A Couple of Love Birds

[Mike Buckley]

Christina,

Your last treatment is very nice. Well done!

For a different treatment that lends to a high-key style, add a new adjustment step to apply a new curve. Move the black point to 14, the white point to 234 and aggressively pull the middle of the curve upward.

[Mike Buckley]

what is the difference between pulling down the very top part of the line versus the point on the first intersection on the top right hand corner, and pulling up the very bottom of the line versus the point on the first intersection on the bottom left hand corner.

Your question is good mostly because it indicates that you haven't digested that you're working with a graph and that you don't yet understand the data provided by the X axis and the Y axis. That information is vital to your adjustment of the curve. You should be able to easily find this information on the Internet. I'm surprised that it wasn't explained in your most recent reading.

To answer your question, the primary difference between the two adjustments that you mentioned is that the first adjustment darkens and lessens the contrast in the very brightest tones in the image and the second adjustment brightens and lessens the contrast in the very darkest tones.

[Brownbear]

Hi Mike,


Thank you. Very encouraging to hear that.

I just tried the high key thing on the image that Terry used, that you used and on a couple of others.. All my attempts were complete disasters. I also watched a LR video on how to do this, and it did not help either. I will save these for later when I'm more skilled...

In LR 4 I can only move the W&B points to 100... I worked these out to a percentage 5 for black and 95 for white, but this blew the whites in the swan, and pulling up the middle of the curve aggressively was also an editing disaster. Far worse than the ones that I have posted here.

All the same, thank you for trying to help. Truly appreciated.

Christina,

Your last treatment is very nice. Well done!

For a different treatment that lends to a high-key style, add a new adjustment step to apply a new curve. Move the black point to 14, the white point to 234 and aggressively pull the middle of the curve upward.

[Brownbear]

Hi Mike,

My book by Scott Kelby does not explain the graph, and the tutorials (tones and curves all finished) only touch on the graph in part, very briefly, showing that one can pull the vertical lines placed along the x axis to lighten the blacks, mid greys and whites, or darken... But it does not show any useful purpose for this application. There may be other tutorials in the set that cover it but they are not in the tones section.

I googled the Curves x/y axis and found a more informative article which explained that the x axis and y axis both represent the black points, middle greys and white points, and from that I can see that if I pull up the say the middle portion of the line I will lighten up the middle greys to lighter greys, and darken these when I pull in the opposite direction. And I can see that pulling up or pulling down any portion of the line is also going to effect the tones in the region that I pull...

Therefore whenever I pull up or down on that curves tool I am effecting the tones (brightening or darkening them) in my image. That said how in the world does one know which tones to adjust? The simplest way seems to be to use that TAT tool and place it on the portion of the image one wishes to lighten or darken, but would this not effect the tones in the entire image, meaning one really has no control over the tones in the image?

I can also see that I can move the black and white points from pure black or pure white but I can't envision a reason for doing this.

I watched the LR tutorial on tones again, and I see that when I went into the separate RGB channels for the swan edits that I was actually introducing a colour cast. I did not realize that when doing so. So I would think that this particular tool would be useful in a landscape scene where I might want to darken or lighten the red or green tones in an image, or in a scene with blue flowers for the same reason?

With respect the x and y axis of the curves and levels tool, am I understanding what you think I've been missing?

Thank you.

Your question is good mostly because it indicates that you haven't digested that you're working with a graph and that you don't yet understand the data provided by the X axis and the Y axis. That information is vital to your adjustment of the curve. You should be able to easily find this information on the Internet. I'm surprised that it wasn't explained in your most recent reading.

To answer your question, the primary difference between the two adjustments that you mentioned is that the first adjustment darkens and lessens the contrast in the very brightest tones in the image and the second adjustment brightens and lessens the contrast in the very darkest tones.

[Mike Buckley]

Yikes, Christina! After reading your post I reviewed the Curve tools in LR5 and Elements 9 for the first time. What a woefully lacking pair of tools! Now I understand why you have not been able to comprehend this stuff for so long about details I have been explaining about a Curve tool.

When I was mentioning the need to understand the data provided in a Curve tool, I was thinking of an X axis that has values of 0 to 255, the standard range of values. Neither of the two apps have that in their Curve tool.

I and others have also mentioned to you in the past to hover your mouse over the image and look in the Curve tool to determine the value of the pixel within that range of 0 to 255. Neither Curve tool allows you to do that, at least not as far as I can tell.

Even GIMP, which is free, allows you to do all of the above. And it took me all of 30 seconds to determine how to do it just now. My guess is that you have to pay for CS to get those capabilities from Adobe.

