Vuescan / Nikon Scan and 'latitude'

Vuescan / Nikon Scan and 'latitude'

Post by Ian M Butterfiel » Tue, 15 Oct 2002 17:01:47



I could do with some help with setting up Vuescan (my prefered tool) or
Nikon Scan (I'll use it if I have too) with a Nikon Coolscan IV ED to
capture the full range of exposures that I've captured on my slides.  Here
is the scenario...

I've just returned from Egypt with several rolls of slide film, The bright
sun and deep shadows there have meant that a lot of the slides have
significant dark areas and significant bright areas.  Not surprising as
latitude (the range of light intencities that can be recorded) on slide film
is not that great (only a few stops).  Looking at the images on a lightbox
the slides have recorded all the details I wanted in the shadows AND all the
details I wanted in highlights.  In fact I'm very pleased with the results -
BUT I cannot reproduce what I see on the light box in the scans.  Either I
can capture the detail in the highlight areas (in which case the details in
the shadows are lost) or I can capture the detail in the shadows (in which
case the highlights are burnt out and the detail lost).

So far I've tried in Vuescan setting black points and white points to 0%,
I've tried adjusting the 'brightness' control in vuescan
In Nikon Scan I've tried adjusting the analogue gain.
Also in Nikon Scan I've tried adjusting the curves control.

The results are always the same - I have to choose between having the
shadows or the highlights correctly scanned.  It is very similar to the
problem photographs have when photographing contrasty scenes - you have to
pick, either expose for the shadows or expose for

Any advice or suggestions would be gratefully recieved.
Thanks,
Ian M Butterfield

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Xander Janssen & Anita Israe » Tue, 15 Oct 2002 22:40:27


Ian,

I use a LS4000 and LS8000, and prefer Vuescan over Silverfast and Nikonscan,
although on some slides Nikonscan is doing a better job compared to Vuescan.
As example very darkslides with small bright colored artifacts or so, here
Vuescan sometimes blows away the highlights were nikonscan is just fine. The
best of the two is experience, and after a while easily recognized.

I always look at my slides on a calibrated lightpanel using Schneider
Kreuznach loupes for checking them, a 3* for MF a 6* for 35m and a 10* for
looking into the finest details, checking grain and sharpness, but also for
looking in the darkest shadows. And yes its very hard to get every detail
out of your scan, if you are looking through these kind of loupes at your
slide on a lightpanel!

If you open a scan in Photoshop and you use the brightness slider in the
level tool in photoshop, and move it al the way to the left, making the
picture very bright, it will show you there is actually much more detail in
the dark area's then what you see in the picture with a normal brightness.

How do we get this detail out there without ruining the overall balance, the
answer is "CONTRAST MASKING" in photoshop. This trick is giving stunning
results. And often much faster then fidling around with your
scanner(software) settings. It makes use of a layer which is desaturated and
then inverted, together with an amount of Gaussian Blur and the an optacity
percentage of the desaturated and inverted second layer it will bring out
your details in the dark, the amount is trial and error, but with your
lightpanel beside your computer you can get very nice matches.
See for a nice tutorial this link:
http://www.luminous-landscape.com/tutorials/contrast_masking.shtml

Regards,

Xander Janssen



> I could do with some help with setting up Vuescan (my prefered tool) or
> Nikon Scan (I'll use it if I have too) with a Nikon Coolscan IV ED to
> capture the full range of exposures that I've captured on my slides.  Here
> is the scenario...

> I've just returned from Egypt with several rolls of slide film, The bright
> sun and deep shadows there have meant that a lot of the slides have
> significant dark areas and significant bright areas.  Not surprising as
> latitude (the range of light intencities that can be recorded) on slide
film
> is not that great (only a few stops).  Looking at the images on a lightbox
> the slides have recorded all the details I wanted in the shadows AND all
the
> details I wanted in highlights.  In fact I'm very pleased with the
results -
> BUT I cannot reproduce what I see on the light box in the scans.  Either I
> can capture the detail in the highlight areas (in which case the details
in
> the shadows are lost) or I can capture the detail in the shadows (in which
> case the highlights are burnt out and the detail lost).

> So far I've tried in Vuescan setting black points and white points to 0%,
> I've tried adjusting the 'brightness' control in vuescan
> In Nikon Scan I've tried adjusting the analogue gain.
> Also in Nikon Scan I've tried adjusting the curves control.

