Lightroom vs. Apertu Curves

In article 2014081211291587230-savageduck1@REMOVESPAMmecom, Savageduck wrote:

Sandman:
Is the concept of larger dynamic range in RAW totally lost on you?
Is this just a case of you not knowing that the sensor of a given
digital camera can capture more data than your monitor can show?
You seem oblivious to this fact where data beyond an 8-bit scope
is "phantom" data to you.

I record 14-Bit NEFs on my DSLR. I adjust those 14-bit RAW files in
either LR or ACR. After conversion I work in 16-bit mode. The only
time I work in 8-bit is the occasional jpeg. All of my images in LR
are either NEF, CR2, DNG, 16-bit PSD, or 16-bit TIF. My 8-bit JPEGs
are produced via the LR export dialog and exist elsewhere. I don't
have an 8-bit workflow until I export from LR to produce a JPEG
version of what is in my LR catalog.

But that's not the point. In Lightroom the workflow is always 8 bit. Your
monitor can only show you 8 bit of color information from your 14 bit
images.

That's why a histogram has only 255 levels of data. That's why curve points
have X/Y values of 0-255. Because your monitor can't show you any more
data. So regardless of the amount of data in the RAW files, your eyes will
never ever see any more than 8 bits of data when using a monitor. That's
the entire point of this thread.

Sandman:
The data is clearly shown in my images, unquestionably! If you
have questions or there are parts you do not understand, do not be
afraid to ask and I'll try to explain further.

Don't bother, I am not going to be using Aperture.

Huh? I didn't say you would. I meant that I could explain further about the
difference between 14 bits of data and a 8 bit workflow and how you can
combine them.

Sandman:
As I have shown you in two images, there *IS* data to recover and
I recovered it. What part of that do you not understand?

http://sandman.net/files/aperture_curves_normal.png

Unedited 8 bit view of image

http://sandman.net/files/aperture_curves_extended.png

Recovered data from 11 bit image data. See how the images differ?

I see that you adjusted the linear curve on the right down which
brought the highlight levels down in the upper right of the image.

Then you still don't understand it. I didn't adjust it *down* I adjusted
it to the *right*

Here's another image to show the difference:

http://sandman.net/files/normal_vs_extended.png

On the left, you see the normal 0-255 range of 8 bit data and on the left
you see the extended range of data. Note how the histogram in the left
image is shown in a lighter part of the right image. This means that the
point value is *higher* than 255.

Here's another to more clearly illustrate what is happening:

http://sandman.net/files/normal_vs_extended2.png

As you can see, the extended version of curves shows data *outside* the 8
bit color space, and I have moved the white point to include it.

You didn't need curves to achieve that result and what was gained
was useless data. The highlights were blown and nothing you can say
or do will change that.

No, the highlights were blown (more) *before* I did it. This way, the
highlights are *less* blown as the images clearly shows you.

Savageduck:
If there was you should be able to do that with the magic tools
you have available in Sweden, and then present us with this
wonderful image with its extended histogram data restored in all
its glory, but you can't because that data does not exist on
your computer or mine. The best that can be done is band-aid
work.

Sandman:
See above, and you see the recovered data.

You can't quantify that data can you?

What do you mean?

Sandman:
Just for once relax, take a deep breath and note how in the second
image, there is data information in the extended range in the
curves editor, how I've moved the white point to the right of the
normal 8 bit range and it is now covering all of the original 11
bit of data.

That's recovered data.

All you are showing is the clipping indicator, not data, and all I
will concede is that when you moved the linear end point on the
right of the curves histogram down, you reduced the amount of fully
blown data. Most importantly you can make the same adjustment in LR,
ACR, PS anr any other editing software which has curves adjustment
capability, even PSE, PS Touch, Pixelmator, and dare I say it GIMP.

This is a misconception on your part. I did *not* move the linear end point
*down*, I moved it to the *right*, revealing *more* data, not clipping any
data.

Sandman:
And again - if something confuses you or you just didn't know
there was such data, just let me know and I'll try to help
explaining it. Not meant sarcastically or anything. I just assumed
that everyone knew that RAW data had more bits of data than what
your monitor could show, or what could be stored in the JPG
format.

I have been doing this for some time and you are being a tad
presumptuous regarding your idea of my ignorance.

I thought so too, which is why I didn't bring it up from the beginning, I
thought you were fully aware of the fact that as long as you're working on
a computer monitor, all data is 8 bit that you can see.

Sandman:
In fact, if you've ever used Photomatix, or any other HDR tool,
you probably know that they don't need to have several bracketed
images to perform HDR tone mapping, you can do it with just one
RAW file, because the software can tone map the extended range of
sensor data found in that one file.

