Photoshop Lab Color: Ch. 1 [Sitemap] - Digital Grin Photography Forum

rutt

Sep-09-2005, 11:20 AM

This chapter describes the most cornerstone technique of LAB colorspace and then explains why it is effective and why it is eitehr much harder or impossible in other colorspaces.

This technique is:

Convert the image to LAB
Curves
Steepen A and B channels symetrically by bringing in the endpoints of each curve toward the center equally. After this, the curve (line actually) will still cross the center horizontal center at the vertical center.
Steepen the L curve through the areas where the detail is of most interest.
Apply the curves
Activate only the L channel
USM, trying the values 200, 1.0, 10

In the course of teaching this, Dan also points out perhaps the most important tool for writing good curves.

With the curves dialog open, you can mouse over the immage with the (left) mouse button held down. A point will appear on the curve showing exactly the point on the curve of that point on the image. By moving the mouse accross an area of interenst in the image, you can figure out which part of the curve controls that area.

Dan uses a few images of canyons and one of Yellowstone lake in the fog to illustrate. I'll repeat with an image of my own.

http://rutt.smugmug.com/photos/35348119-L.jpg

I took this just before 6pm. When I took it, it felt like early evening. But the image seems flat and there is little sign of that great sweet light that I remembered.

I followed Dan's recipe, steepening both the A and B curves by pulling in both endpoints by 16 points:

http://rutt.smugmug.com/photos/35344737-L.jpg

http://rutt.smugmug.com/photos/35344866-S.jpghttp://rutt.smugmug.com/photos/35344869-S.jpg

Now some of the color I remember is restored to the sky and the sand is warmer and, well, more sandy just as the dune grass is greener. But the image is still a lot paler than my memory of it. Using the mouse on the L curve, I discovered that the areas of interest are in the midtones and that my shadows aren't nearly at the top of the curve. Also my highlights aren't very close to the bottom of the curve. So there is lots of unused contrast which can be applied to the areas of interest in this image:

http://rutt.smugmug.com/photos/35344785-L.jpg

http://rutt.smugmug.com/photos/35344860-S.jpg

Just as the A and B curves increased the color differences, this curve increases the brightness differences, increasing the contrast. It's really the same principal

Lastly, I sharpened the L channel, using Dan's parameters:

http://rutt.smugmug.com/photos/35344849-L.jpg

http://rutt.smugmug.com/photos/35344872-S.jpg
http://rutt.smugmug.com/photos/35344876-S.jpg

Voila!

What's going on? For me the interesting part of this chapter is the theory behind it. Granted the technique is a super simple way of improving tons of shots, but understanding why it works opens the door to a greater understanding of how to use (and not misuse) LAB color corrections and of digital color correction in general. Human sight is a marviously complex and highly evolved system. It self calibrates to emphasize differences in color and in shade. When we are in the middle of the forest surrounded by infinite subtle shades of green, our vision pulls these shades apart allowing us to see enormous variation. On the beach in the late afternoon, we can discern traces of the impending sunset and see the complexity of hue in the sand and sky. But the camera doesn't do this. It doesn't "know" when to do it. When we look at a picture, our vision does not self calibrate because the picture only occupies a small portion of our field of view and is surrounded by all kinds of competing cues. So the picture won't capture the same intensity and variety of color we saw when we took it. Even if it is a faithful rendering of the "true" colors of the scene, it does not capture what we saw.

The basic LAB enhancement can restore the picture so that it captures the vision of our memory. Steepening the A and B curves moves colors apart from each other. This kind of symetrical curve steepening applies a constant multiplier to the colors. If it was X amount of green before, it will be X*N green after and similarly for magenta, blue, and yellow. The green gets greener in proportion to how green it was to begin with. The difference between blue and yellow is magnified as slight variations from neutral are pushed away from neutral toward in the direction of their tint. So the sky of my lighthouse shot becomes more of a tapestry of blue and red. It looks more like the sky just before sunset, more they way I remember it.

Steepening the L curve over the areas of interest in the image similarly apes the way our vision works. The steeper the curve, the greater the detail, as Dan likes to say. I found no interesting detail beyond darker than the point in the grass just to the right of the road. So I pulled the right end of the L curve in almost to that point. Then I steepened it through the midtones to bring up the detail in the sand, lighthouse, and sky.

Dan spends a certain amount of time comparing thte results of alternative non-LAB techniques for accomplishing the same thing. I don't want to recap these arguments in detail. If there is someone who really feels that it's important to be able to do this correction without using LAB, please raise the issue. Suffice to say:

This isn't going to be nearly as easy any other way.
Learning this technique will open the door for variations and different powerful uses of LAB. So why resist?

I guess there is one more thing worth mentioning. This 1st simple LAB technique works well for images which seem duller than our memories of the scene. The worse this difference, the better the technique can work and the more steepening is called for. But try applying to a picture that already has very bright distinct colors or to an image with were delicate subtle color variations are important (portraits) and the results won't be so fortuitous. Also if the image has a color cast, this technique will only make it worse.

OK, does someone else want to post some examples of using this technique? Have I missed or mistaken some points of theory?

Enjoy (LAB book discussion)

pathfinder

Sep-09-2005, 07:11 PM

"I guess there is one more thing worth mentioning. This 1st simple LAB technique works well for images which seem duller than our memories of the scene. The worse this difference, the better the technique can work and the more steepening is called for. But try applying to a picture that already has very bright distinct colors or to an image with were delicate subtle color variations are important (portraits) and the results won't be so fortuitous. Also if the image has a color cast, this technique will only make it worse.

Very good discussion John, you have defintely set the bar high.