> The results are always the same - I have to choose between having the
> shadows or the highlights correctly scanned.  It is very similar to the
> problem photographs have when photographing contrasty scenes - you have to
> pick, either expose for the shadows or expose for

> Any advice or suggestions would be gratefully recieved.
> Thanks,
> Ian M Butterfield



 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Ian M Butterfiel » Wed, 16 Oct 2002 01:16:27


Xander,
Thank you for your response!  You described the situation exactly.  I've not
come across contrast masking before, so I'll have to work through the
tutorial link you've given.

Once again thank you.
Ian



> Ian,

> I use a LS4000 and LS8000, and prefer Vuescan over Silverfast and
Nikonscan,
> although on some slides Nikonscan is doing a better job compared to
Vuescan.
> As example very darkslides with small bright colored artifacts or so, here
> Vuescan sometimes blows away the highlights were nikonscan is just fine.
The
> best of the two is experience, and after a while easily recognized.

> I always look at my slides on a calibrated lightpanel using Schneider
> Kreuznach loupes for checking them, a 3* for MF a 6* for 35m and a 10* for
> looking into the finest details, checking grain and sharpness, but also
for
> looking in the darkest shadows. And yes its very hard to get every detail
> out of your scan, if you are looking through these kind of loupes at your
> slide on a lightpanel!

> If you open a scan in Photoshop and you use the brightness slider in the
> level tool in photoshop, and move it al the way to the left, making the
> picture very bright, it will show you there is actually much more detail
in
> the dark area's then what you see in the picture with a normal brightness.

> How do we get this detail out there without ruining the overall balance,
the
> answer is "CONTRAST MASKING" in photoshop. This trick is giving stunning
> results. And often much faster then fidling around with your
> scanner(software) settings. It makes use of a layer which is desaturated
and
> then inverted, together with an amount of Gaussian Blur and the an
optacity
> percentage of the desaturated and inverted second layer it will bring out
> your details in the dark, the amount is trial and error, but with your
> lightpanel beside your computer you can get very nice matches.
> See for a nice tutorial this link:
> http://www.luminous-landscape.com/tutorials/contrast_masking.shtml

> Regards,

> Xander Janssen



> > I could do with some help with setting up Vuescan (my prefered tool) or
> > Nikon Scan (I'll use it if I have too) with a Nikon Coolscan IV ED to
> > capture the full range of exposures that I've captured on my slides.
Here
> > is the scenario...

> > I've just returned from Egypt with several rolls of slide film, The
bright
> > sun and deep shadows there have meant that a lot of the slides have
> > significant dark areas and significant bright areas.  Not surprising as
> > latitude (the range of light intencities that can be recorded) on slide
> film
> > is not that great (only a few stops).  Looking at the images on a
lightbox
> > the slides have recorded all the details I wanted in the shadows AND all
> the
> > details I wanted in highlights.  In fact I'm very pleased with the
> results -
> > BUT I cannot reproduce what I see on the light box in the scans.  Either
I
> > can capture the detail in the highlight areas (in which case the details
> in
> > the shadows are lost) or I can capture the detail in the shadows (in
which
> > case the highlights are burnt out and the detail lost).

> > So far I've tried in Vuescan setting black points and white points to
0%,
> > I've tried adjusting the 'brightness' control in vuescan
> > In Nikon Scan I've tried adjusting the analogue gain.
> > Also in Nikon Scan I've tried adjusting the curves control.

> > The results are always the same - I have to choose between having the
> > shadows or the highlights correctly scanned.  It is very similar to the
> > problem photographs have when photographing contrasty scenes - you have
to
> > pick, either expose for the shadows or expose for

> > Any advice or suggestions would be gratefully recieved.
> > Thanks,
> > Ian M Butterfield


 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Jeff Randa » Wed, 16 Oct 2002 02:41:00


Ian:

Your eye adapts to the light and dark of your slide.  The CDD's of
your scanner can't.  A potential work-round for this problem is for
you to create two scans, one optimized for the highlights and one for
the shadows and then blend them in your photo-editor.  The specific
blending workflow may be a little different depending on which program
you use.

Hope this helps.

Jeff Randall


> I could do with some help with setting up Vuescan (my prefered tool) or
> Nikon Scan (I'll use it if I have too) with a Nikon Coolscan IV ED to
> capture the full range of exposures that I've captured on my slides.  Here
> is the scenario...

