Drab Scans of Slides with Nikon LS-4000 w/Nikon Scan 3.1

Discussion in 'Digital Photography' started by john chapman, Feb 9, 2004.

  1. john chapman

    john chapman Guest

    I do bulk scans of Slides with a Nikon LS4000 using Nikon Scan 3.1,
    with final manipulation in Photoshop, but am trying to minimize the
    amount of post scan manipulation as much as possible. Right now my
    scanned slides are dark and rather lifeless (dull colors and low
    contrast)-- nowhere near the life of the original slide. I am in
    dire need of some suggestions on how to set up Nikon Scan parameters
    for my bulk scans.

    I am able to brighten the scans by increasing the Master Gain (to
    about +.6) in the Analog Gain Palette, but the colors and contrast
    still remain a bit muted. Most scans still require increasing the
    contrast, at a minimum. But I am not sure this is the approach I
    should be using since the Brightness Function in the Color Balance
    Palette and the functions of the LCH Editor Palette would seem also to
    have some effect on the exposure of the scans, as well. It is obvious
    to me that I really do not understand the use/purpose of these other
    controls.

    I would appreciate some suggestions on the initial setup, as a
    starting point, rather than my continuing to flail around in the dark
    (pun not intended) on a trial and error basis. Thanks in advance for
    assistance.
     
    john chapman, Feb 9, 2004
    #1
    1. Advertising

  2. john chapman

    Bill Hilton Guest

    >From: (john chapman)

    >I do bulk scans of Slides with a Nikon LS4000 using Nikon Scan 3.1,
    >with final manipulation in Photoshop, but am trying to minimize the
    >amount of post scan manipulation as much as possible. Right now my
    >scanned slides are dark and rather lifeless (dull colors and low
    >contrast)-- nowhere near the life of the original slide. I am in
    >dire need of some suggestions on how to set up Nikon Scan parameters
    >for my bulk scans.


    The overall dynamic range should have been set with auto exposure. This seems
    to do a good job with the brightest areas (ie, doesn't overexpose them) but for
    the black point it will pick up the slide border as the blackest area and maybe
    skew the results, especially if you don't have a solid black point in the
    slide. Then when you crop off the mount border you may need to reset the black
    point (you should be able to see this on the histogram or in Levels after you
    crop ... if there's a gap on the left side of the histogram this is what
    happened).

    For the "overall dark" problem I'd try increasing the gamma setting, you may
    need to go as high as 2.8 or so ... take a slide with a known medium gray
    (ideally an IT8 target) and scan it with the default settings and measure the
    grey point in Photoshop, then bump up gamma and repeat (you have to get out of
    NikonScan for the change to be updated). The end points should still be set by
    autoexposure but this might help you with brighter midtones.

    >Most scans still require increasing the
    >contrast, at a minimum. But I am not sure this is the approach I
    >should be using since the Brightness Function in the Color Balance
    >Palette and the functions of the LCH Editor Palette would seem also to
    >have some effect on the exposure of the scans, as well. It is obvious
    >to me that I really do not understand the use/purpose of these other
    >controls.


    I'm pretty sure the Color Balance and Brightness controls are linear, meaning
    if you increase brightness you also move the black point up by the same amount,
    which would explain why the scans are lacking in contrast.

    Good luck.

    Bill
     
    Bill Hilton, Feb 9, 2004
    #2
    1. Advertising

  3. john chapman

    Jim Guest

    Have you created a custom profile for your scanner? I used Monaco, and that
    helped a lot..

    Jim
    "john chapman" <> wrote in message
    news:...
    > I do bulk scans of Slides with a Nikon LS4000 using Nikon Scan 3.1,
    > with final manipulation in Photoshop, but am trying to minimize the
    > amount of post scan manipulation as much as possible. Right now my
    > scanned slides are dark and rather lifeless (dull colors and low
    > contrast)-- nowhere near the life of the original slide. I am in
    > dire need of some suggestions on how to set up Nikon Scan parameters
    > for my bulk scans.
    >
    > I am able to brighten the scans by increasing the Master Gain (to
    > about +.6) in the Analog Gain Palette, but the colors and contrast
    > still remain a bit muted. Most scans still require increasing the
    > contrast, at a minimum. But I am not sure this is the approach I
    > should be using since the Brightness Function in the Color Balance
    > Palette and the functions of the LCH Editor Palette would seem also to
    > have some effect on the exposure of the scans, as well. It is obvious
    > to me that I really do not understand the use/purpose of these other
    > controls.
    >
    > I would appreciate some suggestions on the initial setup, as a
    > starting point, rather than my continuing to flail around in the dark
    > (pun not intended) on a trial and error basis. Thanks in advance for
    > assistance.
     