I wonder if the Nik plug-ins provide that capability. I haven't tried any of them even though I licensed the package months ago.

It seems to me that Adobe has designed those Curve tools to be able to use presets and sliders rather than simply control every aspect of the curve using anchor points, which to my understanding is the traditional manipulation of a curve. After getting used to the interface, I might conclude that those systems are indeed easy to use. However, until I figure that out they are actually considerably more difficult to control than simply dragging the curve and using anchor points.

[Mike Buckley]

I can also see that I can move the black and white points from pure black or pure white but I can't envision a reason for doing this.

That's actually not what is happening.

Assume the total dynamic range shown in the histogram begins with no tones on the far left or far right sides of the graph. When you move the black point to the right until it coincides with the darkest point tone in the image, you are mapping that tone so its luminance value is zero rather than its original tone. Similarly, when you move the white point to the left, you are remapping the brightest tone in the image so its luminance value is now 255 rather than its original tone. The technical gurus will surely explain that that is a vastly, overly simplified explanation, but I hope it suffices.

So, to answer your question of why you would move the black and white points, the reason to move the black point is to map the darkest tones in you image to make them darker. Similarly, the reason to move the white point is to map the brightest tones in your image to make them brighter.

[Brownbear]

Hi Mike,

Thank you for sharing. It is helpful to know why I am having a hard time comprehending and trying to follow your advice. Once I finish the tutorials in LR, I will move on to learning about Elements and layers (likely signing up for more tutorials) and once I'm comfortable with the basics perhaps I will move on to CS, and perhaps Nik, too... Right now I will just work on mastering the basics.

Thank you.

Yikes, Christina! After reading your post I reviewed the Curve tools in LR5 and Elements 9 for the first time. What a woefully lacking pair of tools! Now I understand why you have not been able to comprehend this stuff for so long about details I have been explaining about a Curve tool.

When I was mentioning the need to understand the data provided in a Curve tool, I was thinking of an X axis that has values of 0 to 255, the standard range of values. Neither of the two apps have that in their Curve tool.

I and others have also mentioned to you in the past to hover your mouse over the image and look in the Curve tool to determine the value of the pixel within that range of 0 to 255. Neither Curve tool allows you to do that, at least not as far as I can tell.

Even GIMP, which is free, allows you to do all of the above. And it took me all of 30 seconds to determine how to do it just now. My guess is that you have to pay for CS to get those capabilities from Adobe.

I wonder if the Nik plug-ins provide that capability. I haven't tried any of them even though I licensed the package months ago.

It seems to me that Adobe has designed those Curve tools to be able to use presets and sliders rather than simply control every aspect of the curve using anchor points, which to my understanding is the traditional manipulation of a curve. After getting used to the interface, I might conclude that those systems are indeed easy to use. However, until I figure that out they are actually considerably more difficult to control than simply dragging the curve and using anchor points.

[Brownbear]

Hi Mike,

Thank you. Very helpful and understood!

That's actually not what is happening.

Assume the total dynamic range shown in the histogram begins with no tones on the far left or far right sides of the graph. When you move the black point to the right until it coincides with the darkest point tone in the image, you are mapping that tone so its luminance value is zero rather than its original tone. Similarl
y, when you move the white point to the left, you are remapping the brightest tone in the image so its luminance value is now 255 rather than its original tone. The technical gurus will surely explain that that is a vastly, overly simplified explanation, but I hope it suffices.

So, to answer your question of why you would move the black and white points, the reason to move the black point is to map the darkest tones in you image to make them darker. Similarly, the reason to move the white point is to map the brightest tones in your image to make them brighter.

[Loose Canon]

Christina I did my particular high key mono version in Photoshop, which is why I didn’t explain the steps I took. I’m sorry but I don’t have Lightroom.

However, I do have a version of Elements 6 (yikes, 6!) on my machine and I took a little stroll down Memory Lane with it!

Would you be interested in joining me on that little stroll? I tried to reproduce the results and got pretty close. Though again I didn’t take the time the shot deserved, I at least got the steps and I would be glad to work up a post for you to play around with if you would like.

If not, no worries! Sometimes its better to step away for a time and come back to it later (at least for me anyway)!

[Brownbear]

Hi Terry,

I would love to take a stroll with you down Memory Lane and see if I can come close to your results, as I adore the look of the swan edit you did, and I have a ton of swans to work with. I might even have an advantage as I have the raw file and you worked with one of my wonky edited jpegs.