I have Photomatix and NIK HDR Efex Pro 2, and I am well aware of the
single image tone-mapping capabilities of both

And how they compress the dynamic 14 bit data range in a RAW image into a 8
bit color space - just like Aperture does in this instance? It's the same
thing (only HDR software does lots of other image processing).

I hope my latest screenshots will help you understand what I've been saying
all along.


--
Sandman[.net]

In article , nospam wrote:

Floyd L. Davidson:
Isn't that rather obvious from what I said. "Exposure" can only
be changed with shutter speed and aperture, before the picture
is taken. Which does not stop several software programs from
incorrectly labeling the brightness adjustment as "exposure".

Sandman:
God, you're ignorant.

An "exposure" slider in software is *not* a brightness editor. It
does *not* edit brightness. A true brightness slider edits all
color values uniformly, increasing or decreasing the color value
of every pixel.

This is how the brightness slider have worked in Photoshop since
the dawn of time, until CS5 (I think) where they changed it, but
you can still use it if you click the "Use legacy" checkbox.

it changed in cs3.

Ah, ok.

Sandman:
The new PS brightness slider spreads the spectrum from the bottom,
meaning that darker tones are changed less than brighter tones.

greatly simplifying, brightness is levels and contrast is curves,
with a single easy to use slider adjust for each.

Sure, I'll go with that. As you said - greatly simplified.

Sandman:
An *exposure* slider focuses on the midrange, trying to emulate
(yes, Eric, this is where emulate is the proper word) how the
exposure of the camera works.

actually it focuses (ahem) on the highlights and shadows by
protecting them from clipping. the midrange isn't as critical.

Well, depends on what you mean by focus - I meant focuses its changes on
the midrange, protecting the lows and highs, as you say.

it does emulate what would happen in the camera and is effectively
the same had you changed exposure in camera.

As much as it can be in post, yes.

Using curves, you can edit the whitepoint, which works exactly
like the new brightness editor in PS, meaning it extends the range
beyond the colorspace.

eh? what does that even mean?

It means that data from 0-255 is evenly spread out between 0-300 for
instance when moving the slider to represent 45 steps.

See this image:

http://sandman.net/files/curves_whitepoint.png

First the unaltered histogram, then I've moved the whitepoint to the left,
and the altered histogram evenly spreads out the values from 0-255 to
0-something over 255 and clips at 255.

Hope that made it a bit more clear.

Sandman:
No one wants or use a brightness editor, they're worthless. That's
why exposure and curves white point gives you the control needed.
I much prefer the curves way, as anyone should.

the old style brightness/contrast were broken.

The old style brightness was a true brightness editor, i.e. it edited the
brightness value accordingly for every pixel in the image, uniformly. The
new one acts pretty much like the curves whitepoint, only it tries to
protect the end/low range a bit.




--
Sandman[.net]

In article , Floyd L. Davidson wrote:

Floyd L. Davidson:
The mere concept of having the information presented via a
"curves" tool is what is confusing you. They might well show it
to you in that context, but what they are doing is allowing you
to go back to the RAW converter and change brightness. If you
are aware of that, conceptually, it isn't at all hard to
understand.

Sandman:
No one is trying to "understand" anything here, ignorant Floyd.
You're just posting obvious basic stuff that no one asked about
because you can't read to save your life.

Same old, same old, ey?

Yep, same old ****. It's pretty obvious who does understand it and
who does not.

Indeed it is, ignorant Floyd. You never learn.

--
Sandman[.net]

In article , Rikishi42 wrote:

Eric Stevens:
Before I comment, I would like to try another approach. Can you
post a copy of the original in an editable form?

Sandman:
Certainly.

http://sandman.net/files/DSC01476.ARW

First: thanks for bringing some photo discutions back in this group.

I have a few newbie questions. Not criticicsms, no openings for a
religious debate, just a few questions from someone with more
knowledge in IT than in photo...

You put that image in ARW format, which - as far as I can see - is a
Sony 'raw' format, right?

Correct, from the wonderful Sony A7 camera.

Now, from my IT point of view (opening, interpreting, converting the
file) a 'raw' format is NOT a format. Not in the sense that there is
a clear unique data format definition. Every manufacturor has their
own 'raw' format, some even different ones in different
models/generations of their camera's. Of course, the ARW format
seems to be from Sony (even my Linux tells me so), and probably
doesn't pretend to be universal. But:

Yes, each RAW format is more or less unique, there is no standard format
other than the DNG (digital negative) format from Adobe that is an option
on some cameras, but not from the big consumer brans (sony, nikon, canon).

1. is the ARW format that common that you would assume anyone can
read it, or do you happen to know that he also has a Sony ?

Most modern photo editing software has up to date support for most raw
formats. This version of ARW is about a year old so most
photography-interested people that would be inclined to use it would have
software that supports it.