    Jim, Feb 9, 2004
    #3
  4. In article <>, john
    chapman <> writes
    >I do bulk scans of Slides with a Nikon LS4000 using Nikon Scan 3.1,
    >with final manipulation in Photoshop, but am trying to minimize the
    >amount of post scan manipulation as much as possible. Right now my
    >scanned slides are dark and rather lifeless (dull colors and low
    >contrast)-- nowhere near the life of the original slide. I am in
    >dire need of some suggestions on how to set up Nikon Scan parameters
    >for my bulk scans.
    >
    >I am able to brighten the scans by increasing the Master Gain (to
    >about +.6) in the Analog Gain Palette, but the colors and contrast
    >still remain a bit muted. Most scans still require increasing the
    >contrast, at a minimum. But I am not sure this is the approach I
    >should be using since the Brightness Function in the Color Balance
    >Palette and the functions of the LCH Editor Palette would seem also to
    >have some effect on the exposure of the scans, as well. It is obvious
    >to me that I really do not understand the use/purpose of these other
    >controls.
    >
    >I would appreciate some suggestions on the initial setup, as a
    >starting point, rather than my continuing to flail around in the dark
    >(pun not intended) on a trial and error basis. Thanks in advance for
    >assistance.


    Before messing about with the scanner settings, make sure that your
    monitor is set up correctly. You have Photoshop, so you could use Adobe
    Gamma for this. Be careful, the first step - getting the black and
    white level - is critical and a poor black level will inevitably result
    in you selecting the wrong gamma.

    Apart from the Analogue Gain controls, every other adjustment is made on
    the scanned data. Specifically, the LCH palette has NO effect on the
    exposure. It merely permits post scan adjustments exactly the same as
    the Curves palette, but according to a Lightness-Chroma-Hue model
    instead of the RGB model. The only difference between the master RGB
    Curve and the Lightness is the weighting of the red green and blue
    channels.

    In NS, go to the Preferences section Colour Management tab and enable
    Nikon Colour Management, making sure you have the colorspace you want to
    use in Photoshop selected under RGB. While there, make sure that you
    have autoexposure and autofocus switched on under Preferences|Single
    scan. If you are going to scan film strips make sure the same setting
    is used for Batch scan as well. Also go into the Advanced Colour
    section and adjust the sample point size to 5x5 pixels andAutocontrast
    excluded level to something like 0.1% or lower fro the black and white
    levels. Reset the white, grey and black point targets as well, just in
    case you have changed these at any time in the past.

    Take a slide which has a reasonable amount of pure white and pure black
    in it and preview that. The autoexposure should result in a scan which
    is pretty well exposed but, to be sure, look carefully at the histogram
    in the Curves Palette.

    Be sure to clear any adjusments that you might previously have applied
    to the black, white and gamma levels by pressing the master reset icon
    (bottom right with three lines). The histogram should extend from 0 to
    255. If you find that the white is clipped at a level significantly
    below 255 then adjust the Master Gain control and preview again to bring
    the clipped level closer to that value, but don't exceed it. If there
    is a lot of saturation at 255 then adjust the Master Gain in the
    opposite direction and preview again. If you notice two or three spikes
    near the right hand side of the histogram, look at the histograms of the
    individual colours and adjust the individual Gain for these colours
    accordingly to get something close to 255 in each one.

    Having achieved a reasonable exposure, select an area of the preview
    using the crop tool which includes most of the image (certainly the near
    black and white areas) but none of the slide mount. Then return to the
    curves window and click on the autocontrast icon (circle with half
    white, half black) and this should return the curves window
    automatically to the RGB view and adjust the histogram levels to almost
    exactly the full 0 to 255 range. However you might want to adjust
    manually the upper and lower ranges in each colour to reduce the slight
    clipping that inevitably results. The idea is to get the black and
    white points set to exactly where black and white lie in the image.

    Having done all of that, select your crop to exactly the area you want
    for all future scans, and select the ICE, GEM & ROC setting you desire,
    together with 8-bit depth. Save the setting under the "Settings" menu
    to a name you will remember and also "Set User Settings", which will
    save these settings as your defaults for subsequent scans.

    That should get you pretty close to optimum scans, so try a couple of
    images to begin with. If there is a slight residual colour cast, then
    look for an area that you know is neutral mid-grey in an image and use
    the grey dropper in the Curves Palette to select it in the preview. This
    should take out any "slope variation" in the film dye transmission that
    might exist between the broad spectrum that your eyes see and the narrow
    spectrum that the LEDs see. When you are happy, save the setting and
    Set User Setting.

