So this didn't work...

CWSkopec · July 1, 2008

A little while ago, I posted this image in the whipping post,

And one commenter suggested "tinkering with the histogram" to acheive a "real white." I'm familiar with Photoshop, but more from a graphic design stand point and not so much from a photography standpoint yet. But I had see the "Make your photos pop" under the tutorials, so I followed that to find the white points (and black point... HEY I was following step by step!! :rofl) Anyway, with just the white point curves adjustment this is what I get...

:huh NOT GOOD!! Even with that layer's opacity knocked down, it just looks washed out and doesn't get the colors to really pop, which was the point of the suggestion... SO I know this photo doesn't have a true white point in it to measure from and that may be the problem... but what's a good method to really make this shot all that it could be? :scratch

Thanks!
And by all means, feel free to take and tinker with this image all you want!

Miguel Delinquento · July 1, 2008

more than skin deep
Chris:
This took about 2.5 minutes in Lightroom

It was a matter of jacking the exposure and contrast up, lots of sharpening, increasing shadowing, increase saturation. This image has more pop now. I'm curious to see other attempts.

It is still a weak image however. The DOF is too wide open to me and the angle of the shot is a little strange. I couldn't get a satisfying crop, yet it calls out for one. And it really should be sharper to begin with. I'm assuming this was hand held?
Perhaps using a tilt-shift lens would enabled a more engaging capture.
I would have taken the shot too as the potential of the colors is irresistable.
Have any out-takes?

M

CWSkopec · July 1, 2008

Miguel Delinquento wrote:

Chris:
This took about 2.5 minutes in Lightroom

It was a matter of jacking the exposure and contrast up, lots of sharpening, increasing shadowing, increase saturation. This image has more pop now. I'm curious to see other attempts.

It is still a weak image however. The DOF is too wide open to me and the angle of the shot is a little strange. I couldn't get a satisfying crop, yet it calls out for one. And it really should be sharper to begin with. I'm assuming this was hand held?
Perhaps using a tilt-shift lens would enabled a more engaging capture.
I would have taken the shot too as the potential of the colors is irresistable.
Have any out-takes?

M

WOW!! That sure pops more... I'll have to play with those settings more... I'm assuming it's better to tweak those settings in Camera Raw before "opening the image," yeah?

As for it being a weak image, I kind of agree with you... but not whole heartedly... it just begged me to inspect it for a week or more before uploading it, and I never could decide if it was for good reasons or bad. And it may entirely be the angle and the DOF leading to the corner... BTW the angle was a result of me not fully commiting myself to just lie down on the ground where I could have had a better look in the view finder, so speaking of "out-takes," Yes, there were plenty. It was hand held, but on a well lit day so I figured it was safe, maybe my shutter speed was just a bit slow...

Anyway, out-takes... here's a couple with a slightly less "unique" (

) angle, I haven't processed them beyond Raw's auto tweaks as I don't have the time at the moment, but I wanted to respond because I always appreciate your critiques, Miguel.

Thanks for checking them out!
And thanks for the comments, Miguel!

Mnemosyne · July 10, 2008

The problem with the first original photo is that you don't have a true white point. And most of the "blacks" in the photo look like shadows (which aren't true blacks either).

So the color skewing came from the lack of a white point. You fooled the software into thinking one color was pure white when it actually wasn't.

CWSkopec · July 14, 2008

Mnemosyne wrote:

The problem with the first original photo is that you don't have a true white point. And most of the "blacks" in the photo look like shadows (which aren't true blacks either).

So the color skewing came from the lack of a white point. You fooled the software into thinking one color was pure white when it actually wasn't.

Ok, that's what I was assuming. Silly Photoshop shouldn't have listened to me!!

So pretty much, the photo used in the "Making your photos pop" tutorial is a best case scenario in that it has both the white shirt and the black hat, right? Which leads me to wonder if there's a better method for photos that don't have one or both... Anyone know of such a thing? Or should I spend a little more time looking at the tutorials we already have here?

Thanks!!