When a new camera is released, it may use a verion of the RAW format that
is not yet supported. Adobe is usually the quickest to support new formats,
Apple a bit slower.

2. I once saw a definition for a type of an open raw format,
supposedly independant from manufactorors: DNG. It had a few flaws,
but the idea was there. Is there still no tendency to go to such
formats in exchanges, or has everyone just given in and accept that
each brand talks it's own language?

Pretty much. I don't know much about the merits of DNG over proprietary
formats for the manufacturers, but I suppose Nikon wouldn't want to feel
limited by someone else's format when it comes to recording sensor data.

Most camera manufacturers have their own software to read this data, and
some formats have functions that wouldn't be supported by DNG.


--
Sandman[.net]

In article , Floyd L. Davidson wrote:

Floyd L. Davidson:
"All adjustments" are applied *after* the data is demosaiced.
It *cannot* be appled to the raw data.

nospam:
the demosaicing is a step that's automatically included and your
nitpicking does not change anything.

it is *not* a separate step as far as the user is concerned (which
is all that matters).

Sure it isn't.

That's an interesting weasel squirm, but your lack of understanding
shows fairly well when you don't even know what the basic steps are,
or that they even are basic steps!

Every time balloon boy Floyd posts, my irony meter self-destructs.



--
Sandman[.net]

In article 2014081217184870530-savageduck1@REMOVESPAMmecom, Savageduck wrote:

nospam:
virtual copies are a benefit from keeping everything in raw, but
it's not required to use it.

Agreed, but it is easy enough to do and there is no time cost
involved to do so. That is how I do things.

So, what *is* a virtual copy? Is it like versions in Aperture, where you
can select "create new version from original" which creates a new unedited
version in the library that you can edit differently than the one you've
already played around with?

Or is like offline versions, jpg masters that are kept for editing even
when the RAW is unavailable? And when it does get available, the edits use
the RAW instead? Not sure I understand it.



--
Sandman[.net]

On Tue, 12 Aug 2014 13:29:37 -0400, nospam
wrote:

In article , Floyd L. Davidson
wrote:

Here's an example of a feature I'm missing in Lightroom.

I use curves extensively, it's the holy grail of exposure editing.
If you're not using it, start using it!

A curves tool does not edit exposure

Captain literal strikes again.

The fact is that a curves tool does not change
"exposure" (brightness is the correct term). It is a
tone mapping tool.

adjusting exposure is not the same as adjusting brightness.

Isn't that rather obvious from what I said.

no. exposure does not change brightness or contrast. it changes
exposure which is why it's called exposure and not brightness or
contrast.

"Exposure"
can only be changed with shutter speed and aperture,
before the picture is taken.

in an ideal world, it's correct when taken, but that's almost never the
case, which is why can also be adjusted afterwards.

If the original image was underexposed, changing exposure in the
editor won't enable previously lost details to be recovered from the
deep shadows. Nor, if the original image was overexposed will changing
exposure allow details to be recovered from burned out highlights. All
it will do is raise or lower brightness between the two extremes of
bright and dark.

Which does not stop
several software programs from incorrectly labeling the
brightness adjustment as "exposure".

maybe some apps do but not all.

in fact, adjusting brightness or contrast is rarely needed, since
levels & curves do a much better and more effective job.

Curves, as I've noted, simply doesn't do that and
therefore cannot be "more effective".

it does do that, and is more effective because some implementations of
brightness and contrast can clip.

Until you understand that, all you'll do is make silly
statements such as you did for the rest of your post.

take your own advice.

you don't use the software being discussed, which would be step one.

I suppose for someone who doesn't understand it that would seem to
be true. It isn't.

so far, what you've said does not apply to the software he's using.
--

Regards,

Eric Stevens

On 12 Aug 2014 18:03:05 GMT, Sandman wrote:

In article , Floyd L. Davidson wrote:

Floyd L. Davidson:
The fact is that a curves tool does not change "exposure"
(brightness is the correct term). It is a tone mapping tool.

nospam:
adjusting exposure is not the same as adjusting brightness.

Isn't that rather obvious from what I said. "Exposure" can only be
changed with shutter speed and aperture, before the picture is
taken. Which does not stop several software programs from
incorrectly labeling the brightness adjustment as "exposure".

God, you're ignorant.

An "exposure" slider in software is *not* a brightness editor. It does
*not* edit brightness. A true brightness slider edits all color values
uniformly, increasing or decreasing the color value of every pixel.

This is how the brightness slider have worked in Photoshop since the dawn
of time, until CS5 (I think) where they changed it, but you can still use
it if you click the "Use legacy" checkbox.

The new PS brightness slider spreads the spectrum from the bottom, meaning
that darker tones are changed less than brighter tones.

