D100 Underexposure solved here

[alfred_schiele]

Hi

So, if you have no problem with your exposures, don't use my curve.

The "problem" I evaluated is more subtile. If I meter a gray
subject I expect it to be in the center of the histogram

WHAT , your smoking crack, just because you meter grey doesnt mean
the histrogram is in the centre,what if 2/3 of the subject is black
??,
think your theory again

There is nothing wrong with my view on metering.

Maybe you understood wrong and I have could worded better.

I expect when metering a reference gray to produce something like a bell curve with the center ideally in the middle of the histogram.

There are two parts, the metering itself (no matrix metering here please) that refers to the ideal gray and the subject that is metered. If the metered subject is dark - as a result to the metering standard - it should still come out gray... refering to uncorrected metering. If you want something different for darker or brighter subjects you have to compensate EV then. If you do so, right, the historgram changes then. And: the historgramm tells of course more than the graypoint does...

If the gray metered subject does not produce a gray image there are two ways to go:

1. EV compensation. This is somehow inconvenient just to shift the gray point to where it should be, at ZERO EV...

2. A custom curve. More convenient as the graypoint standard can be kept close to EV comp. ZERO. But on the other hand it introduces some nonlinear behaviour that can be positive or negative depending on the curve shape used and your need, preference or taste.

In the meantime a downloaded the curves and analyzed them. The one that is called ev0_... has a confusing name as it does shift the graypint up a good deal as well. Nevertheless the curves are interesting as they reduce contrast in the highlights and give a bit better detail visibility there.

All these curves seem better to me than the predefined N* standards.

Regards, A. Schiele

 

[alfred_schiele]

Hi

I am not sure what you want to show, but concerning my experiences
with my D100 your "uncorrected" shots with your D100 look pretty ok.

The corrections you show are detail nuances but not what I
experienced from my D100 concerning exposure or even "problems"
with exposure.

How is you methering method? Where is your garypoint in the
historgram when you shoot a gray card??

After downloading your curves and having a closer look to them I understand better... but ev0_.. sounded a bit confusing for a curve that pushes up the graypoint a good deal as well...

Regards, A. Schiele.

 

[Zorpie]

Thom Hogan's observations may be right in that there is no 'ideal' curve for every situation - but then arguably - is the linear (default) setting - then ideal for every situation ?? Obviously - it isn't. Else we wouldn't be having these discussions. Those who 'leave things as they are' and swear that there is 'no problem' are then to be found deep in post-processing. This is surely - the same thing - just another way of achieving the same results. But there are no doubt those - who will disagree with me !!

So, if you have no problem with your exposures, don't use my curve.

The "problem" I evaluated is more subtile. If I meter a gray
subject I expect it to be in the center of the histogram but this
does not happen. To compensate this I have to either correct EV
and/or use a custom curve either in camera or in NC3 depending on
the circumstance. Matrix metering is no good reference to check
exposure problems.

I don't belive cameras from different production charges really
differ. For several reasons I have replaced my fisrt D100 and my
second is absolutely the same concerning exposure...

Contrast is another thing that can differ from shot to shot and
this is not corrected in camera as well. In this case a single
custom curve alone does not help. Even more subtile, contrast
differs with used colourspace too. I was somehow surprised to see
in NC3 that using my custom curve on shots with ColourSpace
produced slightly differnt contrast than Colourspace III. Even
changing colourspace in NC3 later didn't change this effect of the
initial colourspace selection during shooting.

I sometimes think the in camera and NC3 workflow has some magic
differences even when the same curves or settings are used...

Regards, A. Schiele

 

[Apolinario]

I deleted the +0 EV curve and all others except for the the +0.3 and +0.7 curves over the weekend. I also added a new curve, the +0.5 EV curve.

My goal in creating this new curve (like my +0.3 curve) was to make neutral corrections to EV without any color/contrast/saturation changes. I also wanted to retain softer highlight blowouts. My new curve mimics a +0.5 EV adjustment without having to set +0.5 and it retains highlights a lot better.

When I originally made my +0 EV curve, I think I was trying to increase saturation and other color changes which in fact did increase graypoint but kept blackpoint and whitepoint the same. I now believe those types of changes should be done in Photoshop instead. Since then I have concluded that the only way for a custom tone curve to be useful for general use (which is all it can be useful for unless the D100 is tethered) is if it made neutral EV corrections and retained highlights better.

I made this curve because after using my +0.3 curve over the weekend, I still found myself disappointed with the brightness of photographs. With my +0.7 curve, it was just slightly too bright. My new +0.5 curve meets those two in the middle and produces photos that are so far perfect in my eyes.

I'll try and upload samples later.

I am not sure what you want to show, but concerning my experiences
with my D100 your "uncorrected" shots with your D100 look pretty ok.

The corrections you show are detail nuances but not what I
experienced from my D100 concerning exposure or even "problems"
with exposure.

