However, for your subject I like the unpopped version a little better. You've done a fine job with the pop technique, I just like the softer feel of the unpopped version.
before/after popping question
David, thanks for the tutorial.
Here is my "before" popping image:
and here is the "after popping" image:
question: it looks as if contrast and saturation have both increased. - could simply increasing both in ACR or CS2 have achieved the same effect, or does this technique do something that goes beyond just boosting contrast and color? thanks
Cletus is right that portraits often call for a softer approach. I believe the answer for this shot lies somewhere in between the two treatments. You could put the popped version over the original and reduce the opacity, see what you get...
question: it looks as if contrast and saturation have both increased. - could simply increasing both in ACR or CS2 have achieved the same effect, or does this technique do something that goes beyond just boosting contrast and color? thanks
I'm not going to pretend that I understand enough to answer that question. I do know that adding pop (however you do it) is spreading out the values to take full advantage of the range available. This could be part of it. Setting the black point like that neutralizes it, which could effect colors, and lastly, when working in RGB it is difficult to separate these changes because of the way the color mode works. Color and contrast are both represented in the same channels, so shifts occur.
If you are interested you could read Rutt's excellent tutorial on adding pop in LAB mode, where luminosity is separated from color information. That tute is here.
The challenge with this image is that you want to add contrast to some areas (face, hair and skin), but you can't afford to raise the brightness of some of the brighter areas like the brightest areas of the dress or the background which would normally happen with the simpler means of adding contrast. You also have to make sure you don't go too far with skin.
There are several ways to accomplish this. The most common would be to just use a soft brush and paint a loose fitting mask over the brightest areas to just shield them from your changes. It might also possible to make the changes differently to avoid influencing the brighter tones and it is also possible to use blend-if settings to keep the change from the brighter areas. If any of these sounds too complicated, just stick with a simple mask over the dress and background.
Here's an in-between that keeps the changes off the dress and background. The changes are more subtle, but the goal was to bring out color and detail in the face and hair.
first time player
A major improvement, I suppose I'll have to read the tutorial if I want to play. A couple observations though. Isn't there any texture that can be salvaged in the black outfit? It really stands out as a distracting black hole. Also, the edges of the taller person's head have a bit of a cut-out quality. I might add a little noise to the edge to soften it up, blend it in. Finally, depending on how much effort is warranted, I would "powder" the faces to hold back the shine.
A major improvement, I suppose I'll have to read the tutorial if I want to play. A couple observations though. Isn't there any texture that can be salvaged in the black outfit? It really stands out as a distracting black hole. Also, the edges of the taller person's head have a bit of a cut-out quality. I might add a little noise to the edge to soften it up, blend it in. Finally, depending on how much effort is warranted, I would "powder" the faces to hold back the shine.
There's a lot of texture in the black dress. I assigned a combination of profiles at lower gamma settings than sRGB, converting back to sRGB each time. Then used Blend if... sliders to isolate the effect to the darkest regions in the original layer, masked out the odd stuff that always shows up with Blend if... sliders, then adjusted the contrast in that layer.
Some stock LAB moves for color and brightness.
There are two ways to slide through life: to believe everything or to doubt everything; both save us from thinking.
—Korzybski
Just want to note that the pop tutorial is not intended to be anything more than a basic first step, a PS 101. There are other tutorials that are intended to take post processing to the next level.
I read through the tut's. I think what I've been doing to get a similar effect to try in succession: Auto Levels - undo, Auto Contrast - undo, Auto Color - undo, and then pick which one looked best. (Always use keyboard shortcuts. :P)
Then go from there, maybe using Shadow/Highlight to pick up some detail in the shadows, or curves for brightening a bit. Sound reasonable?
I read through the tut's. I think what I've been doing to get a similar effect to try in succession: Auto Levels - undo, Auto Contrast - undo, Auto Color - undo, and then pick which one looked best. (Always use keyboard shortcuts. :P)
Then go from there, maybe using Shadow/Highlight to pick up some detail in the shadows, or curves for brightening a bit. Sound reasonable?