One thing not mentioned is that rather than steepening the a and b curves and spreading out the hue, flattening the a and b curves should flatten out the color also if colors were overdone.

rutt

Sep-10-2005, 03:29 AM

Besides missing the part about curve flattening, I also missed a trick which Dan shows (off). Extreme A&B curve sharpening can bring out colors which aren't really there at all in the original. For example, here is a shot from this year's Boston marathon:

http://rutt.smugmug.com/photos/35423686-L.jpg

And here I have applied some some very extreme curves to get what Dan calls "the man from mars" effect:

http://rutt.smugmug.com/photos/35423705-L.jpg

http://rutt.smugmug.com/photos/35423566-S.jpghttp://rutt.smugmug.com/photos/35423567-S.jpg

Dan says that no client would want this and probably he is right since it's no longer the late 60s. But after blending this back into the original at 30%, I got this:

http://rutt.smugmug.com/photos/35423667-L.jpg

Now the blues introduced by the steep curves make interesting highlights on the too flat faces of the original, and perhaps some client would want.

Questions for discussion:

Could I have accomplished the same thing with less steep curves and no blend? Why?
This trick doesn't seem to work best with the darkest skinned people and not at all with light skinned people. Why?
The runner's shirts look awfully blue in the final version, probably too much for my client. Why? What could I have done about it?

rutt

Sep-10-2005, 03:49 AM

Here is one more topic for discussion. I think that people who don't buy into Dan's basic philosophy about photographic images versus vision are going to balk pretty quickly at his techniques. The discussion in Chapter 1 of Professional Photoshop covers this more directly than what I have read so far in the LAB book, but it's a pretty deep assumption in both books. In a nutshell:

The "truth" the camera captures is often only a pale shadow of our vision. Even the best shot images often fail to capture the colors and contrast of our vision. The objective truth the camera captures is very different from what we see. Good postprocessing can exegerate the colors and contrast (among other things) to bring the image closer to our original vision.

Opinions?

Khaos

Sep-10-2005, 10:13 AM

The "truth" the camera captures is often only a pale shadow of our vision. Even the best shot images often fail to capture the colors and contrast of our vision. The objective truth the camera captures is very different from what we see. Good postprocessing can exegerate the colors and contrast (among other things) to bring the image closer to our original vision.

Opinions?
I agree completely. The brain deciphers what the eye gives it, so we don't always see what is truly there, thus we have those optical illusions.

Also, photos tend to, like music, have an emotional attachment to them. They talk to us through our eyes and if we need to adjust the colors and shadows to get the feeling across we should. Photography shouldn't be about getting a technically great shot every time, it should be about the photographers interpretaion of what he\she is shooting.

DavidTO

Sep-10-2005, 10:25 AM

Thanks for the in-depth summary, Rutt. Looking forward to getting my copy and joining in...and I'm a bit nervous about meeting the high standards you've set in your summary.

rutt

Sep-10-2005, 10:33 AM

Thanks for the in-depth summary, Rutt. Looking forward to getting my copy and joining in...and I'm a bit nervous about meeting the high standards you've set in your summary.

Don't worry. This chapter was the easiest chapter and the material was very familiar to me. Moreover, this wasn't even the first time I've written an explanation of this particular technique. Looking forward, I'm sure we'll all get out of our depth and the playing field will be level.

Right now, I'd like to see some practice with the basic technique from Ch. 1. Before and afters of shots were it really doesn't work are just as interesting as ones where it does.

DavidTO

Sep-10-2005, 10:56 AM

How do you get your curves dialog to show percentage change instead of absolute value?

DavidTO

Sep-10-2005, 12:11 PM

OK, here's my first.

Here's the curves I applied:

http://davidrosenthal.smugmug.com/photos/35460242-M.jpghttp://davidrosenthal.smugmug.com/photos/35460243-M.jpghttp://davidrosenthal.smugmug.com/photos/35460244-M.jpg

The straight RAW conversion:

http://davidrosenthal.smugmug.com/photos/35460295-L.jpg

And the result:

http://davidrosenthal.smugmug.com/photos/35460389-L.jpg

DavidTO

Sep-10-2005, 01:23 PM

K, one more for now.

Original RAW conversion:

http://davidrosenthal.smugmug.com/photos/35473524-L.jpg

a and b curves:

http://davidrosenthal.smugmug.com/photos/35473482-M.jpghttp://davidrosenthal.smugmug.com/photos/35473486-M.jpg

The result:

http://davidrosenthal.smugmug.com/photos/35473552-L.jpg

Lightness curve:

http://davidrosenthal.smugmug.com/photos/35473491-M.jpg

And the final:

http://davidrosenthal.smugmug.com/photos/35473577-L.jpg

Now, I may be getting ahead of myself with this one. You'll notice that I didn't adjust the a and b curves symmetrically. This is one of those sweet light shots that we were talking about in another thread. I found the first to magenta, so when I adjusted the a/b curves, I just guessed about what I should be doing and got a color that I found more pleasing, but still seemed sweet to me. The original RAW conversion definitely would have left her looking sunburned. How'd I do?

I'm also not sure about the lightness adjustment I made. Is it too much? There's a very fine line between giving the image the pop it deserves and preserving the lovely softness of the light.

BTW: that's Lynnesite's horse Ember on a visit a year ago this month. Gotta back out to see her and her horses, and soon!

rutt

Sep-10-2005, 01:54 PM

How do you get your curves dialog to show percentage change instead of absolute value?

It seems you figured this out. But for others, PS has this annoying property. If the curves are set up with darkness on the right, they show percentages, if the other way, they show absolute values. I wish these were independent options. Perhaps they are and I don't know how to control them.

ginger_55

Sep-10-2005, 03:00 PM

how did you get percentages????

I did a whole bunch of photos, just because I needed to work some stuff up anyway. But I probably didn't do it right. Mine were not "off", they sometimes needed the punch I get in selective colors. If I really wanted a photo, I did that, too.

So I could not get the percentages.
I just subtracted whatever the number was on the page.

And I would need a whole tutorial on how to show those charts and things, but I can show you some photos worked up.