In short if you are willing to take the time to provide me with some directions, I would love to try it and would give it a go tomorrow. I should say that I have less experience with Elements than I do Lightroom and I have only tried a an adjustment layer a couple of times...

Therefore if it is highly complicated and something that you think I could not manage to do, I could study for a few more months and then send you a PM to let you know that I am ready to try it... I don't wish to waste your time if it is beyond my skill level right now. And I'm just fine with waiting... But if it is fairly simple I would like to try it.

Thank you.

Christina I did my particular high key mono version in Photoshop, which is why I didn’t explain the steps I took. I’m sorry but I don’t have Lightroom.

However, I do have a version of Elements 6 (yikes, 6!) on my machine and I took a little stroll down Memory Lane with it!

Would you be interested in joining me on that little stroll? I tried to reproduce the results and got pretty close. Though again I didn’t take the time the shot deserved, I at least got the steps and I would be glad to work up a post for you to play around with if you would like.

If not, no worries! Sometimes its better to step away for a time and come back to it later (at least for me anyway)!

[Mike Buckley]

I highly object to Terry explaining how he created the high-key version before explaining to me the setup that he used for his photo in this thread. It's really simple: I asked first.

[Stagecoach]

Christina,

I have a plugin for Elements (I'm only using Ver 7) called 'SmartCurve' which is a free download, it may be that it is the tool Mike is referring too?

Whilst I have it and it appears to work fine it is one of those things I just have not explored yet

Grahame

[Loose Canon]

Allllllrighty then Christina! Its great having you with me on our Stroll! And it was great seeing my Old Friend Elements again too! This shouldn’t be horrifyingly complicated (famous last words!) and any questions or if I suck at this just fire away! I don’t know how familiar you might be with Elements and there will always be other (probably better) ways to go about it within the program and I don’t know how the newer versions of Elements are set up. But we should get in the same zip code anyway!

With the image open in Elements the first thing I did was duplicate the background layer. In the menu bar go to Layer>Duplicate Layer (Command key/j or Alt key/j with Windows). This duplicate layer (default name Layer 1) is what I will start working on leaving the original BG layer alone. In case everything goes horribly wrong I have a fallback!



Next I wanted to make the B&W conversion. I went to the menu bar and with Layer 1 highlighted (selected) tapped Enhance>Convert to Black & White. In the Convert to Black & White pane I chose under “Select a Style”> Portraits. Then went next to it to the “Adjustment Intensity” box and pushed the Red slider to +85. This is a good place to play around. Styles, intensities, contrast, to get what you like. For the dream-like look I didn’t want it too contrasty so I just left that slider alone. When you have played to your Heart’s content tap “Okay” and boom, conversion done on Layer 1.





I wanted to define what would actually show of the image and what I wanted to “white out”. So I essentially put a white vignette on the image.

To do this I created a new layer. Click the top left Icon above the layer stack.



Default name Layer 2. Next I filled the layer with white. With Layer 2 selected I went to the menu bar and tapped Edit>Fill Layer. In “Contents” box chose white. Blend mode “normal”, Opacity 100%, and the “Preserve Transparency” box unticked. Tap “Okay”. Then I hid the layer by clicking the eye icon beside the layer box.





Next I selected the Elliptical Marquis tool and with Layer 1 highlighted selected an ellipse around the birds. You can move the ellipse around by depressing your space bar while holding your click with the mouse. Release the space bar while still holding the mouse click to finish the ellipse.





After this is done put your cursor inside the ellipse, control/click and a box will appear. Select from the box “Refine Edge”. Or go to the menu bar Select>Refine Edge, and you should be seeing this. I set the smooth at 3, feathered it 86 pixels (to soften and graduate the selection), and expanded the selection +38%.



Go back to your layer stack, make Layer 2 (the white filled layer) visible (eye icon again) and have it selected (highlighted). You should see nothing but a white filled layer with your selection (marching ants!) on it.



With Layer 2 selected hit your delete key and that will delete the white fill that is inside your elliptical selection revealing the birds. Hit Command/d (Alt/d on windows) to deselect (get rid of) the selection outline. Or go to menu bar Select>Deselect.




At this point Christina you could be finished if it looks good to you. I took it a little further by dodging and burning to bring a little local contrast to the birds and to the water ripples/reflections in front of them.

So if you get to here okay Christina and would like to continue our stroll with dodging and burning I would be happy to stroll with you! This can be done without adjustment layers and still have nice local control over what you want to contrast. Plus its big fun and makes for some very nice finishing touches to put the polish on your image!

Here is what I have at this point.



