Vuescan - histogram clips at black point

[Matthew Rex]

Hi,

I'm using Vuescan, Nikon LS-IV, scanning slide film. I have ordered
IT8.7 targets from Wolf Faust but while I'm waiting for them to arrive I
hope someone can help/explain this problem to me:

On most scans if I look at the histogram in Vuescan, I see the top one
showing no clipping, yet with the black and white points set to 0, I get
clipping - almost always red on the bottom histogram.

The reds in my scans just look weird - just blotches of the same shade,
with no change in tone.

I have experimented with the different media and colour balance settings
but I can't get the black points to stop clipping.

What I *think* I want is to set media to "image" and color balance to
"manual" or "none". Certainly setting color to "neutral" as per the
manual is way too dark.

I have just previewd a scan that is mostly red with some green (a flower
closeup) and would you know it? the green and blue channels are clipped!
Help!

I use the manual's "advanced workflow" and am fairly comfortable with
it. I plan to profile the scanner ASAP - is this likely to help? (I
assume that once I profile I can rescan from raw files with out having
to re-scan the actual slides.)

Am I missing something?

Thanks,
Matt.

 

[Bart van der Wolf]

"Matthew Rex" <"mrex iinet net au"> wrote in message
SNIP

On most scans if I look at the histogram in Vuescan, I see the top one
showing no clipping, yet with the black and white points set to 0, I get
clipping - almost always red on the bottom histogram.

I assume your crop area excludes frame edges.

If the top histogram (of the Raw data) is not clipped and close to
clipping to the right hand side, then the exposure is correct. Without
Black/White point clipping set (both are 0%), the bottom histogram
should not clip either, depending only on your profile choices and
Color balancing settings.

Could you elaborate on your Color Tab settings (a crop of a screen
dump on some website would help, and being able to see the histograms
is often an eye opener)?

Bart

 

[Matthew Rex]

"Matthew Rex" <"mrex iinet net au"> wrote in message
SNIP



I assume your crop area excludes frame edges.

I usually scan at maximum crop but for this discussion my test uses a
crop within the frame border.

If the top histogram (of the Raw data) is not clipped and close to
clipping to the right hand side, then the exposure is correct. Without
Black/White point clipping set (both are 0%), the bottom histogram
should not clip either, depending only on your profile choices and
Color balancing settings.

Hmmm, the raw histogram is not clipped - but the "black point triangle"
is set well to the right of where the histogram shows is the black
point. To put it another way - looking at the raw histogram I would like
to move the black point futher to the left but I can't move the
triangle! The black point and white point are set to 0 in the color tab.

So it appears that the output histogram is "correct" based on where the
black/white point triangles are. So perhaps my question should be "Why
can't I set black point to the left-most point on the histogram?"

Could you elaborate on your Color Tab settings (a crop of a screen
dump on some website would help, and being able to see the histograms
is often an eye opener)?

Here's two screenshots:
http://members.iinet.net.au/~mrex/test/histo.jpg
http://members.iinet.net.au/~mrex/test/histo2.jpg

Media type set to "image".

Bart

Thanks,
Matt.

 

[Robert Feinman]

Here's two screenshots:
http://members.iinet.net.au/~mrex/test/histo.jpg
http://members.iinet.net.au/~mrex/test/histo2.jpg

Media type set to "image".

If you lock the image color you will get access to the RGB triangles.
You can then try to move each separately.

 

[Matthew Rex]

Here's two screenshots:
http://members.iinet.net.au/~mrex/test/histo.jpg
http://members.iinet.net.au/~mrex/test/histo2.jpg

Media type set to "image".

If you lock the image color you will get access to the RGB triangles.
You can then try to move each separately.
[/QUOTE]
OK, that moves the RGB triangles to the far left of the raw histogram.
The output histogram is still clipped though :-(

 

[Bart van der Wolf]

Robert Feinman wrote: SNIP
OK, that moves the RGB triangles to the far left of the raw histogram.
The output histogram is still clipped though :-(

Then you may need to increase exposure (after locking it, and doing a
new Preview), or try a different output profile. I also assume you
scan for 48/64 bit images.

Bart

 

[Mendel Leisk]

If you lock the image color you will get access to the RGB triangles.
You can then try to move each separately.

OK, that moves the RGB triangles to the far left of the raw histogram.
The output histogram is still clipped though :-([/QUOTE]

Try Input|Lock film base color

Then in the Color tab, try setting the 3 new fields each to "1"
(regardless of what they start with). If black is still clipping, try
"1.1", "1.2", etc. This should push the left end of the the "before"
histogram to the right.

 

[Matthew Rex]

Then you may need to increase exposure (after locking it, and doing a
new Preview), or try a different output profile. I also assume you
scan for 48/64 bit images.

Bart

Bart, I re-read your earlier post and I think I understand better now.
Exposure settings do seem to be the issue.

Using this as reference:
http://members.iinet.net.au/~mrex/test/histo2.jpg

Basically you are saying that the exposure is good, but that due to the
output colour space there is still some clipping. One way to fix that
might be to increase exposure in the original scan - but colour space
restrictions might still result in a clipped output histogram.

It seems to me that the "Advanced Workflow" won't always give the best
results. I had been blindly using it to ensure no clipping of
highlights, sometimes to the detriment of shadow detail.

I was also setting "exposure clipping" to 0% which is probably a mistake
for all but the lightest scans.

scan for 48/64 bit images.

Yes, I scan 64bit raw. I'll wait for my IT8.7 target to arrive and then
have another go at scanning some slides paying more attention to
exposure settings.

Thanks to the group for all comments.

Matt.

 

[ThomasH]

I usually scan at maximum crop but for this discussion my test uses a
crop within the frame border.


Hmmm, the raw histogram is not clipped - but the "black point triangle"
is set well to the right of where the histogram shows is the black
point. To put it another way - looking at the raw histogram I would like
to move the black point futher to the left but I can't move the
triangle! The black point and white point are set to 0 in the color tab.

So it appears that the output histogram is "correct" based on where the
black/white point triangles are. So perhaps my question should be "Why
can't I set black point to the left-most point on the histogram?"


Here's two screenshots:
http://members.iinet.net.au/~mrex/test/histo.jpg
http://members.iinet.net.au/~mrex/test/histo2.jpg

I think that this the same question which I asked a while ago
with a similar example of a white point = 0% and blue component
outside of the WP marker.

The black and white point markers stand at the location of
the min and max luminance when they are set to zero. That's
why red component can be "smaller than 0% BP" and blue
component can be "larger than 0% WP".

Thomas