I will show you one from RAW, just LAB and with LAB and selective colors.

ginger

ginger_55

Sep-10-2005, 03:22 PM

from RAW

http://gingersnap.smugmug.com/photos/35487829-L.jpg

From LAB

http://gingersnap.smugmug.com/photos/35487846-L.jpg

From RAW plus selective colors with about +10 in both red and yellow

http://gingersnap.smugmug.com/photos/35487850-L.jpg

I didn't have any where it didn't work, what I had were a few where I wanted a little punch in certain areas, such as above and below.

RAW below

http://gingersnap.smugmug.com/photos/35487842-L.jpg

Just LAB below:

http://gingersnap.smugmug.com/photos/35487835-L.jpg

LAB below plus selective colors w about +10 in red. Then I used the burn tool on her lips and eyes, along with the sharpening tool, lightly on her eyes and hairl

http://gingersnap.smugmug.com/photos/35487840-L.jpg

ginger_55

Sep-10-2005, 03:41 PM

I wanted to see how this would work. I am trying to better my bird shots since I seem to like them all by myself. This is a shot of a bird in front of a decrepit shack that I love. I worked it up last week, it was not the best, but I thought maybe it could be better.

This was last week, CMYK black channel, but no LAB:

http://gingersnap.smugmug.com/photos/35148334-L.jpg

This was today with just LAB

http://gingerSnap.smugmug.com/photos/35490056-L.jpg

This was the same as above with, again, selective color added as punch in the yellows and red. I use yellow a lot with grass and with a bird w yellow on it. I use red to enhance wood and grass.

http://gingerSnap.smugmug.com/photos/35490052-L.jpg

Below is a bird in flight, I wanted to see what would happen there. This is with just the RAW below:

http://gingersnap.smugmug.com/photos/35490069-L.jpg

This was the above bird today w LAB plus selective colors upped the yellow, plus I added a cooling filter which improved the sky, I thought.

http://gingersnap.smugmug.com/photos/35490063-L.jpg

All full frame again.

ginger

rutt

Sep-10-2005, 04:23 PM

Ginger, it's hard to understand what you did. Please show your work as David and I did.

rutt

Sep-10-2005, 04:31 PM

David, the bird picture works, but I might have been braver with the A+B steepening.

With the horse picture, you definitely did skip ahead a little too fast. You have introduced a color cast. The neutral horse is now yellow and to me looks like he belongs in the merry-go-round at Disney Land. That's the reason for keeping the center point of the curve in the center: it prevents neutral colors from acquiring a cast. In this case, I don't think there was a cast, but the child's face is too light and may need a little less magenta and/or more yellow. Try steepening the L curve through the area of the girl's (and horse's) face, the highlights. Try splitting the difference and making the A curve a but flatter and the B curve steeper untl you get good CMYK readings on her flesh (Y >= M).

Soon, you'll find out how to do better, but for now keep the curves symetrical.

DavidTO

Sep-10-2005, 04:40 PM

With the horse picture, you definitely did skip ahead a little to fast. You have introduced a color cast. The neutral horse is now yellow and to me looks like he belongs in the merry-go-round at Disney Land. That's the reason for keeping the center point of the curve in the center: it prevents neutral colors from acquiring a cast. In this case, I don't think there was a cast, but the child's face is too light and may need a little less magenta and/or more yellow. Try steepening the L curve through the area of the girl's (and horse's) face, the highlights. Try splitting the difference and making the A curve a but flatter and the B curve steeper untl you get good CMYK readings on her flesh (Y >= M).

Soon, you'll find out how to do better, but for now keep the curves symetrical.

The flesh tones definitely had too much magenta in relation to the yellow, that's why I did that. This shot was taken just after the sun had set behind a hill, so it was very warm.

When I steepened the L curve in her face, it got too contrasty and harsh.

I guess the part that's getting ahead is exactly how to give her more yellow/less magenta in LAB.

ginger_55

Sep-10-2005, 05:58 PM

Rutt, I have no idea how to show what I did. I never have known how to show my work space, and I have commented about that in the past and asked for help.

However, I know what I did, I know what you all did. It is working for me. Once I get started, I just keep working things up, unless there is a problem.

If someone wants to show me how to show a work flow, I will, but not often, and not right now.

I would probably do it once. It is quite simple though. I printed out the pages of your tutorial and followed them, as best as I could, with each photo. The only thing I did differently was to use the numbers and subtract, as I didn't know how to get the percentages.

ginger

PS, it is OK, you can just ignore my stuff up there. I am pleased, smile.

gus

Sep-10-2005, 06:36 PM

Until David sent me 'the photoshop CS2 book for digital photographers' by scott kelby.

I couldnt believe how flat this shot was i took yesterday so i sat down with the book & took it through the 18 steps on colour correcting using the darkest & lightest points (the 2 sheds in the foreground)....only took about 5 mins.

No massive difference but its something that improved the shots colour. Got rid of that purple look also.

Thanks David :thumb

Before

http://wadjelaphotography.smugmug.com/photos/35422251-M.jpg

After

http://wadjelaphotography.smugmug.com/photos/35515044-M.jpg

DavidTO

Sep-10-2005, 06:37 PM

Gus,

Check out this thead (http://www.dgrin.com/showthread.php?t=18308) and try it out on that shot.

gus

Sep-10-2005, 06:46 PM

Hence my issues mate...whats LAB ? I have to have the book open & follow it through step at a time.

DavidTO

Sep-10-2005, 06:52 PM

Hence my issues mate...whats LAB ? I have to have the book open & follow it through step at a time.

LAB is a color mode, like RGB.

RGB: Red, Green, Blue
CMYK: Cyan, Magenta, Yellow, Black
LAB: Lightness, A, B....yeah, I know the logic falls apart there.

But read the first post in the thread, it should tell you everything you need to know. If not, ask a question.

You enter LAB mode by menu Image>Mode>LAB. When you're done, go back to RGB Image>Mode>RGB.