Edit Note: The intensity of the white vignette may be controlled by the opacity control above the layer stack. With the white fill layer selected, adjust the opacity control accordingly. I left it at 100%. To use the control as a slider, click and hold on the word "Opacity" and move the mouse left or right.

[Loose Canon]

I highly object to Terry explaining how he created the high-key version before explaining to me the setup that he used for his photo in this thread. It's really simple: I asked first.

I apologize profusely Mike! I guess I forgot I had a thread and that's kind of irresponsible of me!

Remind me not to call you names (Internationally Famous I believe it was) again will you?!!

[Brownbear]

Thank you so much Terry. It looks simple enough to do especially since you have presented all the steps so nicely. I will try this later today and hope to post something half decent. And perhaps we will take a stroll through Dodge and Burn land, too.

[Mike Buckley]

I have a plugin for Elements (I'm only using Ver 7) called 'SmartCurve' which is a free download, it may be that it is the tool Mike is referring too?

I'll explain the type of tool that is so much more helpful than the tools in LR5 and Elements 9. This one happens to be a Levels tool and Curve tool combined into one tool, but that's not important. I'm referring only to the Curve tool now.

In both screenshots shown below, the horizontal axis indicates the luminosity values of zero to 255 from left to right. The vertical axis indicates the relative number of pixels in the image that have a particular luminosity from none at the bottom to the most at the top. The grey pattern displayed within the graph is the composite histogram of the photo before the curve is adjusted, so it displays each luminosity value (on the horizontal axis) and the relative number of pixels (on the vertical axis) that have a particular luminosity value.

In Screenshot 1, the diagonal line is the linear curve before the curve has been adjusted by the user. Notice the black dot near the center of that diagonal line. That is there because I had held the mouse pointer over a pixel in the image. That black dot graphically displays the luminosity value. More important, notice the Input and Output values, which are the same for now. (More about that below.) That value is the luminosity value of the pixel identified by the mouse pointer.

In Screenshot 2, I have applied a standard S-curve. Notice that the histogram has not changed; it still displays the original histogram before the S-curve has been applied. (To see the histogram of the image after the S-curve has been applied, I would simultaneously display the histogram panel in my software.)

Also in Screenshot 2, notice the two large black dots in the curve. Each of those dots are anchor points. They are called that because they are the two places in the curve that I used to drag the curve and because that very specific point is anchored in place and will not be moved regardless of what I do to other parts of the curve. I can place many, many anchor points in the curve to make whatever shape happens to make the best photograph.

Notice the two small black dots in Screenshot 2. As in Screenshot 1, my mouse pointer was hovering over a particular pixel in the image. Those small black dots indicate the luminosity value on the linear curve (the curve before I adjusted it) and on the S-curve. As in Screenshot 1, the luminosity value of that pixel is also displayed as Input and Output values. The Input value is the luminosity value of that pixel before the Curve has been adjusted. The Output value is the luminosity value that that pixel has been mapped to as a result of having adjusted the Curve.

Any Curve tool that lacks any of the above capabilities is lacking the very important attributes that give the user total control over the curve and information about the pixels in the image. Just my opinion.


Screenshot 1



Screenshot 2

[Loose Canon]

Thank you so much Terry. It looks simple enough to do especially since you have presented all the steps so nicely. I will try this later today and hope to post something half decent. And perhaps we will take a stroll through Dodge and Burn land, too.

You are most welcome Christina. You are great company on a Stroll! There is not even the slightest doubt that you could bring us something absolutely gorgeous.

But here is the most important overall Big-Picture thing you can do...

Have fun with it!

If we miss that then we miss the whole point!

[Brownbear]

Hi Terry,

Thank you. You certainly know how to make a stroll through post-processing fun. I'm afraid that I did not manage to create anything beautiful, in fact I made my white swans grey but the important thing is that I learned to do something new.

Your lesson on creating a high key b&w photo was also a good lesson in using layers. I expect that the next time I am ready to try layers again, I will be referring to this thread and your truly wonderful step by step instructions that even a five year old could follow. Thank you for that.

I forgot that my Elements 9 does not recognize my raw files from my new camera, therefore I processed in LR and did the last steps the B&W and the vignette in Elements on a jpeg. In LR I burned the water ripples and the water by 5 points.

When I converted to B&W and dragged the red slider up to 85 the swans looked really horrid, lots of white clipping so I pulled it back to 66, decreased the greens to 4 and the blues to 15... When I look at my histogram I can see that I changed a low contrast photo to a higher contrast photo (2 mountain peaks)

Below is my try.... I don't know why my vignette is so obvious whereas yours isn't... I did follow your step to the letter, except for dragging the reds so far over and only because it was clipping the swans terribly, plus my swans turned grey.