    You should be ready to go for bulk scans now but, depending on what mix
    of film and age they are, you might need to revisit the settings and
    save new settings for specific films. Remember to use the Kodachrome
    setting if you are using KC emulsion - which will almost certainly
    require fine adjustment of the Curves for every batch of film. You
    might also want experiment with the other settings for global gamma in
    the Curves palette, the Colour Balance palette or Colour Saturation
    (chroma in the LCH palette), and Unsharp Mask, and your final preferred
    default setting for these can be saved under the settings menu as above.

    Hope that helps.
    --
    Kennedy
    Yes, Socrates himself is particularly missed;
    A lovely little thinker, but a bugger when he's pissed.
    Python Philosophers (replace 'nospam' with 'kennedym' when replying)
     
    Kennedy McEwen, Feb 9, 2004
    #4
  5. john chapman

    Ken Guest

    In article <>,
    (john chapman) wrote:

    > I do bulk scans of Slides with a Nikon LS4000 using Nikon Scan 3.1,
    > with final manipulation in Photoshop, but am trying to minimize the
    > amount of post scan manipulation as much as possible. Right now my
    > scanned slides are dark and rather lifeless (dull colors and low
    > contrast)-- nowhere near the life of the original slide. I am in
    > dire need of some suggestions on how to set up Nikon Scan parameters
    > for my bulk scans.
    >


    Which slide film are you using?
     
    Ken, Feb 9, 2004
    #5
  6. john chapman

    john chapman Guest

    At the moment Fuji Sensia 100, 200, and 400, with some Provia.

    Ken <> wrote in message news:<>...
    > In article <>,
    > (john chapman) wrote:
    >
    > > I do bulk scans of Slides with a Nikon LS4000 using Nikon Scan 3.1,
    > > with final manipulation in Photoshop, but am trying to minimize the
    > > amount of post scan manipulation as much as possible. Right now my
    > > scanned slides are dark and rather lifeless (dull colors and low
    > > contrast)-- nowhere near the life of the original slide. I am in
    > > dire need of some suggestions on how to set up Nikon Scan parameters
    > > for my bulk scans.
    > >

    >
    > Which slide film are you using?
     
    john chapman, Feb 10, 2004
    #6
  7. john chapman

    john chapman Guest

    I am still playing around with your suggestions, which have been very
    helpful. I do, however, have several other questions.

    It seems to have no effect whether I set autoexposure to on or leave
    off. Does this mean the AE is defective in the ls4000? Or is AE a
    software based control?

    I have also been experimenting with Vuescan which does not seem to
    have a setting for AE. However, with little or no setup or tweaking,
    it gives better results than Nikon Scan. The Vuescan images are a bit
    more contrasty and have more of the vibrance and depth that slides are
    known for.

    Are the settings saved in the Tool Pallette part of the overall
    settings, or must they be loaded separately each time?

    If I understand your instructions correctly, I should take a good
    slide with range of values and containing black and white areas. I
    should scan the slide and set the values in the tool pallette to
    provide good results, save those settings. Then each time I run with
    that setting, the software will provide the same tweaking as specified
    in the save settings. Is that essentially correct?