DavidTO

Sep-10-2005, 07:01 PM

Here's a rough try with your M image.

I pulled in the A+B curves 15 points on either end (85/15) and adjusted the curves on the Lightness channel slightly.

gus

Sep-10-2005, 07:05 PM

Here's a rough try with your M image.

I pulled in the A+B curves 15 points on either end (85/15) and adjusted the curves on the Lightness channel slightly.
Ok...you convinced me...i will give it a burl & report back.

DavidTO

Sep-10-2005, 07:11 PM

Rutt,

Another question I have is, how do you ensure that you aren't making colors that aren't real? I know that LAB can make colors that don't really exist. I've tried proofing with sRGB and CMYK, but I don't really understand how the Gamut check thing works. When I change the mode back to RGB do I automagically lose the colors that RGB can't represent? That would make sense, since that's my understanding of what happens if you go through CMYK; it's lossy.

It just seems that I should be able to make colors in LAB that are so vibrant as to be beyond the gamut of RGB and, of course, CMYk.

gus

Sep-10-2005, 07:12 PM

OK got the curves worked out (L being for lightness i assume) but where is the USM ?

DavidTO

Sep-10-2005, 07:14 PM

OK got the curves worked out (L being for lightness i assume) but where is the USM ?

That's Unsharp Mask, under sharpen in the Filter Menu. I actually use Smart Sharpen in CS2, but either is fine. And you should probably stick with Rutt's recommendations for now, he understands this much better than I do.

gus

Sep-10-2005, 07:29 PM

Hmmm...any better ?

http://wadjelaphotography.smugmug.com/photos/35522757-M.jpg

DavidTO

Sep-10-2005, 07:43 PM

Hmmm...any better ?

http://wadjelaphotography.smugmug.com/photos/35522757-M.jpg

To my eye you could be more aggressive with the A+B curves, making the color more intense.

DavidTO

Sep-10-2005, 07:53 PM

Rutt, I have no idea how to show what I did. I never have known how to show my work space, and I have commented about that in the past and asked for help.

Ginger, you may want to start another thread asking the Windows folks (can't help you, I'm a Mac guy) how to do screen and window captures. It should be fairly easy to do.

pathfinder

Sep-10-2005, 07:55 PM

The flesh tones definitely had too much magenta in relation to the yellow, that's why I did that. This shot was taken just after the sun had set behind a hill, so it was very warm.

When I steepened the L curve in her face, it got too contrasty and harsh.

I guess the part that's getting ahead is exactly how to give her more yellow/less magenta in LAB.

Dave, the background shifted from a neutral grey to a greenish cast also behind the horse. The flesh tones seemed ok, but their def was a color shift.

I tried John's recipe taken from chapter 1 on a wedding shot I did last weekend, just for grins to see what effect it would have on people rather than vegetation or canyon walls like Margulis talked about in chapter 1.

John's post:
"This technique is:

1. Convert the image to LAB
2. Curves
3. Steepen A and B channels symetrically by bringing in the endpoints of each curve toward the center equally. After this, the curve (line actually) will still cross the center horizontal center at the vertical center.
4. Steepen the L curve through the areas where the detail is of most interest.
5. Apply the curves
6. Activate only the L channel
7. USM, trying the values 200, 1.0, 10 ""

This is exactly the recipe I used for the second image I am posting here. I did use a slightly different amount of USM as I was shooting 16 Mb files - so I used 200, 2.0 10 for sharpening

http://pathfinder.smugmug.com/photos/35408683-L.jpg

And the shot with the increase in the a and b curves - straight lines through the center, but one block steeper along the top and bottom row, minimal changes in the L channel.

http://pathfinder.smugmug.com/photos/35408490-L.jpg

Definitely more color in the flowers, the vegetation, and the skin tones. I was surprised by how much more I liked the second shot. The first seems to have a faint grey cast to it by comparison.

gus

Sep-10-2005, 08:02 PM

Before...

http://wadjelaphotography.smugmug.com/photos/35421385-M.jpg

After....

http://wadjelaphotography.smugmug.com/photos/35527015-M.jpg

DavidTO

Sep-10-2005, 08:10 PM

Nice.

DavidTO

Sep-10-2005, 08:13 PM

Dave, the background shifted from a neutral grey to a greenish cast also behind the horse. The flesh tones seemed ok, but their def was a color shift.

Well, I gotta figure out how to get the skin tones right without screwing up the background. The dirt may look neutral in the first, but the magenta is too high on the flesh tones. How do I remedy that without screwing everything else up?

ginger_55

Sep-11-2005, 02:51 AM

Thanks, David. Those colors, they are real, they just are exagerrated versions of what is there. They can be brought out by the manipulation. The problem comes that they can't be printed. They are too something. Saturated, strong, whatever.

The same thing can happen in RGB only with over saturation and things, probably curves there, too. I am quite careful with curves, or anything I do. Less is better lots of times. I don't know how the gamut thing works either, but you can have a bit off, I did read that, just not a lot.

The colors are real. Look in Rutt's first photo, the blue is there. I can see it before it is magnified. I think his yellows are magnified, too. Actually everything is. I, personally, don't like the colors, but............ they were there in the first place, just not as intense.

I seriously doubt that the colors go back to being printable, or change, by going back to RGB. They can "exist" in RGB, too. For the calendar, I had one photo that was not really printable. The green was too saturated. Instead of working more on that photo, I went to a different one. Greens are not my favorite color to work with. That was an RGB issue.

ginger

re how to show my work flow, I don't know what I am going to do about it. I think it is obvious from what I said as to what I did. I was just working yesterday to keep my mind off of stuff here, so it really isn't important unless you all really want to know something.

There are very few people whose directions I understand on the list, if the wrong person answered me, I would be worse off than before I started as everyone would just be more frustrated.

ginger_55

Sep-11-2005, 03:04 AM

Dave, the background shifted from a neutral grey to a greenish cast also behind the horse. The flesh tones seemed ok, but their def was a color shift.