How is you methering method? Where is your garypoint in the
historgram when you shoot a gray card??

After downloading your curves and having a closer look to them I
understand better... but ev0_.. sounded a bit confusing for a curve
that pushes up the graypoint a good deal as well...

Regards, A. Schiele.

--
You cannot perceive beauty but with a serene mind

 

[Apolinario]

I think there can be a curve ideal for every situation, but only if it corrects some problem with underexposure neutrally, without any color/saturation changes. This is what I have been trying to do with my two latest curves.

If the curve changes anything else other than brightness, then it will not work for every situation.

So, if you have no problem with your exposures, don't use my curve.

The "problem" I evaluated is more subtile. If I meter a gray
subject I expect it to be in the center of the histogram but this
does not happen. To compensate this I have to either correct EV
and/or use a custom curve either in camera or in NC3 depending on
the circumstance. Matrix metering is no good reference to check
exposure problems.

I don't belive cameras from different production charges really
differ. For several reasons I have replaced my fisrt D100 and my
second is absolutely the same concerning exposure...

Contrast is another thing that can differ from shot to shot and
this is not corrected in camera as well. In this case a single
custom curve alone does not help. Even more subtile, contrast
differs with used colourspace too. I was somehow surprised to see
in NC3 that using my custom curve on shots with ColourSpace
produced slightly differnt contrast than Colourspace III. Even
changing colourspace in NC3 later didn't change this effect of the
initial colourspace selection during shooting.

I sometimes think the in camera and NC3 workflow has some magic
differences even when the same curves or settings are used...

Regards, A. Schiele

--
You cannot perceive beauty but with a serene mind

 

[Octopuzz]

Apolinario,

Congratulations on your quest. I applaud your efforts and eperimentation. I strongly believe that something great will come about in the end. By the way, have you assembled a panel of color experts to judge the results of these curve adjustments?

Waiting for your samples.

My goal in creating this new curve (like my +0.3 curve) was to make
neutral corrections to EV without any color/contrast/saturation
changes. I also wanted to retain softer highlight blowouts. My
new curve mimics a +0.5 EV adjustment without having to set +0.5
and it retains highlights a lot better.

When I originally made my +0 EV curve, I think I was trying to
increase saturation and other color changes which in fact did
increase graypoint but kept blackpoint and whitepoint the same. I
now believe those types of changes should be done in Photoshop
instead. Since then I have concluded that the only way for a
custom tone curve to be useful for general use (which is all it can
be useful for unless the D100 is tethered) is if it made neutral EV
corrections and retained highlights better.

I made this curve because after using my +0.3 curve over the
weekend, I still found myself disappointed with the brightness of
photographs. With my +0.7 curve, it was just slightly too bright.
My new +0.5 curve meets those two in the middle and produces photos
that are so far perfect in my eyes.

I'll try and upload samples later.

I am not sure what you want to show, but concerning my experiences
with my D100 your "uncorrected" shots with your D100 look pretty ok.

The corrections you show are detail nuances but not what I
experienced from my D100 concerning exposure or even "problems"
with exposure.

How is you methering method? Where is your garypoint in the
historgram when you shoot a gray card??

After downloading your curves and having a closer look to them I
understand better... but ev0_.. sounded a bit confusing for a curve
that pushes up the graypoint a good deal as well...

Regards, A. Schiele.

--
You cannot perceive beauty but with a serene mind

 

[Apolinario]

I assembled a panel of 9 world-renowned color experts at my house and they unanimously concluded these curve adjustments give no significant color shift, with a 99.5% level of confidence.

Congratulations on your quest. I applaud your efforts and
eperimentation. I strongly believe that something great will come
about in the end. By the way, have you assembled a panel of color
experts to judge the results of these curve adjustments?

Waiting for your samples.

My goal in creating this new curve (like my +0.3 curve) was to make
neutral corrections to EV without any color/contrast/saturation
changes. I also wanted to retain softer highlight blowouts. My
new curve mimics a +0.5 EV adjustment without having to set +0.5
and it retains highlights a lot better.

When I originally made my +0 EV curve, I think I was trying to
increase saturation and other color changes which in fact did
increase graypoint but kept blackpoint and whitepoint the same. I
now believe those types of changes should be done in Photoshop
instead. Since then I have concluded that the only way for a
custom tone curve to be useful for general use (which is all it can
be useful for unless the D100 is tethered) is if it made neutral EV
corrections and retained highlights better.

I made this curve because after using my +0.3 curve over the
weekend, I still found myself disappointed with the brightness of
photographs. With my +0.7 curve, it was just slightly too bright.
My new +0.5 curve meets those two in the middle and produces photos
that are so far perfect in my eyes.

I'll try and upload samples later.

I am not sure what you want to show, but concerning my experiences
with my D100 your "uncorrected" shots with your D100 look pretty ok.

The corrections you show are detail nuances but not what I
experienced from my D100 concerning exposure or even "problems"
with exposure.