Mnemosyne · July 14, 2008

CWSkopec wrote:

Ok, that's what I was assuming. Silly Photoshop shouldn't have listened to me!!

So pretty much, the photo used in the "Making your photos pop" tutorial is a best case scenario in that it has both the white shirt and the black hat, right? Which leads me to wonder if there's a better method for photos that don't have one or both... Anyone know of such a thing? Or should I spend a little more time looking at the tutorials we already have here?

Thanks!!

Exactly, but even then you have to be careful. Just because on your screen it looks white, doesn't mean it actually is. Deep shadows and specular highlights (reflections on glass and other reflective materials) are outside the actualy color spectrum. 245 to 250 is basically white. 255 is beyond white.

As to a better method, if you just play around with them, you'll learn. I have been playing around for the past year and a half, balancing my own photos. And I was photo editor last spring at my school paper, so I got to edit a LOT of photos. For the most part, I don't use the "find the white/black" points. I just play with the levels until I like the results. I've gotten good at recognizing when there is a color cast now.

It just takes practice.

BinaryFx · July 15, 2008

Before I attempt to answer the original or subsequent points raised...

What colour space is the image in?

The file on the web is tagged with an Adobe RGB ICC profile, which yields a different result viewing between non colour managed web browser and Photoshop which is colour managed. Most "net" images are considered to be sRGB (for right or wrong).

Should the saturation look like Adobe RGB when viewed in Photoshop, with the embedded Adobe RGB ICC profile tag - or should it look like A98 numbers being presumed as sRGB and not A98??? The difference may be subtle, although it should be more noticeable in the yellow/orange hues and the grouting between the tiles.

Stephen Marsh
http://members.ozemail.com.au/~binaryfx/

CWSkopec · July 15, 2008

BinaryFx wrote:

Before I attempt to answer the original or subsequent points raised...

What colour space is the image in?

The file on the web is tagged with an Adobe RGB ICC profile, which yields a different result viewing between non colour managed web browser and Photoshop which is colour managed. Most "net" images are considered to be sRGB (for right or wrong).

Should the saturation look like Adobe RGB when viewed in Photoshop, with the embedded Adobe RGB ICC profile tag - or should it look like A98 numbers being presumed as sRGB and not A98??? The difference may be subtle, although it should be more noticeable in the yellow/orange hues and the grouting between the tiles.

Stephen Marsh
http://members.ozemail.com.au/~binaryfx/

Stephen,
It's in Adobe RGB. I briefly experimented with that color space as I had glanced at an article saying it offered a wider gamut of colors. Then I read the article (I believe) on Smugmug about the differences and why they use sRGB as their standard for printing. I switched the camera back to sRGB after that, but this must have been taken durring the "experimental" time!

Thanks for posting the side-by-side! The difference is subtle enough that if they weren't next to each other, I might never notice the difference but I can see it here.

I think I understand what you're getting at, but just to be sure... One of the essential issues here is that while I processed in an Adobe controlled medium (Photoshop) it's being displayed in sRGB through everyone's various web browers. And thus the image I saw, which could still use some work I'm sure, isn't what everyone else is viewing. So it appears to be in worse shape than it really is. Right?

Thanks for your comments/time!

BinaryFx · July 15, 2008

CWSkopec wrote:

I think I understand what you're getting at, but just to be sure... One of the essential issues here is that while I processed in an Adobe controlled medium (Photoshop) it's being displayed in sRGB through everyone's various web browers. And thus the image I saw, which could still use some work I'm sure, isn't what everyone else is viewing. So it appears to be in worse shape than it really is.

That is pretty much right Chris, non-CM web browsers/monitors may not be in sRGB, although that is a common standard that is close for many. As CM web browsers are not that common, it is usually* a safe bet to presume that any given image downloaded from the web is "approximately sRGB" - so when I opened up the image in Photoshop, I did not know for sure if the ICC profile was valid or not.

*Photo sites like this are often an exception, general "public display" images are usually safe to presume as intended to be in or close to sRGB.

It is fine to shoot/work in Adobe RGB or another RGB space instead of sRGB, as long as you rembember to convert all duped web/monitor display images to sRGB (which is a defacto "web output" colour space, just as one would convert a dupe to the approprite print space for inkjet or printing press when the output is ink on paper).