I tried John's recipe taken from chapter 1 on a wedding shot I did last weekend, just for grins to see what effect it would have on people rather than vegetation or canyon walls like Margulis talked about in chapter 1.

John's post:
"This technique is:

1. Convert the image to LAB
2. Curves
3. Steepen A and B channels symetrically by bringing in the endpoints of each curve toward the center equally. After this, the curve (line actually) will still cross the center horizontal center at the vertical center.
4. Steepen the L curve through the areas where the detail is of most interest.
5. Apply the curves
6. Activate only the L channel
7. USM, trying the values 200, 1.0, 10 ""

This is exactly the recipe I used for the second image I am posting here. I did use a slightly different amount of USM as I was shooting 16 Mb files - so I used 200, 2.0 10 for sharpening

And the shot with the increase in the a and b curves - straight lines through the center, but one block steeper along the top and bottom row, minimal changes in the L channel.

Definitely more color in the flowers, the vegetation, and the skin tones. I was surprised by how much more I liked the second shot. The first seems to have a faint grey cast to it by comparison.

PF, there is definitely more pop in the second one.

You, also did not show your workflow, and I am sure you know how, don't you?

The only thing I would be interested in knowing, except as to how aggressively you did the a and b curves, I was not aggressive about it. I went about 3/4 of the way to the first "line", at the most. I know that I was not completely symmetrical about it, in other words, the top was not necessarily the same as the bottom as to how far over I went, as my mind had not grasped that in Rutt's recipe.

Also, I think the lighten curve is more subjective. Rutt's look strange to me from what I have been doing in RGB, and it would drive me nuts to have to keep the center point all the time. I don't quite understand the words "flatten", "steepen", though I did try. I thought it was interesting that you could find your area of interest by using the cursor. I was not quite certain what to do about it, but I did try to do what I thought was "steepening" at around those points.

One, PF, did you start working on your wedding photo right out of RAW? Did you do much in RAW? Your shots, like mine, only look like a matter of "degree", I do like the second one better, but it just looks like a bit more pop than the first one.

ginger

rutt

Sep-11-2005, 03:25 AM

how did you get percentages????

Click on the bar at the bottom of the curves dialog. It will reverse the scale of darkness to lightness (or warm to cool in A or B). At the same time it will show percentages instead of absolute numbers.

Dan likes the curves with darkness on the right and I've gotten used to that as well. PS doesn't have consistent defaults. RGB defaults to lightness on the left and LAB and CMYK are the opposite. Dan always has lightness on the left.

rutt

Sep-11-2005, 03:33 AM

Rutt, I have no idea how to show what I did. I never have known how to show my work space, and I have commented about that in the past and asked for help.

I don't have a windows box at hand but I did a google search for "windows screen snapshot" and found this:

http://www.google.com/url?sa=t&ct=res&cd=4&url=http%3A//www.softpedia.com/get/Multimedia/Graphic/Graphic-Capture/MWSnap.shtml&ei=ahQkQ7ONGsbSaPjwuMAD

It also seems that there is a builtin function to copy a particular window to the clipboard:

http://www.cardbox.co.uk/tech_screen.htm

Once you do this, you should be able to past it into a PS window and then save.

In either case, once you have a graphics file just upload to smugmug and then use in you posts just like a smugmug photo.

ginger_55

Sep-11-2005, 03:36 AM

http://gingerSnap.smugmug.com/photos/35558606-L.jpg

I took your version, the first one, David and uploaded it. Then I applied the a curve, symmetrically. The B curve is where the sweetlight and color shift comes in, IMO. I have not changed from numbers to percentages, so when I talk top and bottom, it might confuse you. But, in essence, I decided to put some sweet light in (I can't actually put it in, I balanced the curve so that the warmer colors, or the yellow, was more pronounced). I moved the "top" part of the curve in, it was in moving the "bottom" part of the curve that I realized that was where the yellow was. I did not move it symmetrically to the top part. (In other words, it is not moved over as far as the top, I don't think. They were not the same) I eyeballed it. I am just messing around, so that is what I did. I made it so a bit of sweet light showed up, but not a lot.

I loved your first photo, David, so I think it is a matter of taste as to which one prefers. I don't really care much about "truth" in colors, I care about "preference" in colors.

On the L curve, I moved them in pretty symmetrically, but I don't like too much contrast, particularly since I already liked the contrast that was there. I ran the cursor over it, the skin tones are always my area of interest with a shot of a person. They were towards the top. I moved that a bit to the left, I think.

No pictures of the work flow, but it was slightly eyeballed. I don't know what was true. I think there is more green in the background in my version, but that is just how I see it.

I forgot to sharpen it.

rutt

Sep-11-2005, 03:38 AM

rutt

Sep-11-2005, 03:41 AM

Well, I gotta figure out how to get the skin tones right without screwing up the background. The dirt may look neutral in the first, but the magenta is too high on the flesh tones. How do I remedy that without screwing everything else up?

Patience. We'll get there soon enough. For now, concentrate on solving the puzzle of finding symetric curves which decrease magenta while increasing yellow. All the clues are here.

ginger_55

Sep-11-2005, 03:42 AM

I don't have a windows box at hand but I did a google search for "windows screen snapshot" and found this:

http://www.google.com/url?sa=t&ct=res&cd=4&url=http%3A//www.softpedia.com/get/Multimedia/Graphic/Graphic-Capture/MWSnap.shtml&ei=ahQkQ7ONGsbSaPjwuMAD

It also seems that there is a builtin function to copy a particular window to the clipboard:

http://www.cardbox.co.uk/tech_screen.htm

Once you do this, you should be able to past it into a PS window and then save.

In either case, once you have a graphics file just upload to smugmug and then use in you posts just like a smugmug photo.

