Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 

   


   


   















      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[filmscanners] Minolta - CCD vs software problem



    Same problem for me with my Minolta Scan Speed ( yes, quite old for now
I know , but still ready! ): tram lines always located at the same place ,
running through the whole image.
    Very few tram lines in the red channel , some more on the green channel
and a bit lesser in the blue channel. I process the images with PS ,
selecting tram lines and balancing brightness with adjacent lines for each
channel (noise remove filter , size 1 , level 0). Works great.
    I'm suspecting some sort of CCD degradation problem. When becoming old ,
some cells appear to be quite "lazy"... Especially in dark areas of the scan
( which become bright after inverting neg film ). What a matter of joke
seeing some psychedelic purple lines right in the sky!
    I don't use the Minolta software for scanning , to bad for me. Vuescan
is far better ( despite the fact it becomes slower and slower after each new
release , where is the fantastic speed of v7.3.10? I can't believe new
version lazyness is due to extra color computation!!! Don't speek about
noise reduction , i don't use it in any case. )
    Something strange concerning these tram lines is that they appears ( on
my own DSS ) at different locations in the three channels. Never could show
up a tram line in a channel other than the one it clearly appears:there's no
link between tram lines in the red channel and tram lines in the green
channel (e.g.). I tried to find tram evidence in the 2 other channels than
the one where the tram line is clearly shown up: nothing! The image looks
perfect in these 2 others channels.

    Here's the point I can't explain myself:

    I scan only neg films , using Vuescan for color conversion. Ed said
color conversion cannot be achieved with LUT. I can't disagree this point ,
it's obvious. Ed said color conversion is done by 3x3 matrix computations ;
the film dye response is considered to be linear ( that's not true , but
it's better than LUT process ). OK. That means , the red channel ( or green
or blue channel ) of the final image is composed of some parts of the THREE
channels composing the RAW scan. So, the tram line present in the raw data
should appear in the three channels of the final image , right? That is not.
Why?

    Then afterwards , CCD problem or software problem?
    Any advice will certainly contribute to solve this question!

    Pascal S.

----- Original Message -----
From: "Arthur Entlich" <artistic-1@shaw.ca>
To: <asap@evc.net>
Sent: Wednesday, October 23, 2002 1:39 AM
Subject: [filmscanners] Re: A partial fix for Minolta scanners?


Hi Tony,

After I wrote the message (in part quoted below) I did look at the
sample images, and yes, it is very similar looking.  I also brought both
samples into Photoshop, upped the levels to see each channel more
clearly, and these "tram" lines do indeed appear to be in random
locations, different with each channel (the red channel is very banded
in this case).  The defects I have seen have looked more like the green
channel showing in the Vuescan example.  One line, one pixel in thickness.

Having said that they look the same, however, doesn't mean they are the
same (although they may be).

1) Do these tram lines always show up in the same locations relative to
  the CCD position?  The ones I saw do.  Do they run the full length of
the frame?

2) The lines I saw show up with the Minolta software, (and as I recall
also with Vuescan,... I tested this over a year ago.)

3) I have seen the problem on three different Dual Scan II units, and
each had repeatable lines in unique locations on each channel depending
upon the specific unit.  The problem has also shown up on Elite II
models, but I only saw one time samples, so I can't say if these defects
were stationary or if they moved around.  My current model has one in
the green.  Earlier models has several in each channel.  Same computer
system and software.

The thing that makes me wonder about the similarity is if the lines
always show up in the same location.  I would think that if the problem
is interface, or timing related, the tram lines would tent to be more
random in terms of positioning.  It would be great if this was just a
software related problem that was correctable, but I tend to think there
may be two unique issues, or it might be that Vuescan is showing an
inherent hardware defect in the CCD or firmware calibration routine
which is diminished in visibility with the Minolta software???


I certainly wouldn't mind being kept up to date on developments
regarding this matter.

Thanks,

Art




Tony Sleep wrote:

> On Mon, 21 Oct 2002 18:15:18 -0700  Arthur Entlich (artistic-1@shaw.ca)
> wrote:
>
>
>>In the case of the DUal Scan II, it is a USB
>>interface, rather than SCSI.
>>
>
> Duh, yes, sorry. Mind you, USB can be weird too.
>
>
>>The "tram lines" I am speaking of are
>>specific, in that they occur in certain exactly repeatable locations,
>>are different locations (and numbers) depending on the color channel,
>>and they are always one pixel wide.  In the case I am speaking of, this
>>is an apparent hardware and or calibration defect.
>>
>
> What do you make of:
>
>>>>http://www.onedition.com/Minolta.jpg
>>>>http://www.onedition.com/Vuescan.jpg
>>>>
>
> Same fault or different?
>
>
> Regards
>
> Tony Sleep - http://www.halftone.co.uk
>
>


----------------------------------------------------------------------------
------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title
or body

----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe 
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or 
body



 




Copyright © Lexa Software, 1996-2009.