"Auto" anything is usually not a great idea. Each image requires different areas and ranges to be tweaked. Auto moves are a sledge hammer where a scalpel is called for. Curves are your first line of defense, followed by Selective Color and Hue/Saturation. Shadow/Highlights is always useful but you want to make sure you target only the areas that really need it.
Beyond that there are a whole host of techniques involving sharpening, forays into LAB space and plate blending. Lots of good info in this vein is wrapped up in the two discussion forums dealing with Dan Margulis' books.
There are two ways to slide through life: to believe everything or to doubt everything; both save us from thinking.
—Korzybski
I think you made all the rigtht moves. I've learned, however, that if a move looks good at 100% this afternoon, it will look a lot better at 70% tomorrow morning. The tendency is to push things incrementally and so lose track of how far a move has gone. The colors strike me as a little bit harsh but that might just be the slightly excessive sharpening. I'm not sure that those beautiful old buildings and bricks would come up that bright and neutral in the real world, however.
Like I said, every move feels right, just a little too far.
There are two ways to slide through life: to believe everything or to doubt everything; both save us from thinking.
—Korzybski
You definitely made a significant improvement on your image. There are several different ways you can go on the color of the original and I thought I'd show you a different approach that has a warmer tone (e.g. less blue). You can see which you prefer.
On this one, I only did two things to your original. Because the histogram looks like this:
With almost nothing on the right edge and no important details in the little tiny tail to the right, I applied this curve to enhance the contrast (the exact same move can be done in levels since all I did was push in the ends):
And, then sharpened a little to end up with this result:
More can be done (the curve could have a little s-shape for even more contrast, the sky could be dealt with separately to deepen the blue, etc...), but 90% of the effect comes from the simple part so I thought I'd just show that.
What jfriend's work points out is that your "popped" image ended up having a blue cast to it, although I believe his is too yellow. Often times it's good to use the gray dropper to neutralize and set the overall cast as one final step to adding pop. Or other times this simple pop recipe is simply not the best way to go. The advantage of this method is its simplicity, and its effectiveness on a wide range of shots. But not every shot is better off with it.
Anyway, I clicked the gray dropper on your original, slightly blue "popped" image, and got this (the red dot being the approximate area where I clicked):
Thanks David, that's the final step that does it for me.
I'm only just developing my PP workflow and now I've added a step at the end to:
1. go away from the monitor
2. have a beer (optional, but encouraged)
3. come back and re-assess whether all the colors are possible and realistic/desired.
I like the simplicity of your method at the end, so that's one added to the memory bank.
Thanks to the others that responded as well.
I've really only been doing anything with PP for 4 days now, having previously just clicked auto levels and then blindly applied an S-Curve to the master RGB.
So, I think after 4 days of taking color correction seriously, I've definitely been able to produce something that is;
1. better than the starting image; and
2. better than the simple method used before.
The sharpening here was something I was aware of, but thanks for the comment. I applied the sharpening on a separate layer in the larger version and then after re-sizing for the forum, I lazed out and just reduced opacity rather than trashing the layer and repeating. Should have taken the extra care and that's another one for the memory bank.
I'm hoping that in another 4 days I can repeat this image from scratch and come up with an even better image and then again in a few weeks.
Thanks all for the comments.
Regards,
Peter
It's not my camera's fault, I'm just visually illiterate
A tip I found somewhere around the net really helped me to find black and white points. Click on the new fill or adjustment layer button on the layers pallet. At the bottom you will see threshold. When you click threshold, a panel will pop up. SLide the slider all the way to the left until the last little bit of black is showing, this will be your blackest spot in the picture. Set a marker. Now slide the slider all the way to the right. When there is just a little bit of white showing, set you marker there. When you have your two markers set, hit cancel. Don't make layer. The threshold dialog box will go away. This helped me a whole lot.
