200dpi and 2400dpi translated into C with usbsnoop2libusb.pl.
they are : http://ggastebois.free.fr/lide90_snoop/lide90_1200dpi_216x7.c
and http://ggastebois.free.fr/lide90_snoop/lide90_2400dpi_216x7.c

I don't see big differences....

> Guillaume Gastebois schrieb:
>> > Hello,
>> > 
>> > No progress on 2400dpi. DPIHW is 2400 and dpiset 600 for a 300dpi scan.
>> > Don't know where to look yet for that problem !
> 
> I'd try to get a log from a 2400dpi scan from windows and compare.
> 
>> > about byte nibbles modifying reg_0x79 from 0x3f to 0x3e (very few 
>> > contrasted image) or 0x40(black image) gives me bad images !!!
>> > 
>> > What is effect of this register on clk3 ?? It's very sensible.
> 
> register 0x77 to 0x79 describe a bitmask, that determines when the clk3 
> signal is high or low. For 0x0003f(lsb is register 0x79), this would give:
> 
> time===>>                          ! next pixel
> bit of 0x79    ! 0x78          !0x77! 0x79
> 0 1 2 3 4 5 6 7!0 1 2 3 4 5 6 7!0 1!0 1 2 3 4 5 6...
> 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 ....
> ___________                         ___________
>             |_______________________|           |____
> 
> 0x40 is this:
> 
> bit of 0x79    ! 0x78          !0x77! 0x79
> 0 1 2 3 4 5 6 7!0 1 2 3 4 5 6 7!0 1!0 1 2 3 4 5 6...
> 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 ....
>              _                                   _
> ___________| |_________________________________| |____
> 
> As you can see, this gives a very short pulse, which will probably not 
> be recognized by the ccd, while 0x3e just shortens the clock pulse at 
> the beginning. The datasheet mentions that the scanner does not need to 
> use the full 18 bits of the bitmask, but uses the first 6, 12 or 18 bits 
> depending on the scanning mode. 12 seems to be correct for the mode the 
> backend programs for cis scanners.
> 
> Please try if modifying bitmask for clk4(0x7a to 0x7c iirc) shows any 
> effects. Also, other values you may want to try on clk3: 0xff03f, 
> 0xff801f, 0xffc00f. Another thing to try may be to send two clock 
> pulses, for example 0x001c7. If clk3 is connected to the ccd, this may 
> give interesting results(like, reducing the resolution by half), but for 
> the 4-bit latch/D-FlipFlop, only the last clock pulse before the gl841 
> samples the data bits is relevant.

I tryed lot's of differents values for clk3 (differents pulse width and positions) succesless.
Only entropy program on offset files gives me sometimes malformed crosses.

> 
>> > regards,
>> > Guillaume
> 
> Regards,
>    Pierre

Regards
Guillaume