    Kennedy McEwen <> wrote in message news:<>...
    > Before messing about with the scanner settings, make sure that your
    > monitor is set up correctly. You have Photoshop, so you could use Adobe
    > Gamma for this. Be careful, the first step - getting the black and
    > white level - is critical and a poor black level will inevitably result
    > in you selecting the wrong gamma.
    >
    > Apart from the Analogue Gain controls, every other adjustment is made on
    > the scanned data. Specifically, the LCH palette has NO effect on the
    > exposure. It merely permits post scan adjustments exactly the same as
    > the Curves palette, but according to a Lightness-Chroma-Hue model
    > instead of the RGB model. The only difference between the master RGB
    > Curve and the Lightness is the weighting of the red green and blue
    > channels.
    >
    > In NS, go to the Preferences section Colour Management tab and enable
    > Nikon Colour Management, making sure you have the colorspace you want to
    > use in Photoshop selected under RGB. While there, make sure that you
    > have autoexposure and autofocus switched on under Preferences|Single
    > scan. If you are going to scan film strips make sure the same setting
    > is used for Batch scan as well. Also go into the Advanced Colour
    > section and adjust the sample point size to 5x5 pixels andAutocontrast
    > excluded level to something like 0.1% or lower fro the black and white
    > levels. Reset the white, grey and black point targets as well, just in
    > case you have changed these at any time in the past.
    >
    > Take a slide which has a reasonable amount of pure white and pure black
    > in it and preview that. The autoexposure should result in a scan which
    > is pretty well exposed but, to be sure, look carefully at the histogram
    > in the Curves Palette.
    >
    > Be sure to clear any adjusments that you might previously have applied
    > to the black, white and gamma levels by pressing the master reset icon
    > (bottom right with three lines). The histogram should extend from 0 to
    > 255. If you find that the white is clipped at a level significantly
    > below 255 then adjust the Master Gain control and preview again to bring
    > the clipped level closer to that value, but don't exceed it. If there
    > is a lot of saturation at 255 then adjust the Master Gain in the
    > opposite direction and preview again. If you notice two or three spikes
    > near the right hand side of the histogram, look at the histograms of the
    > individual colours and adjust the individual Gain for these colours
    > accordingly to get something close to 255 in each one.
    >
    > Having achieved a reasonable exposure, select an area of the preview
    > using the crop tool which includes most of the image (certainly the near
    > black and white areas) but none of the slide mount. Then return to the
    > curves window and click on the autocontrast icon (circle with half
    > white, half black) and this should return the curves window
    > automatically to the RGB view and adjust the histogram levels to almost
    > exactly the full 0 to 255 range. However you might want to adjust
    > manually the upper and lower ranges in each colour to reduce the slight
    > clipping that inevitably results. The idea is to get the black and
    > white points set to exactly where black and white lie in the image.
    >
    > Having done all of that, select your crop to exactly the area you want
    > for all future scans, and select the ICE, GEM & ROC setting you desire,
    > together with 8-bit depth. Save the setting under the "Settings" menu
    > to a name you will remember and also "Set User Settings", which will
    > save these settings as your defaults for subsequent scans.
    >
    > That should get you pretty close to optimum scans, so try a couple of
    > images to begin with. If there is a slight residual colour cast, then
    > look for an area that you know is neutral mid-grey in an image and use
    > the grey dropper in the Curves Palette to select it in the preview. This
    > should take out any "slope variation" in the film dye transmission that
    > might exist between the broad spectrum that your eyes see and the narrow
    > spectrum that the LEDs see. When you are happy, save the setting and
    > Set User Setting.
    >
    > You should be ready to go for bulk scans now but, depending on what mix
    > of film and age they are, you might need to revisit the settings and
    > save new settings for specific films. Remember to use the Kodachrome
    > setting if you are using KC emulsion - which will almost certainly
    > require fine adjustment of the Curves for every batch of film. You
    > might also want experiment with the other settings for global gamma in
    > the Curves palette, the Colour Balance palette or Colour Saturation
    > (chroma in the LCH palette), and Unsharp Mask, and your final preferred
    > default setting for these can be saved under the settings menu as above.
    >
    > Hope that helps.
     
    john chapman, Feb 10, 2004
    #7
  8. (john chapman) wrote in message news:<>...
    >
    > I have also been experimenting with Vuescan which does not seem to
    > have a setting for AE. However, with little or no setup or tweaking,
    > it gives better results than Nikon Scan. The Vuescan images are a bit
    > more contrasty and have more of the vibrance and depth that slides are
    > known for.
    >

    As far as I see, VueScan will autoexpose unless you tell it not to.
    You can do that with the "lock exposure" setting on the Input tab
    after doing a preview. You can crop part of the image and lock
    exposure to what it has calculated for that part crop. Then recrop to
    the whole image and have it use the locked setting. (I find this is
    helpful with sunsets, where the whole image contains a lot of
    silhouette or shadow which is of no interest, together with beautiful
    sky which you want perfectly exposed). Or you can choose lock exposure
    and overtype VueScan's calculated values with values of your choice.
    But there does not seem to be any way of doing that independently for
    red, green and blue. However, based on my limited experience, VueScan
    does indeed produce a much better result than NikonScan, with or
    without my intervention!

    Even VueScan attempts to give you as much of the dynamic range as it
    can from the image by being conservative with white and black points.
    If you want punch, without using an image editor, you can be much more
    aggressive with the white and black points and sacrifice some
    highlight or shadow detail.

    I have found that VueScan can produce images that look on my monitor
    to be identical in both colour and punch to the slide as projected by
    my slide projector. Playing around with manual colour balance is very
    powerful, particularly with right mouse click.
     