Thanks, Rutt. You were posting as I was working on David's photo. I will work on showing the workflow for something later.

g

rutt

Sep-11-2005, 03:44 AM

rutt

Sep-11-2005, 04:05 AM

WOW, 'Gus. David accomplished what I never could with you. Great!

'Gus, since you are a moderator, please would you mind moving the posts in this thread into the Ch 1 thread? They relate directly and I'd like to consolidate the discussion for later readers. Thanks.

rutt

Sep-11-2005, 04:08 AM

Before...

http://wadjelaphotography.smugmug.com/photos/35421385-M.jpg

After....

http://wadjelaphotography.smugmug.com/photos/35527015-M.jpg

'Gus, try steepening the B curve less. There is no rule that you have to steepen both A and B curves the same. The green of vegatation has both a green (A) and a yellow (B) component. Often you want to steepen the A cuve more than the B curve to make vegation more vibrant without making it more yellow.

rutt

Sep-11-2005, 04:24 AM

Also, I think the lighten curve is more subjective. Rutt's look strange to me from what I have been doing in RGB, and it would drive me nuts to have to keep the center point all the time. I don't quite understand the words "flatten", "steepen", though I did try. I thought it was interesting that you could find your area of interest by using the cursor. I was not quite certain what to do about it, but I did try to do what I thought was "steepening" at around those points.

This recipe does not call for the L curve to be symetric or even linear. Look at my L curve for the lighthouse shot. Writing the L curve for this recipe is less mechanical than the symetric steepening of A and B curves. As I wrote in my initial post:

With the curves dialog open, you can mouse over the immage with the (left) mouse button held down. A point will appear on the curve showing exactly the point on the curve of that point on the image. By moving the mouse accross an area of interenst in the image, you can figure out which part of the curve controls that area.

First master this technique. Bring up the L curve and move the mouse over:

The lightest points of the images
The darkest points of the images
The areas of greatest interest in your image.

Now bring the light endpoint of the curve inward not quite as far as the lightest points of the image. Bring the dark endpoint of the curve inward not quite as far as the darkest point of the image. At this point you have accomplished approximately what "Auto Levels" would do, increased the contrast with a linear (but not symetric) steepening of the L curve. Now experiment with some interior points to further steepen the curve through the areas of interest. This last part takes practice and a light hand. But it can be very effectve; this was the lesson of LAB and the Turner sky (http://www.dgrin.com/showthread.php?t=2042) In that shot, the whole goal was to make the curve steep through the highlights to bring out the subtle variations in the sky. Often, much more subtle curves are required. But take heart, this is a lot easier to do in LAB because brightness is given by a single curve L. Using the composite curve in RGB doesn't quite do the same thing as there are interactions between the channels.

ginger_55

Sep-11-2005, 04:55 AM

David, David's A curve is not symmetrical. I have been sitting in the living room obsessing on that information. That is probably what threw the color off.

I have also thought that bringing in the curves symmetrically should not cause a shift, but neither of us, David nor I, brought them in symmetrically. The A curve in David's case and the B curve in my case.

One thing I wanted to say, and this might be what you do not want me to say, Rutt, but doing this, moving curves, IMO, takes a lot of faith. One end of the curve changes what the other end does. It is difficult for our minds to follow a "recipe" and do this. I have to have faith and move both ends before making a judgement.

OK, now I can get my ten lashes.

Also, what is a clipboard? People keep mentioning that. When you all write a book of definitions, you might want to include that word.

g

If it is going to be self evident, in the googled sites, I can find out the definition myself.

I never knew what we were doing in Turner Sky. That is one reason it was so frustrating to me. I have learned a lot since then, but I have gaps in my learning, like fundamental definitions.

rutt

Sep-11-2005, 05:06 AM

David, David's A curve is not symmetrical. I have been sitting in the living room obsessing on that information. That is probably what threw the color off.

Exactly. If you keep the curves symetrical, you will not introduce change neutral colors, though you might emphasize a cast which already exists.

I have also thought that bringing in the curves symmetrically should not cause a shift, but neither of us, David nor I, brought them in symmetrically.

Exactly.

One thing I wanted to say, and this might be what you do not want me to say, Rutt, but doing this, moving curves, IMO, takes a lot of faith. One end of the curve changes what the other end does. It is difficult for our minds to follow a "recipe" and do this. I have to have faith and move both ends before making a judgement.

Right now we are just trying to learn one thing at a time so that we really understand it at a deep level. This symetrical A+B curve recipe isn't the best thing for all shots, or even most shots. But understanding it completely is the point of this thread and will form the foundation which we will build on. Very soon, we'll branch out. But for now, tunnel vision and faith are required. If you need to skip ahead, buy the book. Else, hang in there.

I never knew what we were doing in Turner Sky. That is one reason it was so frustrating to me. I have learned a lot since then, but I have gaps in my learning, like fundamental definitions.

Instead of fretting about asymetric curves before we get there, go back and reread the Turner Sky thread until you understand it. Ask questions on this thread and I will answer them. The real point of that thread is learning to write an L curve which is well within the scope of this chapter.

Nikolai

Sep-11-2005, 07:41 AM

I just got it last week, didn't have time to read through.
I LOVE LAB, and find myself to use it more and more lately. As a fromer mathematician, I have no psychological problems with "theoretical" spaces in general, and this one is practical enough to be immediately useful.:-)
Just wait till I get finished with "Canyon conundrum":-)
Cheers!:1drink

DavidTO

Sep-11-2005, 10:13 AM

Patience. We'll get there soon enough. For now, concentrate on solving the puzzle of finding symetric curves which decrease magenta while increasing yellow. All the clues are here.

I'll have to try adjusting the curves symmetrically, but not necessarily the same, which is what I've been doing. Whatever the a curve gets, that's what I've been giving b. When I'm symmetrical, that is. When I have time I'll try handling them each separately, but symmetrically.

rutt

Sep-11-2005, 10:15 AM

I'll have to try adjusting the curves symmetrically, but not necessarily the same, which is what I've been doing. Whatever the a curve gets, that's what I've been giving b. When I'm symmetrical, that is. When I have time I'll try handling them each separately, but symmetrically.