Also, don't know if it was discussed earlier, but you can bring up a curves layer and select the black eyedropper and select your dark marker. Then select the white eyedropper and click on your white marker. Then hit ok on the curves dialog box. Then proceed to make your S-curve with a second curves layer to bring out some contrast and "pop".
A tip I found somewhere around the net really helped me to find black and white points. Click on the new fill or adjustment layer button on the layers pallet. At the bottom you will see threshold. When you click threshold, a panel will pop up. SLide the slider all the way to the left until the last little bit of black is showing, this will be your blackest spot in the picture. Set a marker. Now slide the slider all the way to the right. When there is just a little bit of white showing, set you marker there. When you have your two markers set, hit cancel. Don't make layer. The threshold dialog box will go away. This helped me a whole lot.
Also, don't know if it was discussed earlier, but you can bring up a curves layer and select the black eyedropper and select your dark marker. Then select the white eyedropper and click on your white marker. Then hit ok on the curves dialog box. Then proceed to make your S-curve with a second curves layer to bring out some contrast and "pop".
You've just summarized what this discussion is about. If you go to the first post in this thread, you'll find the link to the tutorial that we're discussing!
Comments
This is the after....
015 retouched color.jpg
trying to figure it all out
welcome to dgrin
Nice shot and nice work with the pop!
However, for your subject I like the unpopped version a little better. You've done a fine job with the pop technique, I just like the softer feel of the unpopped version.
Looking forward to seeing more of your work!
David, thanks for the tutorial.
Here is my "before" popping image:
and here is the "after popping" image:
question: it looks as if contrast and saturation have both increased. - could simply increasing both in ACR or CS2 have achieved the same effect, or does this technique do something that goes beyond just boosting contrast and color? thanks
Cletus is right that portraits often call for a softer approach. I believe the answer for this shot lies somewhere in between the two treatments. You could put the popped version over the original and reduce the opacity, see what you get...
Dgrin FAQ | Me | Workshops
I'm not going to pretend that I understand enough to answer that question. I do know that adding pop (however you do it) is spreading out the values to take full advantage of the range available. This could be part of it. Setting the black point like that neutralizes it, which could effect colors, and lastly, when working in RGB it is difficult to separate these changes because of the way the color mode works. Color and contrast are both represented in the same channels, so shifts occur.
If you are interested you could read Rutt's excellent tutorial on adding pop in LAB mode, where luminosity is separated from color information. That tute is here.
Dgrin FAQ | Me | Workshops
The challenge with this image is that you want to add contrast to some areas (face, hair and skin), but you can't afford to raise the brightness of some of the brighter areas like the brightest areas of the dress or the background which would normally happen with the simpler means of adding contrast. You also have to make sure you don't go too far with skin.
There are several ways to accomplish this. The most common would be to just use a soft brush and paint a loose fitting mask over the brightest areas to just shield them from your changes. It might also possible to make the changes differently to avoid influencing the brighter tones and it is also possible to use blend-if settings to keep the change from the brighter areas. If any of these sounds too complicated, just stick with a simple mask over the dress and background.
Here's an in-between that keeps the changes off the dress and background. The changes are more subtle, but the goal was to bring out color and detail in the face and hair.
Homepage • Popular
JFriend's javascript customizations • Secrets for getting fast answers on Dgrin
Always include a link to your site when posting a question
CRW_1649_small.jpg
_CRW_1649_small.jpg
A major improvement, I suppose I'll have to read the tutorial if I want to play. A couple observations though. Isn't there any texture that can be salvaged in the black outfit? It really stands out as a distracting black hole. Also, the edges of the taller person's head have a bit of a cut-out quality. I might add a little noise to the edge to soften it up, blend it in. Finally, depending on how much effort is warranted, I would "powder" the faces to hold back the shine.
http://blue-dog.smugmug.com
http://smile-123.smugmug.com
http://vintage-photos.blogspot.com/
Canon 7D, 100-400L, Mongoose 3.5, hoping for a 500L real soon.