> I've just returned from Egypt with several rolls of slide film, The bright
> sun and deep shadows there have meant that a lot of the slides have
> significant dark areas and significant bright areas.  Not surprising as
> latitude (the range of light intencities that can be recorded) on slide film
> is not that great (only a few stops).  Looking at the images on a lightbox
> the slides have recorded all the details I wanted in the shadows AND all the
> details I wanted in highlights.  In fact I'm very pleased with the results -
> BUT I cannot reproduce what I see on the light box in the scans.  Either I
> can capture the detail in the highlight areas (in which case the details in
> the shadows are lost) or I can capture the detail in the shadows (in which
> case the highlights are burnt out and the detail lost).

> So far I've tried in Vuescan setting black points and white points to 0%,
> I've tried adjusting the 'brightness' control in vuescan
> In Nikon Scan I've tried adjusting the analogue gain.
> Also in Nikon Scan I've tried adjusting the curves control.

> The results are always the same - I have to choose between having the
> shadows or the highlights correctly scanned.  It is very similar to the
> problem photographs have when photographing contrasty scenes - you have to
> pick, either expose for the shadows or expose for

> Any advice or suggestions would be gratefully recieved.
> Thanks,
> Ian M Butterfield


 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Mike Engle » Wed, 16 Oct 2002 03:19:38



> Xander,
> Thank you for your response!  You described the situation exactly.  I've not
> come across contrast masking before, so I'll have to work through the
> tutorial link you've given.

> Once again thank you.
> Ian



> > Ian,

> > I use a LS4000 and LS8000, and prefer Vuescan over Silverfast and
> Nikonscan,
> > although on some slides Nikonscan is doing a better job compared to
> Vuescan.
> > As example very darkslides with small bright colored artifacts or so, here
> > Vuescan sometimes blows away the highlights were nikonscan is just fine.
> The
> > best of the two is experience, and after a while easily recognized.

> > I always look at my slides on a calibrated lightpanel using Schneider
> > Kreuznach loupes for checking them, a 3* for MF a 6* for 35m and a 10* for
> > looking into the finest details, checking grain and sharpness, but also
> for
> > looking in the darkest shadows. And yes its very hard to get every detail
> > out of your scan, if you are looking through these kind of loupes at your
> > slide on a lightpanel!

> > If you open a scan in Photoshop and you use the brightness slider in the
> > level tool in photoshop, and move it al the way to the left, making the
> > picture very bright, it will show you there is actually much more detail
> in
> > the dark area's then what you see in the picture with a normal brightness.

> > How do we get this detail out there without ruining the overall balance,
> the
> > answer is "CONTRAST MASKING" in photoshop. This trick is giving stunning
> > results. And often much faster then fidling around with your
> > scanner(software) settings. It makes use of a layer which is desaturated
> and
> > then inverted, together with an amount of Gaussian Blur and the an
> optacity
> > percentage of the desaturated and inverted second layer it will bring out
> > your details in the dark, the amount is trial and error, but with your
> > lightpanel beside your computer you can get very nice matches.
> > See for a nice tutorial this link:
> > http://www.luminous-landscape.com/tutorials/contrast_masking.shtml

> > Regards,

> > Xander Janssen



> > > I could do with some help with setting up Vuescan (my prefered tool) or
> > > Nikon Scan (I'll use it if I have too) with a Nikon Coolscan IV ED to
> > > capture the full range of exposures that I've captured on my slides.
> Here
> > > is the scenario...

> > > I've just returned from Egypt with several rolls of slide film, The
> bright
> > > sun and deep shadows there have meant that a lot of the slides have
> > > significant dark areas and significant bright areas.  Not surprising as
> > > latitude (the range of light intencities that can be recorded) on slide
> > film
> > > is not that great (only a few stops).  Looking at the images on a
> lightbox
> > > the slides have recorded all the details I wanted in the shadows AND all
> > the
> > > details I wanted in highlights.  In fact I'm very pleased with the
> > results -
> > > BUT I cannot reproduce what I see on the light box in the scans.  Either
> I
> > > can capture the detail in the highlight areas (in which case the details
> > in
> > > the shadows are lost) or I can capture the detail in the shadows (in
> which
> > > case the highlights are burnt out and the detail lost).

> > > So far I've tried in Vuescan setting black points and white points to
> 0%,
> > > I've tried adjusting the 'brightness' control in vuescan
> > > In Nikon Scan I've tried adjusting the analogue gain.
> > > Also in Nikon Scan I've tried adjusting the curves control.