An *exposure* slider focuses on the midrange, trying to emulate (yes, Eric,
this is where emulate is the proper word) how the exposure of the camera
works.

With the base values of rgb(0, 30, 250), these are the results:

True brightness +10: rgb(10, 40, 255)
New PS brightness +10: rgb(3, 36, 255)
Exposure +10: rgb(3, 40, 253)

Using curves, you can edit the whitepoint, which works exactly like the new
brightness editor in PS, meaning it extends the range beyond the
colorspace.

No one wants or use a brightness editor, they're worthless. That's why
exposure and curves white point gives you the control needed. I much prefer
the curves way, as anyone should.

nospam:
in fact, adjusting brightness or contrast is rarely needed, since
levels & curves do a much better and more effective job.

Curves, as I've noted, simply doesn't do that and therefore cannot
be "more effective".

It does, you just don't know anything about these matters.

Floyd L. Davidson:
Until you understand that, all you'll do is make silly
statements such as you did for the rest of your post.

nospam:
take your own advice.

you don't use the software being discussed, which would be step
one.

I suppose for someone who doesn't understand it that would seem to
be true. It isn't.

Ironic.

So what does the 'exposure' control do to the image?
--

Regards,

Eric Stevens

On Wed, 13 Aug 2014 21:10:50 +1200, Eric Stevens
wrote:

On 12 Aug 2014 18:03:05 GMT, Sandman wrote:

In article , Floyd L. Davidson wrote:

Floyd L. Davidson:
The fact is that a curves tool does not change "exposure"
(brightness is the correct term). It is a tone mapping tool.

nospam:
adjusting exposure is not the same as adjusting brightness.

Isn't that rather obvious from what I said. "Exposure" can only be
changed with shutter speed and aperture, before the picture is
taken. Which does not stop several software programs from
incorrectly labeling the brightness adjustment as "exposure".

God, you're ignorant.

An "exposure" slider in software is *not* a brightness editor. It does
*not* edit brightness. A true brightness slider edits all color values
uniformly, increasing or decreasing the color value of every pixel.

This is how the brightness slider have worked in Photoshop since the dawn
of time, until CS5 (I think) where they changed it, but you can still use
it if you click the "Use legacy" checkbox.

The new PS brightness slider spreads the spectrum from the bottom, meaning
that darker tones are changed less than brighter tones.

An *exposure* slider focuses on the midrange, trying to emulate (yes, Eric,
this is where emulate is the proper word) how the exposure of the camera
works.

With the base values of rgb(0, 30, 250), these are the results:

True brightness +10: rgb(10, 40, 255)
New PS brightness +10: rgb(3, 36, 255)
Exposure +10: rgb(3, 40, 253)

Using curves, you can edit the whitepoint, which works exactly like the new
brightness editor in PS, meaning it extends the range beyond the
colorspace.

No one wants or use a brightness editor, they're worthless. That's why
exposure and curves white point gives you the control needed. I much prefer
the curves way, as anyone should.

nospam:
in fact, adjusting brightness or contrast is rarely needed, since
levels & curves do a much better and more effective job.

Curves, as I've noted, simply doesn't do that and therefore cannot
be "more effective".

It does, you just don't know anything about these matters.

Floyd L. Davidson:
Until you understand that, all you'll do is make silly
statements such as you did for the rest of your post.

nospam:
take your own advice.

you don't use the software being discussed, which would be step
one.

I suppose for someone who doesn't understand it that would seem to
be true. It isn't.

Ironic.

So what does the 'exposure' control do to the image?

Ignore that question. I see that the answer is already in the article.
--

Regards,

Eric Stevens

On Tue, 12 Aug 2014 16:48:02 -0400, nospam
wrote:

In article , Floyd L. Davidson
wrote:


Floyd L. Davidson:
Until you understand that, all you'll do is make silly
statements such as you did for the rest of your post.

nospam:
take your own advice.

you don't use the software being discussed, which would be step
one.

I suppose for someone who doesn't understand it that would seem to
be true. It isn't.

Ironic.

isn't it?

You two can call it whatever you like, all of it is "brightness" and
not a bit of it is exposure.

i'm calling what adobe calls it, and that's exposure. it is *not* the
same as brightness.

Exposure is how many photons are captured
by the sensor...

more of your weaseling.

Floyd is quite correct.

nobody expects to go back and change the f/stop or shutter speed.

changing exposure in lightroom looks the same as if the exposure in the
camera was changed by the same amount.

It can't, for the simple reason that it can't recover detail in the
extremes. Changing real exposure would recover detail at one end or
the other.

nothing is perfect, but nobody is going to notice a difference by
looking at the results unless it's extreme (and even then, probably
not).
--

Regards,

Eric Stevens