That's the right idea. Now I've started to reach Ch 3 and this is really the topic there. Still, this is a good puzzle which will help your understanding if you solve it. Symetric curves add yellow and subtract magenta.

ginger_55

Sep-11-2005, 10:38 AM

AAAAAAAAAAAAGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGHHHHH HHHHHHHH!

http://gingersnap.smugmug.com/photos/35601376-L.jpg

http://gingersnap.smugmug.com/photos/35601381-M.jpg

http://gingersnap.smugmug.com/photos/35601382-M.jpg"

http://gingersnap.smugmug.com/photos/35601375-M.jpg

That was a lot of work

ginger_55

Sep-11-2005, 10:47 AM

The first time I did it, the lightness curve was a bit more complicated. By the third try, it had deteriorated to THIS above.

g (And that was not the end of things, as you saw, Rutt) I think a clipboad would be nice.

rutt

Sep-11-2005, 10:49 AM

AAAAAAAAAAAAGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGGHHHHH HHHHHHHH!

Please translate into English. Is this good or bad? Are you frustrated with the result on this image?

The basic recipe here won't be good for every image. What has gone wrong here in you opinion? Can you guess why?

Consider more subtle curves and consider a non linear L curve to focus contrast on your subject instead of background.

ginger_55

Sep-11-2005, 10:55 AM

Please translate into English. Is this good or bad? Are you frustrated with the result on this image?

The basic recipe here won't be good for every image. What has gone wrong here in you opinion? Can you guess why?

Consider more subtle curves and consider a non linear L curve to focus contrast on your subject instead of background.
1) A Lot of damn work

2) I have no idea what went wrong, the colors are accurate as far as I can see.

3) I had a non linear L curve until I had done it twice with things not coming down to upload smugmug. By that time, I just did it as easily as possible.

to see if it would come down. However, I do not think that my non linear curves were that much different in the end results.

ginger

gus

Sep-11-2005, 12:15 PM

gus

Sep-11-2005, 12:19 PM

rutt

Sep-11-2005, 12:57 PM

Thanks rutt...when you blokes talk about steepening, how much are you doing it ?

Im just curling up the bit at the end...is that roughly what you all do ?

Also is there an easy way to curl it up without the whole line going stupid ?

Go back and look at my initial post. The simplest form of this technique is just to move each endpoint of the curve inward exactly the same amount. That's what I've been calling "symetrical steepeining." It's symetrical, because the endpoints move inward the same amount which keeps the center of the curve (line really) crossing the 0,0 point. This is important because it preserves the color balance of the shot. It will never make a color that was yellow into a blue color or a color that was green into a magenta. It only makes things that are already green (for example) greener, enhancing the differences different greens. It's linear, because only the two endpoints move and the curve remains a line. There are lots of reasons for nonlinear curves, but you can do A LOT with linear A+B curves so learn how to use them first until you get comfortable with them. The L curve is different. Try to find a neighborhood of the curve which encompases the parts of your image where you want to enhance detail and try to make that steep by pulling the curve down at it's start and up at its end. Easy does it. You may also want to try linear L curves which work almost exactly like auto-levels except you get to pick the white and black points.

The curve you posted isn't likely to do much. In the A and B channels, all the action tends to be near the middle of the curve. Small moves there will have a big impact on your image. The ends of the curve, where you have made your big change only impacts very very saturated colors. Very very very very saturated colors. (This is a big part of the subject of Dan's Chapter 2.)

Play with the A and B curves and don't be afraid. You can always undo or just cancel. Try moving the endponts quite a lot, (40%) Try reversing the endpoints by pulling the endpoint that is on the top to the bottom and visa versa. Dan suggests trying experimenting with increments of 10, 15, and 20 to get the hang of symetrical steepening. I often use numbers between 15 and 5.

DavidTO

Sep-11-2005, 02:55 PM

OK, so I reworked the shot of my daughter and horse. I messed up (going too fast) and didn't save my L curves, but here's the A+B:

http://davidrosenthal.smugmug.com/photos/35633264-L.jpghttp://davidrosenthal.smugmug.com/photos/35633270-L.jpg

Here's the result:
http://davidrosenthal.smugmug.com/photos/35633350-L.jpg

And after Lightness and L sharpening:
http://davidrosenthal.smugmug.com/photos/35633445-L.jpg

For reference, here's the original RAW conversion:
http://davidrosenthal.smugmug.com/photos/35473524-L.jpg

rutt

Sep-11-2005, 04:07 PM

OK, so I reworked the shot of my daughter and horse.

Perfect, given the constraints I set. And better than the previous effort which introduced a yellow cast.

gus

Sep-11-2005, 07:53 PM

Didnt you blokes say to keep A & B the same ?

DavidTO

Sep-11-2005, 08:12 PM

Didnt you blokes say to keep A & B the same ?

Apparently each one needs to be symmetrical at this point, but they do not need to be the same. I guess in the future we're going to learn how to adjust them asymmetrically, but for now that's the deal. So the A highlights are moved an equal amount as the A shadows. Same for B, but B does not need to equal A.

rutt

Sep-12-2005, 02:41 AM

In the book, Dan keeps A and B the same through most of Ch 1 and only at the very end of the most technical part does he give an example where he does not. At that point is arguing that this LAB adjustment can't be duplicated in any other colorspace.

Anyway, I just wanted to restrict the recipe as much as possible at this point. More options tend to confuse people at first and the simpler the recipe the more likely it is that people will try it out.

Chapter 2 is a pretty technical (and numerical) explanation of LAB. I'm going to write something about it this week. A few people have told me that it's intimidating. I don't think it has to be, so I'll try to help sort out what you really should know from stuff you can absorb over time.