    Stephen Rogers, Feb 10, 2004
    #8
  9. In article <>, john
    chapman <> writes
    >I am still playing around with your suggestions, which have been very
    >helpful. I do, however, have several other questions.
    >
    >It seems to have no effect whether I set autoexposure to on or leave
    >off. Does this mean the AE is defective in the ls4000? Or is AE a
    >software based control?
    >

    Autoexposure is a software based control, but I think you might have to
    restart NS to enable the change to take effect.

    Autoexposure takes a low resolution pass of the image and then adjusts
    the actual exposure to bring the highlights up to a level which achieves
    a nearly full CCD without saturating it. It isn't like the autoexposure
    of a film camera, which effectively adjusts the exposure to get the
    average brightness level to match a defined reference. If it did that,
    then it is quite likely that highlights would saturate the CCD, which
    has a hard limit rather than soft limit of film.

    >I have also been experimenting with Vuescan which does not seem to
    >have a setting for AE. However, with little or no setup or tweaking,
    >it gives better results than Nikon Scan. The Vuescan images are a bit
    >more contrasty and have more of the vibrance and depth that slides are
    >known for.


    Vuescan has specific algorithms built in to get you close to the correct
    colour balance right away. Many folk prefer that, but I have never been
    really happy with its output myself for reasons I have given in these
    parts many times before.
    >
    >Are the settings saved in the Tool Pallette part of the overall
    >settings, or must they be loaded separately each time?
    >

    Palette settings saved using "Set User Settings" become the default
    settings next time the driver loads. Settings saved using "Save
    Settings..." can be saved to a file for recall later. I have several
    settings saved this way for each of the file types I use.

    >If I understand your instructions correctly, I should take a good
    >slide with range of values and containing black and white areas. I
    >should scan the slide and set the values in the tool pallette to
    >provide good results, save those settings. Then each time I run with
    >that setting, the software will provide the same tweaking as specified
    >in the save settings. Is that essentially correct?
    >

    That's pretty much it.
    --
    Kennedy
    Yes, Socrates himself is particularly missed;
    A lovely little thinker, but a bugger when he's pissed.
    Python Philosophers (replace 'nospam' with 'kennedym' when replying)
     
    Kennedy McEwen, Feb 10, 2004
    #9
  10. john chapman

    john chapman Guest

    Thanks to your suggestions I have been able to substantially improve
    the output from my Nikon Scan scans.

    Vuescan does not seem to have provisions for executing Digital Ice,
    ROC, or GEM (but I have pure software versions of the latter two in
    any case). Although I had sharpening turned off, the Vuescan results
    were noticeably sharper, with more detail, than those from Nikon Scan.
    This difference may be a primary source for my perception that the
    Nikon scans were kind of flat and which triggered my initial question
    in this thread. Test procedures were the same slide at same minimum
    processing through both systems; and then displayed side by side in
    Photoshop, and enlarged multiple times. But even at full screen size,
    the vuescan version was sharper. I assume that vuescan could be doing
    some unrequested hidden post scan processing, but the results on my
    ls4000 were better than with using Nikon Scan.
     
    john chapman, Feb 11, 2004
    #10
  11. john chapman

    HRosita Guest

    Hi,

    I have been disappointed with the Nikon scanner IV. I now have the Minolta
    Elite 5400 that is a bit more expensive but far superior. In addition it can
    scan 4 slides or six negatives at one time and store them as TIFF without
    intervention.
    If the 5400 is too expensive, try the Minolta Scan Elite II.
    Rosita
     
    HRosita, Feb 11, 2004
    #11
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Rafe B.

    Scan Comparison Site, Call for Sample Scans

    Rafe B., Dec 16, 2003, in forum: Digital Photography
    Replies:
    67
    Views:
    1,396
    stacey
    Dec 23, 2003
  2. Adam

    Nikon coolscan 4000 and black / white scans

    Adam, Apr 10, 2006, in forum: Digital Photography
    Replies:
    1
    Views:
    520
    Raphael Bustin
    Apr 10, 2006
  3. geronimo

    How to get good scans of dark slides?

    geronimo, Jan 12, 2007, in forum: Digital Photography
    Replies:
    3
    Views:
    972
    dennis@home
    Jan 12, 2007
  4. jamesnorris

    scans and scan

    jamesnorris, Jan 21, 2007, in forum: Digital Photography
    Replies:
    4
    Views:
    334
    jamesnorris
    Jan 22, 2007
  5. Replies:
    3
    Views:
    856
    Richard Karash
    Aug 9, 2008
Loading...

Share This Page