[Leica] RE:Re: RE: Re: The Last Words on B&W Scanning (Don Dory)

[hoppyman at bigpond.net.au (G Hopkinson)]

Thanks Don, that's thoughtful reasoning.
I know this method does work well for you and Tina as we have seen.

I see the data from all three sensors used to produce the output, whether
RGB or Greyscale. As you said, some question regarding what the scanning
software does with that input. This might equally apply to how it produces
three separate channels of colour where none exists. 

Assuming both outputs are 16 bit, must be identical tonal range. The RGB
method is going to give you three versions of the same 16 bits, each altered
by how that channel is sensitive to the grey tones of the film. Analogous to
the effect of coloured filters on the film for say more dramatic skies or
better skin tones. You can manipulate the grey scale output later to do the
same by changing to RGB and using the channel mixer.

We are in total agreement regarding manipulating the scan as little as
possible, Just different philosophies regarding when and with what the
needed manipulation is performed. (and just what juju is within the scanner
software) I think that PS should be more capable than the scanning software
and so prefer to not even crop nor always auto expose within the scanning
process. I don't want to throw any information away at the scan stage and
would check for clipping, of course.

Thanks again for contributing on this.

Cheers
 Hoppy 


-----Original Message-----

Message: 12
Date: Thu, 13 Jul 2006 21:04:21 -0400
From: "Don Dory" <don.dory@gmail.com>
Subject: Re: [Leica] RE: Re: The Last Words on B&W Scanning (Don Dory)
To: "Leica Users Group" <lug@leica-users.org>
Message-ID:
        <9b678e0607131804g5a281d5bk7b8eeb55cf5bb181@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Hoppy,
I scan in RGB mode as I don't want the scanner to control any part of the
scan process.  In RGB mode you are using all the sensors in the scanner to
record exactly whatever they see.  In grayscale I don't know if the scanner
is norming some pixels as there is no color information; this is too much
for me.

Secondly, the RGB file in PS offers more bits to manipulate so less clipping
off of potentially valuable information, again, I am in control of what goes
into the image.  Part of this is back in the old days of 8 bit files.

My theory on adjustments is that I want to manipulate the image bits as
little as possible.  So, if I adjust the scan then there is less to do in PS
so less loss of data as severe adjustments are done.  An exception to this
would be an image that I think shows promise but I am not sure what I want
to do with it.  In that case I will output a pretty flat image that still
holds data from almost white  almost black.  Indeed, these files will
capture all the information on the negative that the scanner can capture.

Don
don.dory@gmail.com

On 7/13/06, G Hopkinson <hoppyman@bigpond.net.au> wrote:
>
> Don, I'm a keen student on BW scanning and printing currently. A number of
> LUG people have been kindly offering suggestions and other help or
> opinions.
> I know that both you and Tina recommend this scanning of BW as RGB. Others
> are happy to scan at greyscale.
>
> Would you explain why you feel the RGB method is superior? Certainly the
> file is much larger, and you can manipulate those channels for more
> control.
> But it seems to me that the RGB channel information is false, in that
> there
> is no colour information to be had. So the scanner is interpreting the
> grey
> image to invent the colour information.
>
> I'm leaving aside file manipulation after the scan for printing or other
> reproduction. Also, I'm not referring to producing BW images from colour
> negs or transparencies.
> You are describing producing (very desaturated) colour images from a black
> and white original.
> Regarding scanning resolution, my take is that there is no benefit in
> using
> less than the maximum real (not interpolated)resolution. I'm assuming here
> that the system used won't take much longer to produce say a 4000dpi scan
> vs
> 2000dpi.  Certainly the time difference is more significant with flatbeds
> and/or larger than 35mm, though.
> Still, you will have extracted the most possible (scan) information from
> the
> film. Then from that best possible file you can down-sample as you wish,
> dependant on output desired.
>
> I understand your advice regarding setting the black white points and the
> curve dialogue. Another approach is to have the scan software adjust as
> little as possible and do this with the more powerful Photoshop along with
> the crop and sharpening etc after you have your "raw" scan. Different
> philosophy
>
> Cheers
>
> Hoppy
> Very willing to learn
>
>
>
> -----Original Message-----
> ------------------------------
>
> Message: 6
> Date: Thu, 13 Jul 2006 16:44:21 -0400
> From: "Don Dory" <don.dory@gmail.com>
> Subject: Re: [Leica] The Last Words on B&W Scanning
> To: "Leica Users Group" <lug@leica-users.org>
> Message-ID:
>         <9b678e0607131344m17e8691eg7791ca6c2f7acfd4@mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Bob,
> Definitely scan in RGB mode as you will get a larger file with more
> information.  Always scan at the largest bit depth that your scanner will
> allow.  Now comes the hard part, at what resolution to scan at.  All I can
> suggest is to try the same negative at different resolutions and see what
> comes out.  I have seen grain aliasing on one set of negatives that a
> fellow
> lugger helped me with by scanning on his Leaf.
>
> Last, turn off any sharpening in the scan phase.  Also do enough
> adjustments
> after the prescan that you get all the information from the negative.
> Typically that involves adjusting the white and black points as well as
> using a curve dialogue to bring the tonality close to what you want.
>
> Don
> don.dory@gmail.com
>
>
> On 7/13/06, bob palmieri <rpalmier@depaul.edu> wrote:
> >
> > Folks -
> >
> > Okay; I'm trying to get over my general disappointment with the way my
> > B&W film scans looks (Tri-X or Neopan 1600 in Xtol).  I now need to
> > scan several things within the next coupla hours.
> >
> > I see a lot of you folks posting some damn good-looking B&W images from
> > these film/developer combos. Is the collective wisdom that I should
> > scan the negs as color negs??  Howzabout scanning as positives and
> > flipping them in Photoshop??  Does scanning at 3200 or so often lead to
> > less grain aliasing than 4800? (I have a not-so-great Epson 2870 and
> > Silverfast.)
> >
> > Since I use the Digest mode, if anyone feels so kind as to E me
> > directly I'd sure appreciate it.  I gotta get these thing out...
> >
> > Bob Palmieri
> >