> > > The results are always the same - I have to choose between having the
> > > shadows or the highlights correctly scanned.  It is very similar to the
> > > problem photographs have when photographing contrasty scenes - you have
> to
> > > pick, either expose for the shadows or expose for

> > > Any advice or suggestions would be gratefully recieved.
> > > Thanks,
> > > Ian M Butterfield


Hello

If you are using Photoshop, you can load the luninosity of the immge as
a selection. CTRL ALT SHIFT ~(tilde). Invert this and then use curves or
levels. This will also bring out the mids. A little feathering or
bluring of the selection as a channel helps.

Mike Engles

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Kevi » Wed, 16 Oct 2002 04:02:15


First, determine what the raw scan data is that you are getting.  It's
quite possible that your scanner cannot contain the dynamic range of
your transparencies, regardless of the software you use.  In
NikonScan, use only analogue gain to adjust  the scan and in VueScan
use only the Device/Lock Exposure and then Device/RGB Exposure setting
to adjust the scan.

To reveal what the raw scan data contains, take a scan and use either
a curves tool or brightness setting and move it to extremes.  First
one way to determine if the shadow detail you want is present in the
raw data, then to the other extreme to determine if the highlight
detail you want is in the raw data.  If both ends of the range contain
the detail you want, but you can only see it at contrary extremes of
manipulation, then your problem might be solved by editing in
Photoshop.  You would probably want to try this in 16 bit mode using
selections copied from an 8 bit duplicate to avoid posterization.

If either end of the range is deficient and you cannot get the full
range into a single scan, then try using two scans, one optimizing
shadow detail, the other optimizing highlight detail and merge them in
Photoshop.  Depending on what you start with this can be a lot of
work, but in many cases it can also deliver excellent results.

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Erik Kraus » Wed, 16 Oct 2002 05:05:26


Hi, Ian M Butterfield

Quote:you wrote...
> Looking at the images on a lightbox
> the slides have recorded all the details I wanted in the shadows AND all the
> details I wanted in highlights.  In fact I'm very pleased with the results -
> BUT I cannot reproduce what I see on the light box in the scans.  Either I
> can capture the detail in the highlight areas (in which case the details in
> the shadows are lost) or I can capture the detail in the shadows (in which
> case the highlights are burnt out and the detail lost).

The problem is not the scanner, it is the monitor. There is no way to
reproduce the entire contrast of a slide with natural looking contrast
on a standard monitor. You must compress it. This could be simply done
setting the gamma to other values, but the image will get dull.

I have developed some actions for photoshop 6 that do a sophisticated
contrast masking not showing the usual halos around bright areas. They
let you adjust shadows and highlights almost independent from each
other. Just have a look at http://www.erik-krause.de/contrast

My work flow with Nikon LS40 and Vuescan is as follows: I use the
advanced work flow suggestions for scanning slides and I use long
exposure pass with Color balance set to 'None'. This minimizes shadow
noise and ensures no clipping in highlights and shadows as white and
black point setting does. I save as 48 bit Tiff and treat this images
in Photoshop with my 16 bit contrast masking action.

--
Erik Krause
Digital contrast problems: http://www.erik-krause.de/contrast

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Bart van der Wol » Wed, 16 Oct 2002 04:23:05



Quote:> Ian:

> Your eye adapts to the light and dark of your slide.  The CDD's of
> your scanner can't.  A potential work-round for this problem is for
> you to create two scans, one optimized for the highlights and one for
> the shadows and then blend them in your photo-editor.  The specific
> blending workflow may be a little different depending on which program
> you use.

That's correct.

VueScan does have an option called Long Exposure that tries to do that, but
the blending algorithm can cause artifacts. Part of that problem might have
to do with misalignment between scan passes so, if it's done manually on two
scans, make sure the images are perfectly aligned.

Bart

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Ian M Butterfiel » Wed, 16 Oct 2002 19:20:34


Erik,
Thanks for the reply - as I write this, I've played with contrast masking
(as detailed in Xander's post) and that has given me a huge improvement on
what I've been able to achive previously.  It does however, as you note on
your website, give a halo effect.  I've been able to minimise most of it
with careful choice of how much gausian blur to apply to the mask, and
changing the opacity of the mask layer.  But I'm very interested to try the
HDR compression techniques.

At present I'm using PaintShop Pro (it does most of what PhotoShop can), is
it possible for you to describe the various steps which make up the
photoshop actions that you have created, so that I can try the corresponding
steps in PaintShop Pro?  I am seriously considering 'upgrading' to Photoshop
so if the steps ARE very complex to describe could you send me the PhotoShop
actions, I can at least use them if/when I make the move to Photoshop.