How is you methering method? Where is your garypoint in the
historgram when you shoot a gray card??

After downloading your curves and having a closer look to them I
understand better... but ev0_.. sounded a bit confusing for a curve
that pushes up the graypoint a good deal as well...

Regards, A. Schiele.

--
You cannot perceive beauty but with a serene mind

--
You cannot perceive beauty but with a serene mind

 

[Octopuzz]

Apolinario,

Could you disclose who these experts are?

Also have you run a statistical analysis to derive at the 99.5% accuracy?

Why hasn't Nikon come out yet to officially congratulate you on achieving something that they couldn't acheive in the first place?

Congratulations on your quest. I applaud your efforts and
eperimentation. I strongly believe that something great will come
about in the end. By the way, have you assembled a panel of color
experts to judge the results of these curve adjustments?

Waiting for your samples.

My goal in creating this new curve (like my +0.3 curve) was to make
neutral corrections to EV without any color/contrast/saturation
changes. I also wanted to retain softer highlight blowouts. My
new curve mimics a +0.5 EV adjustment without having to set +0.5
and it retains highlights a lot better.

When I originally made my +0 EV curve, I think I was trying to
increase saturation and other color changes which in fact did
increase graypoint but kept blackpoint and whitepoint the same. I
now believe those types of changes should be done in Photoshop
instead. Since then I have concluded that the only way for a
custom tone curve to be useful for general use (which is all it can
be useful for unless the D100 is tethered) is if it made neutral EV
corrections and retained highlights better.

I made this curve because after using my +0.3 curve over the
weekend, I still found myself disappointed with the brightness of
photographs. With my +0.7 curve, it was just slightly too bright.
My new +0.5 curve meets those two in the middle and produces photos
that are so far perfect in my eyes.

I'll try and upload samples later.

I am not sure what you want to show, but concerning my experiences
with my D100 your "uncorrected" shots with your D100 look pretty ok.

The corrections you show are detail nuances but not what I
experienced from my D100 concerning exposure or even "problems"
with exposure.

How is you methering method? Where is your garypoint in the
historgram when you shoot a gray card??

After downloading your curves and having a closer look to them I
understand better... but ev0_.. sounded a bit confusing for a curve
that pushes up the graypoint a good deal as well...

Regards, A. Schiele.

--
You cannot perceive beauty but with a serene mind

--
You cannot perceive beauty but with a serene mind

 

[alfred_schiele]

Hi

Thank you for commenting the latest state of your curve experiments.

With me personally I think I will do curve corrections with NEFs again.

I found that differences in exposure between using different nikkors can not be resolved with any curve...

I will not put much more energy in this also I was inspired by your idea to decrease contrast in the higher tones only and maintaining shadow contrast somehow.

Regards, A. Schiele

My goal in creating this new curve (like my +0.3 curve) was to make
neutral corrections to EV without any color/contrast/saturation
changes. I also wanted to retain softer highlight blowouts. My
new curve mimics a +0.5 EV adjustment without having to set +0.5
and it retains highlights a lot better.

When I originally made my +0 EV curve, I think I was trying to
increase saturation and other color changes which in fact did
increase graypoint but kept blackpoint and whitepoint the same. I
now believe those types of changes should be done in Photoshop
instead. Since then I have concluded that the only way for a
custom tone curve to be useful for general use (which is all it can
be useful for unless the D100 is tethered) is if it made neutral EV
corrections and retained highlights better.

I made this curve because after using my +0.3 curve over the
weekend, I still found myself disappointed with the brightness of
photographs. With my +0.7 curve, it was just slightly too bright.
My new +0.5 curve meets those two in the middle and produces photos
that are so far perfect in my eyes.

I'll try and upload samples later.

I am not sure what you want to show, but concerning my experiences
with my D100 your "uncorrected" shots with your D100 look pretty ok.

The corrections you show are detail nuances but not what I
experienced from my D100 concerning exposure or even "problems"
with exposure.

How is you methering method? Where is your garypoint in the
historgram when you shoot a gray card??

After downloading your curves and having a closer look to them I
understand better... but ev0_.. sounded a bit confusing for a curve
that pushes up the graypoint a good deal as well...

Regards, A. Schiele.

--
You cannot perceive beauty but with a serene mind

 

[alfred_schiele]

Hi

I think there can be a curve ideal for every situation, but only if
it corrects some problem with underexposure neutrally, without any
color/saturation changes. This is what I have been trying to do
with my two latest curves.

If the curve changes anything else other than brightness, then it
will not work for every situation.

I think as well one custom curve is a comprimise that can be acceptable for lots of situations.

Some printer drivers have auto adjustment options for digital camera, softer or harder tones as well and if the basic images is a bit less contrasty these automatisms work fine while they don't with material that is allready too contrasty from the bottom.

Sure, for fine art one curve isn't ideal but D100 has still a lot left even for the average everyday shots... unfortunately...

Regards, A. Schiele