Color Curves Whites/Blacks

>I literally ran out of ideas for threads: the thread

>>2722847
depends on the material, you moron. In your pic the curve raises contrast, which makes sense, cause the histogram shows strong mids. but what when the material already shows deep blacks and much lights? do you think that curve makes sense then? no? fucker? think.

>>2722847
>is it common
It depends. Some people find it dreamy and/or retro-ish. It can look quite nice if done right but on the other hand, it can fuck up the look of a potentially good photo, too.

Yes it is popular on tumblr for those who want to emulate vsco. IMO it looks like garbage compared to vsco but oh well.

>>2722847
kek

that is the S-curve you get when you plot a film density: this means that is the most common curve you'll see when processing images

if your image is bland (which usually is when you shoot digital correctly), that is the curve you firstly apply and correct in the processing

>>2722852
i have no idea what you're saying sorry

also, that is the curve why "curves" are called "curves" instead of "densities", which would be somewhat more correct

>>2722858
He's saying that s-curve is not applicable in every instance, example being an image already high in contrast

File: Hipster curve template.jpg (51KB, 214x246px) Image search: [Google]

51KB, 214x246px

I wish I knew how the fuck to edit curves.

[EXIF data available. Click here to show/hide.]

Camera-Specific Properties:
Camera Software	Adobe Photoshop CS5.1 Windows
Image-Specific Properties:
Image Orientation	Top, Left-Hand
Horizontal Resolution	72 dpi
Vertical Resolution	72 dpi
Image Created	2012:12:15 21:04:23
Color Space Information	sRGB
Image Width	214
Image Height	246

>>2722852
In the pic, the curve does not increase contrast at all. Highs are reigned in, lows are raised, and mids are preserved. An S curve where the mids are adjusted and the whites and blacks stay anchored increases contrast.

>>2722855
This curve is one of the steps involved in creating a VSCO preset. It isn't one of the other, and you can't have the VSCO fade without raising the black point, which is what this curve is doing.

>>2722857
correct, but not in this way.

>>2722850
I've seen this exact thread many times desu

>>2722857
look closer

File: INyDjkz.jpg (52KB, 480x270px) Image search: [Google]

52KB, 480x270px

>>2722852

Any good resources for learning how to use curves? I never touch it in LR as I don't know what I'm doing

>>2722852
>the curve raises contrast

No it doesn't.

It keeps the midtones the same.
It REDUCES contrast in the blacks, shadows, highlights and whites.

It's basically the "my screen has too much dynamic range, let's fuck it up" curve.

>>2723072
Lynda.com
A whole pants-load of YouTube videos.

What are the best VSCO presets, ive downloaded a ton and I literally have no idea which ones are good

>>2723152

Give up now, retard.

>>2723152
What size of socket is best? They're tools, and all have different uses, to be used by an artist to create the look they want. There is no "best".

Photography is not a set of instructions, like lego. You don't buy the only right lens, put it on the only right camera, pick the only correct composition, and select the best preset. It's decisions and priorities of an artist all chosen for a particular photo.

>>2723154
Y

>>2723078
In fairness, with the exception of the raised blacks, it does raise contrast. If you want to be anal, you're raising contrast and reducing dynamic range, which is possible, because both are relative to the untouched image and not necessarily absolute.
>doesn't touch mid tones
Actually, any contrast curve does, just slightly. The "s" curve is a contrast increasing curve.

>>2723156
yeh thats bullshit brah you know for a some there is some good ones in there

stop being artsy and pretentious

>>2723158
>The "s" curve is a contrast increasing curve.
Only when it's applied in a specific way. Look at the curve, dude. The spot between the two mid anchors is very nearly on the center line. There's next to no adjustment being made, contrast wise. And outside of those anchors, the highlights are being pulled in, and the shadows are being pulled in. Contrast is lowered, in the OP curve. To increase contrast, highlights are raised, and shadows are lowered. That is not occurring in any noticeably way in the OP curve. Stop stating what a book told you one time and use your eyes to see what's actually being shown to you.

>>2723160
>you know for a some there is some good ones in there
What?
Whatever you are failing to say in English, what preset is best will depend entirely on the look you are going for. There is no "best" because there is no universal goal to be achieved. You wouldn't go into a hardware store and say "Which tool is best? I want to buy the best tool". The guy would look at you like you were stupid, and then his question would be "best for what ?What project are you doing?"

"Oh, nothing, I just want to buy the best tool now so when something comes up, I can have the best tool. So, like, this big hammer? or... maybe that planer? Which tool is the best tool in here?"