There's a lot of texture in the black dress. I assigned a combination of profiles at lower gamma settings than sRGB, converting back to sRGB each time. Then used Blend if... sliders to isolate the effect to the darkest regions in the original layer, masked out the odd stuff that always shows up with Blend if... sliders, then adjusted the contrast in that layer.
Some stock LAB moves for color and brightness.
—Korzybski
Dgrin FAQ | Me | Workshops
Then go from there, maybe using Shadow/Highlight to pick up some detail in the shadows, or curves for brightening a bit. Sound reasonable?
"Auto" anything is usually not a great idea. Each image requires different areas and ranges to be tweaked. Auto moves are a sledge hammer where a scalpel is called for. Curves are your first line of defense, followed by Selective Color and Hue/Saturation. Shadow/Highlights is always useful but you want to make sure you target only the areas that really need it.
Beyond that there are a whole host of techniques involving sharpening, forays into LAB space and plate blending. Lots of good info in this vein is wrapped up in the two discussion forums dealing with Dan Margulis' books.
—Korzybski
I'm brave so I'll add 1.
First, unprocessed shot.
Regards,
Peter
I think you made all the rigtht moves. I've learned, however, that if a move looks good at 100% this afternoon, it will look a lot better at 70% tomorrow morning. The tendency is to push things incrementally and so lose track of how far a move has gone. The colors strike me as a little bit harsh but that might just be the slightly excessive sharpening. I'm not sure that those beautiful old buildings and bricks would come up that bright and neutral in the real world, however.
Like I said, every move feels right, just a little too far.
—Korzybski
You definitely made a significant improvement on your image. There are several different ways you can go on the color of the original and I thought I'd show you a different approach that has a warmer tone (e.g. less blue). You can see which you prefer.
On this one, I only did two things to your original. Because the histogram looks like this:
With almost nothing on the right edge and no important details in the little tiny tail to the right, I applied this curve to enhance the contrast (the exact same move can be done in levels since all I did was push in the ends):
And, then sharpened a little to end up with this result:
More can be done (the curve could have a little s-shape for even more contrast, the sky could be dealt with separately to deepen the blue, etc...), but 90% of the effect comes from the simple part so I thought I'd just show that.
Homepage • Popular
JFriend's javascript customizations • Secrets for getting fast answers on Dgrin
Always include a link to your site when posting a question
Anyway, I clicked the gray dropper on your original, slightly blue "popped" image, and got this (the red dot being the approximate area where I clicked):
Dgrin FAQ | Me | Workshops
I'm only just developing my PP workflow and now I've added a step at the end to:
1. go away from the monitor
2. have a beer (optional, but encouraged)
3. come back and re-assess whether all the colors are possible and realistic/desired.
I like the simplicity of your method at the end, so that's one added to the memory bank.
Thanks to the others that responded as well.
I've really only been doing anything with PP for 4 days now, having previously just clicked auto levels and then blindly applied an S-Curve to the master RGB.
So, I think after 4 days of taking color correction seriously, I've definitely been able to produce something that is;
1. better than the starting image; and
2. better than the simple method used before.
The sharpening here was something I was aware of, but thanks for the comment. I applied the sharpening on a separate layer in the larger version and then after re-sizing for the forum, I lazed out and just reduced opacity rather than trashing the layer and repeating. Should have taken the extra care and that's another one for the memory bank.
I'm hoping that in another 4 days I can repeat this image from scratch and come up with an even better image and then again in a few weeks.
Thanks all for the comments.
Regards,
Peter
Also, don't know if it was discussed earlier, but you can bring up a curves layer and select the black eyedropper and select your dark marker. Then select the white eyedropper and click on your white marker. Then hit ok on the curves dialog box. Then proceed to make your S-curve with a second curves layer to bring out some contrast and "pop".
You've just summarized what this discussion is about. If you go to the first post in this thread, you'll find the link to the tutorial that we're discussing!
Dgrin FAQ | Me | Workshops
Keep up the great work!!