Thanks for the advice (and thanks to everyone else who replied).

Ian M Butterfield

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Tom Harriso » Thu, 17 Oct 2002 02:52:52


Bart --

This is a little off topic, but in my tests with Long Exposure Pass, it
seems that I am getting a longer exposure on the long pass than the maximum
value I can set manually with the Device | Lock exposure option.  Can you
confirm or deny :-)

Thanks -

Tom





> > Ian:

> > Your eye adapts to the light and dark of your slide.  The CDD's of
> > your scanner can't.  A potential work-round for this problem is for
> > you to create two scans, one optimized for the highlights and one for
> > the shadows and then blend them in your photo-editor.  The specific
> > blending workflow may be a little different depending on which program
> > you use.

> That's correct.

> VueScan does have an option called Long Exposure that tries to do that,
but
> the blending algorithm can cause artifacts. Part of that problem might
have
> to do with misalignment between scan passes so, if it's done manually on
two
> scans, make sure the images are perfectly aligned.

> Bart

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Bart van der Wol » Fri, 18 Oct 2002 03:05:57



Quote:> Bart --

> This is a little off topic, but in my tests with Long Exposure Pass, it
> seems that I am getting a longer exposure on the long pass than the
maximum
> value I can set manually with the Device | Lock exposure option.  Can you
> confirm or deny :-)

In an earlier thread (Vuescan Long Exposure Pass : Help needed) Ed Hamrick
answered to the question about the Long exposure time: "It's scanner
dependant - usually between 6x and 12x".

I have noticed that you can enter (type) a value in some option boxes
exceeding the range you can set with the spin/slider control and it
sometimes seems to indeed have an effect (judging by the histogram). I have
to admit I didn't try all possible boundaries of all option boxes.
It would seem logical that the scanner has a hard/firmware limit for some
settings like the exposure time, but it seems less likely that Ed hardcodes
all those differences for all scanner models into VueScan (but it wouldn't
surprise me if he did because he does it for the R to GB exposure ratio for
negatives).

To make a long story short, Preview, try typing a value in the option box,
Preview again and watch the histogram after the second stage refresh. It
probably helps to switch the histogram display to Logarithmic (my default
setting anyway).

Bart

 
 
 

Vuescan / Nikon Scan and 'latitude'

Post by Tom Harriso » Fri, 18 Oct 2002 11:52:50






> > Bart --

> > This is a little off topic, but in my tests with Long Exposure Pass, it
> > seems that I am getting a longer exposure on the long pass than the
> maximum
> > value I can set manually with the Device | Lock exposure option.  Can
you
> > confirm or deny :-)

[snip]
> I have noticed that you can enter (type) a value in some option boxes
> exceeding the range you can set with the spin/slider control and it
> sometimes seems to indeed have an effect

Bart --

I'll give that a try.  Thanks.

Tom

- Show quoted text -

 
 
 

1. Vuescan files smaller than Nikon Scan's from Coolscan IV ED???

Hi,

can anyone explain to me why the files produced by Vuescan (V 7.5.56) are
typically about half the size of those produced by Nikon Scan?

I am scanning the same slide with settings of 48RGB in Vuescan, and in Nikon
Scan a bit depth of 12. My scanner is a Nikon Coolscan IV ED, and I am
scanning at 2900 dpi.

The Vuescan file is 31Mb, and the Nikon Scan file is 59Mb.

I have also noticed that Vuescan files produced from other slides can be as
low as 15Mb at these settings, whereas the NS scans are consistently around
55-60Mb.

Is this down to TIF file compression? If so, can it be turned off in
Vuescan?

Thanks for your help.
Pete

2. counting hot car

3. printing from vuescan in Linux and Vuescan 'help pages'

4. Cookie problem

5. Vuescan can't see my 'supported scanner' (HR6) on Mac osX

6. Oracle Versions

7. VueScan doesn't help mu Nikon LS-1000 in Win XP

8. QUIZ # 2

9. REPOST: VueScan doesn't help mu Nikon LS-1000 in Win XP

10. Help with Nikon Scan 3.1, LS4000, slide feeder and jpeg's

11. Winfax Pro 9 won't allow me to 'scan and send'

12. upgrade Nikon Cool-Scan LS 1000 's new driver

13. "" "" "" " '' '' '' '' '' '' '' ''