The working space or "editing" space is independent of both the monitor profile and output profiles for say a printer.

I get the impression that you shoot in-camera JPEGs rather than Raw camera data as you mentioned the camera colour setting which is not applicable to Raw data and the curve moves that you made were in Photoshop as the Raw converter does not offer such curves and white balance and exposure would be the most similar (you may still wish to stick with Adobe RGB in the camera, as long as you remember to convert dupes to sRGB for non colour managed monitor viewing)...Although Raw processing is also mentioned later so I am unsure of your workflow.

I think it will be easier for me to make shorter posts on different points raised in the discussion, instead of one long post. More to come!

Regards,

Stephen Marsh
http://members.ozemail.com.au/~binaryfx/

BinaryFx · July 15, 2008

CWSkopec wrote:

Ok, that's what I was assuming. Silly Photoshop shouldn't have listened to me!!

So pretty much, the photo used in the "Making your photos pop" tutorial is a best case scenario in that it has both the white shirt and the black hat, right? Which leads me to wonder if there's a better method for photos that don't have one or both... Anyone know of such a thing? Or should I spend a little more time looking at the tutorials we already have here?

Chris, on the bolding, I agree, it is always a good idea to read up, experiment and then ask others who have been down this road once or twice before for their thoughts (just as you have been doing)!

I would not say that it is a "best case" scenario, it just happens to have areas that can be improved (shadow and highlight points) by adjusting colour balance as the tonal range and contrast is also increased.

A couple of days ago I made a post in another topic which goes into some of your questions...

http://www.dgrin.com/showthread.php?t=99389

When one sets the black/white endpoints, one is usually setting range and altering colour balance in one step.

The original image may have a casted shadow of say 15r 12g 16b, setting the black point at an idealised neutral R=G=B value of say 10r 10g 10b will extend the tonal range as well as affect the original colour balance. Usually this is a good thing, however there will also be exceptions.

The white point is similar.

So, there are three general goals when adjusting the tonal range of an image. 1 - Change tone and colour balance with the one curve operation (usually to each individual channel, ignoring the master/composite curve). 2 - Change tone without affecting hue. 3 - Change hue/saturation without affecting tone.

You wish to change luminosity and saturation...but not hue.

A simple Photoshop curve option is to open up levels or curves and hit the option button, set the Auto Colour Correction Options to the top radio button labeled "Enhance Monochromatic Contrast" (auto contrast). This will move the separate straight line channel curves same distance, attempting to preserve general hue while increasing tonal range and saturation.

Not every image will require endpoint adjustments, it would be crazy to force an image to the extremes if it did not benefit from the move. Some images may require adjustment of tonal range - but not colour balance (luminosity blend mode curves). Some may require adjustment of colour balance, but not range (color blend mode curves). Some may require separate luminosity and colour curves instead of a single normal blend mode curve.

It is not always so much about the extreme (or near to) darkest and lightest areas of the image - but more so the darkest and lightest areas of the image that contain detail or that are more critical to the observer than areas of lesser importance (targeting the curve moves to the content, rather than to the darkest/lightest areas).

Curves in ALR/ACR generally preserve hue, while in Photoshop there are other ways to affect contrast without messing with hue too much or at all, with auto contrast being one (or manual edits which also preserve the general colour balance in normal blend mode)...luminosity blend mode curves are perhaps the simplest way to do this, followed by separate saturation adjustments (RGB contains both colour and tone, sometimes it helps to work these independently of each other).

CWSkopec wrote:

WOW!! That sure pops more... I'll have to play with those settings more... I'm assuming it's better to tweak those settings in Camera Raw before "opening the image," yeah?

Chris, if the file starts life as Raw camera data, then it is common accepted practice to do as much in the raw converter as possible. Some break this accepted practice though, but for many this is the only way. If you are opening a non Raw camera data file into ALR/ACR, then I think the jury is still out on whether it is better to tweak the rendered/gamma encoded, non-Raw camera data in Lightroom/Camera Raw or in Photoshop.