File: level_curve_ex_contrast_pane.png (3KB, 262x262px) Image search: [Google]

3KB, 262x262px

>>2723158
>The "s" curve is a contrast increasing curve.
That's not how it works.

The dotted line indicated the original contrast.

Where the curve is steeper than the dotted line (ie: Δy/Δx > 1) you get more contrast.
Where it's less steep (ie: Δy/Δx < 1) you get less contrast.
In OP's example the curve never gets any steeper than 1. (well, perhaps a tiny little bit in the middle but nothing noticeable)

In this example you do get more contrast in the mid tones, because the curve is steeper in the middle.

>>2723164
>barely
Yeah, it's not strong, but it's there.

And again, contrast is increased, dynamic range is decreased. This is the same as if one were to not touch the black point and white point, but did dick with the output levels.

Look, when speaking about contrast, it's those two middle anchors that matter. The difference between them is increased, albeit slightly. This is increased contrast. Like say we have two values of 20 and 30. We make that 20 into 17 and then chop off all values under 5 and over 40. The range between 17 and 30 remains greater than 20 and 30 irrespective of what happens outside of those numbers. That's what's happening here.

>>2722971
kek
now I noticed that he's clipping on the blacks

that's plain dumb

>>2723168
If you have smooth curves turned on, put an anchor on the left side of the histogram and pull it below the 1:1 line, you will increase the slope, period. If you put an anchor in the right have and pull up any amount, you'll likewise increase the slope, period.

The only time an s curve does not increase contrast is if you have smoothing turned off and do some crazy rotated shit that isn't an s curve anyway.

File: 2015-12-16 11_40_44-_MG_6538.jpg @ 100% (Curves 1, Layer Mask_8) _.jpg (279KB, 2041x1046px) Image search: [Google]

279KB, 2041x1046px

>>2723171
Look how much I increased my contrast you guy! This guy is definitely right, and knows exactly what he's talking about! He read a book you guys!

[EXIF data available. Click here to show/hide.]

Camera-Specific Properties:
Camera Software	Greenshot
Image-Specific Properties:

>>2723192
Being technically right and pedantic on the internet is much more important than actual results, real world usage, or communicating useful ideas to people who may otherwise not understand them, therefore, you're still wrong. Thanks for playing, and be sure to pick up the /p/ home game! See you next week!

>>2723192
>>2723195
That the same curve is a contrast curve is a mathematical fact. Telling anyone anything different is like telling them that 1+1=3.

Usefulness doesn't come from half truths and outright lies, it comes from actually understanding how things like dynamic range and contrast interact. It comes from understanding how to use tools to get the effect you want which is a result of knowing how the magnitude of change affects the image.

This isn't about winning Internet points by winning an argument, it's about getting people real information and not some ignorant ass idiocy.

>>2723167
if i went to a hardstore and said which tool is the best they woud point me to the best tool, faggot

>>2723214
So you're saying that, no matter what, that an S curve increases contrast, and that in the image posted, the image on the right has more contrast than the image on the left. This is what you're saying?

If contrast is about the steepness of the curve, then sure, OP's curve does increase contrast.
But that's a rather irrelevant definition, perceived contrast is more about the difference between the white point and the black point. And there, OP's definitely reduces it.

Let's start by defining contrast, people.

>>2723219
Doesn't look like those anchor points were moved at all, which if they weren't, it's not an s curve. If they were, then sure, an imperceptible amount.
>>2723220
Black point and white point are dynamic range, not contrast. They do give an upper and lower limit to what the contrast of an image can be, but they don't define contrast themselves. Contrast, especially perceived contrast, is more about the range of tonal densities and the gradation between them than the total tonal range of the image.

Take for instance a black to white grade against a series of dark gray and white bars. The bars are far more contrasty even though they have a smaller dynamic range because they have fewer gradations between them and the tonal densities of the image are firmly separated. The grade acts as a visual bridge between the two extreme tonalities easing the transition.

>>2723231
>it's not an s curve
No, it's just a curve shaped like an S....

Are you fucking seriousing me right now...

>>2723168
thanks for explaining that. best way i've heard it explained

>>2723243
If there's no curve at all, it's not shaped like an s. It's just three lines.

>>2723171
>This is the same as if one were to not touch the black point and white point

No it isn't.
There still aren't any very dark or very light tones in the image.

like >>2723192 shows: the shadows get less contrast.
That's because the black point was raised.

>>2723255
>If there's no curve at all, it's not shaped like an s
Are you saying that there is no curve in >>2723192
Because there is... two of the four anchors have been moved, which changes the linear adjustment to be the shape of an S. A sort of "S curve" if you will...