I think I do know how to dodge and burn but the problem is that whenever I try it, it looks artificial... ie the water ripples in the photo above dodged by just five points and the water does not look real.. Do you have any tips to share on how to dodge and burn well?

Yesterday, I tried a couple of high key swan shots. I am going to edit them following your steps here... I presume it will work better on a photo that is already high key. I will post in a separate thread so this one does not become too confusing.

Thank you.

PS
You're a great teacher, it is the pupil who is a little inept.

You are most welcome Christina. You are great company on a Stroll! There is not even the slightest doubt that you could bring us something absolutely gorgeous.

But here is the most important overall Big-Picture thing you can do...

Have fun with it!

If we miss that then we miss the whole point!

[Brownbear]

Thank you for the wonderful detailed explanation of the curves tool and the explanation.

Why is it good to know the luminosity value of the pixels you adjusted, ie; the final input and output points?

I'll explain the type of tool that is so much more helpful than the tools in LR5 and Elements 9. This one happens to be a Levels tool and Curve tool combined into one tool, but that's not important. I'm referring only to the Curve tool now.

In both screenshots shown below, the horizontal axis indicates the luminosity values of zero to 255 from left to right. The vertical axis indicates the relative number of pixels in the image that have a particular luminosity from none at the bottom to the most at the top. The grey pattern displayed within the graph is the composite histogram of the photo before the curve is adjusted, so it displays each luminosity value (on the horizontal axis) and the relative number of pixels (on the vertical axis) that have a particular luminosity value.

In Screenshot 1, the diagonal line is the linear curve before the curve has been adjusted by the user. Notice the black dot near the center of that diagonal line. That is there because I had held the mouse pointer over a pixel in the image. That black dot graphically displays the luminosity value. More important, notice the Input and Output values, which are the same for now. (More about that below.) That value is the luminosity value of the pixel identified by the mouse pointer.

In Screenshot 2, I have applied a standard S-curve. Notice that the histogram has not changed; it still displays the original histogram before the S-curve has been applied. (To see the histogram of the image after the S-curve has been applied, I would simultaneously display the histogram panel in my software.)

Also in Screenshot 2, notice the two large black dots in the curve. Each of those dots are anchor points. They are called that because they are the two places in the curve that I used to drag the curve and because that very specific point is anchored in place and will not be moved regardless of what I do to other parts of the curve. I can place many, many anchor points in the curve to make whatever shape happens to make the best photograph.

Notice the two small black dots in Screenshot 2. As in Screenshot 1, my mouse pointer was hovering over a particular pixel in the image. Those small black dots indicate the luminosity value on the linear curve (the curve before I adjusted it) and on the S-curve. As in Screenshot 1, the luminosity value of that pixel is also displayed as Input and Output values. The Input value is the luminosity value of that pixel before the Curve has been adjusted. The Output value is the luminosity value that that pixel has been mapped to as a result of having adjusted the Curve.

Any Curve tool that lacks any of the above capabilities is lacking the very important attributes that give the user total control over the curve and information about the pixels in the image. Just my opinion.


Screenshot 1



Screenshot 2

[Mike Buckley]

Why is it good to know the luminosity value of the pixels you adjusted, ie; the final input and output points?

When the Output and Input values are different, I think the Output values are more important because they indicate the changed luminosity. Beyond that...

The brightest parts of an image are generally the parts that the viewer sees first, though there are also other factors such as saturation, contrast and sharpness. You want to "make" an image that leads the viewer's eye through the image and you want to control where in the image that journey is to begin. If you're wondering which parts of the image have the brightest luminosity (two different colors can have the same luminosity), you can easily determine that by reviewing the luminosity values.

You might also want to control the tonal range of part of the image, such as the background of a studio shot or the vignette around your swans. As an example, very recently Grahame made an image of a tomato and he wanted the white background to be uniformly white. Rather than rely on his eyes, perhaps the inadequacies of a monitor such as mine that doesn't cost several thousand dollars, and all of the deficiencies that go along with evaluating an image using a monitor when the goal is to make a large print, he was able to determine that the range of luminosity in his background was within the values of 253 to 255 -- very uniform throughout.

If you want to ensure that certain very dark tones are true black, the easiest and perhaps only way to reliably do that is to determine whether their luminosity value is zero. (I should use that method more often when making photographs of glass using a black background.)

There are probably other examples that aren't coming immediately to mind, but you get the point.