Sincerely,

Stephen Marsh
http://members.ozemail.com.au/~binaryfx

CWSkopec · July 15, 2008

BinaryFx wrote:

I get the impression that you shoot in-camera JPEGs rather than Raw camera data as you mentioned the camera colour setting which is not applicable to Raw data and the curve moves that you made were in Photoshop as the Raw converter does not offer such curves and white balance and exposure would be the most similar (you may still wish to stick with Adobe RGB in the camera, as long as you remember to convert dupes to sRGB for non colour managed monitor viewing)...Although Raw processing is also mentioned later so I am unsure of your workflow.

I think it will be easier for me to make shorter posts on different points raised in the discussion, instead of one long post. More to come!

Regards,

Stephen Marsh
http://members.ozemail.com.au/~binaryfx/

First, let me just say thank you for the depth of your discussion on this, Stephen!! It overwhelmed me a bit this morning when I looked at it as I was just to busy to fully read and comprehend it at work!!

I've been shooting in Raw for just about a month or more... I admit, I was "afraid" to try it at first...

I was imagining that even in Raw the color space dictated in the camera had a bearing... but then again, I don't fully understand the Raw format yet, and just have had time to get more than a few chapters into the book I bought about it. Taking a quick look through everything I've processed in the last few days and they're all Adobe RGB images because they're all run through Photoshop Camera Raw I'm guessing. I'll add "Convert to sRGB" to the list of steps for a image to be uploaded.

The curves and other adjustments I did were done on the resulting tif file after the Raw conversion because that's what I found in the tutorial I origionally started with.

In regards to your second post, I'm not sure how to respond other than to say Thank You once again!!

I just bookmarked it, because I'm sure I'll need to refer back to it a few times before it truly sets in and I have a better feeling of what adjusments I should be making.

BinaryFx wrote:

Chris, if the file starts life as Raw camera data, then it is common accepted practice to do as much in the raw converter as possible. Some break this accepted practice though, but for many this is the only way. If you are opening a non Raw camera data file into ALR/ACR, then I think the jury is still out on whether it is better to tweak the rendered/gamma encoded, non-Raw camera data in Lightroom/Camera Raw or in Photoshop.

I think that instead of going out shooting this weekend, I'll reserve that time for some serious reading of the Raw conversion book I bought.

That sounds like a better way to spend my time. Also will be looking over your site, after a cursory glance, it looks like you've got a number of helpful tutorials and tips there!!

Duffy Pratt · July 17, 2008

Here's an example of a one minute job using LAB, and a single set of curves, following the basic recipe from one of the first five chapters of Margulis' The Canyon Conundrum.

CWSkopec · July 18, 2008

Duffy Pratt wrote:

Here's an example of a one minute job using LAB, and a single set of curves, following the basic recipe from one of the first five chapters of Margulis' The Canyon Conundrum.

Thanks, Duffy!!

By LAB, I'm assuming you're talking about the LAB color mode in Photoshop, yeah? What sort of advantages does that offer over RGB color?

I'll put the book on my reading list, but it's insanely too long ever since I got my camera... seems like I'm always out taking photos in my free time and not sitting back reading anymore...

Duffy Pratt · July 18, 2008

CWSkopec wrote:

Thanks, Duffy!!

By LAB, I'm assuming you're talking about the LAB color mode in Photoshop, yeah? What sort of advantages does that offer over RGB color?

I'll put the book on my reading list, but it's insanely too long ever since I got my camera... seems like I'm always out taking photos in my free time and not sitting back reading anymore...

There's a sticky thread at the top of this forum which summarizes the chapters of Margulis' book. It goes into more than enough detail to show the sort of thing that I did here.

Duffy

CWSkopec · July 18, 2008

Duffy Pratt wrote:

There's a sticky thread at the top of this forum which summarizes the chapters of Margulis' book. It goes into more than enough detail to show the sort of thing that I did here.

Duffy

Even better!! That'll be far easier to read up on!!

Thanks again, Duffy!!

So this didn't work...

Comments