Chapter 3 is where we'll start changing the recipe. David volunteered to kick off that discussion by writing a summary of it once he gets his book and gets through the first two chapters.

TristanP

Sep-24-2005, 07:49 PM

I don't know if this is the best image to try it on, but this technique has been intriguing me for awhile and I thought I'd try it tonight.

Original

http://home.comcast.net/~sarahpanek/dgrin/path.jpg

a+b curves

http://home.comcast.net/~sarahpanek/dgrin/a_curve.jpg
http://home.comcast.net/~sarahpanek/dgrin/b_curve.jpg

After a+b curves

http://home.comcast.net/~sarahpanek/dgrin/path_ab_curves.jpg

L curve

http://home.comcast.net/~sarahpanek/dgrin/l_curve.jpg

After L curve

http://home.comcast.net/~sarahpanek/dgrin/path_l_curve.jpg

Final after sharpening L only (100/1.0/10)

http://home.comcast.net/~sarahpanek/dgrin/path_final.jpg

DavidTO

Sep-24-2005, 08:00 PM

I don't know if this is the best image to try it on, but this technique has been intriguing me for awhile and I thought I'd try it tonight.

Original

http://home.comcast.net/~sarahpanek/dgrin/path.jpg

.........

Final after sharpening L only (100/1.0/10)

http://home.comcast.net/~sarahpanek/dgrin/path_final.jpg

How much processing did this image get before LAB? I'm curious, because the road is cool in the B value and warm in A. That makes for magenta and blue, which ends up looking kinda purpley.

But your greens are very yellow. So if you were to add yellow for the road you'd throw the vegetation out.

This image (IMO) is a more advanced problem. This requires work that won't be covered until Chap. 4 or later.

Aside from those challenges, what you did (given the constraints of where we are with this) made sense and improved the image (while exagerating some problems already in the image).

Try again with another image. It's good to work with an image that you know has an area that's neutral, or nearly so. Any neutral area will have A and B values of 0.

gefillmore

Oct-02-2005, 07:05 AM

rutt

just to let you know I'm still here--

trying to catch up on chapters 1-5 while working on chapter 6--

will take chapter 9 or 10 but you might want to wait til you see how I do on 6--

this is a pretty neat way to go thru a book, especially so technical, and get all the different viewpoints--

interesting how you've asked for us to take these directions on faith and how some have difficulty doing just that--

if any of these go to church I'm sure they're not the preacher's favorite--
I never was but it was for other reasons--

please don't get me wrong--nothing like some healthy skepticism and seeing all the different personalities coming thru--

my guess is that when I get the opportunity to look at the galleries of your respondents that the photos will reflect their personalities--

anyway, enjoying this and thanks to all who are participating--

george

gefillmore

Oct-02-2005, 07:11 AM

ps got called in to the hospital in the middle of the night (ultrasound tech) (always taking pictures) and did not get any sleep--

just took some ambien which makes me somewhat looney and really shows when I'm on a forum--had to quit the high school sports forum because I was probably way too embarassing to my kids--

so if I say something that offends, then f-f-forgive me, please--

rutt

Oct-02-2005, 07:14 AM

rutt

Oct-03-2005, 09:08 AM

You need to say more about what you did and what you were trying to accomplish here. Can you post the curves you used? Off hand, I'd say this image is also in need of a little TLC in the L channel.

Really these posts belong in one of the chapter discussions if we only knew which one.

gefillmore

Oct-03-2005, 09:37 AM

sorry--

meant to put this in chapter 1 dialog--

so engrossed in trying to get the photos posted that I lost track--

I can get too 'focused'--

use pc vs mac--for those who know, is it easier to post with mac??

how are you posting your curves?

does anyone know what a gates keyboard is?? (answer at bottom)

administrator--if you need to or want to move my last few posts, no problem with me--mainly making sure I can get the photos up

a keyboard with just three keys (ctrl-alt-del)

rutt

Oct-03-2005, 09:47 AM

how are you posting your curves?

I gave these instructions to Ginger and they seem to have worked for her:

http://www.dgrin.com/showpost.php?p=162050&postcount=38

gefillmore

Oct-03-2005, 10:44 AM

DavidTO

Oct-03-2005, 10:45 AM

trial run curves

Whoa. Check out those curves!

DavidTO

Oct-03-2005, 10:55 AM

ok--is everybody posting to smugmug including graphics files and then on to the forum?

I was using the manage attachments and it was very tedious--

oh well, I will keep at it but right now I need to get back to chapter 4--

thanks for the help--

george

I use attachments if I'm only attaching one file. But for the LAB posts you need to attach (usually) a before and after and three curves. So in that case I would upload to smugmug first.

In the future in order to show your LAB work show us what you started with, what you ended up with, and describe how you got there, showing the curves that you used to get there.

We can't be much help without that info, and others can't learn from your experience without it.

Keep on having fun. It's kind of addictive, ain't it?

Oh, and also if you need specific help with the mechanics of posting, feel free to PM me.

rutt

Oct-03-2005, 11:06 AM

We can't be much help without that info, and others can't learn from your experience without it.

I want to add something. It's very important when you retouch a photograph to have a goal in mind. If you don't, then at least when you are done you should be able to say what you acheived. Without at least one of these, it's very hard to access the technical success of your work. If you get in the habit of thinking in terms of goal directed photo retouching, you will learn more and do it more quickly than if you just fool around and try stuff until you think it "looks better".

gefillmore

Oct-03-2005, 12:23 PM

rutt and david--

thanks for the help and the advice--

yes, it's addictive--

george

photomomo

Oct-10-2005, 06:38 PM

The flesh tones look fine but there is too much green with everything else???

Stan

Oct-11-2005, 05:35 AM

I will try not to post too much while I catch up but if anyone can help with some early pointers...

1 The sky turned the nastiest blue in the lake and desert images and did not resemble the colour in the book. I take it this gets addressed later in the book.(?) I noticed it also in Davidto's correction of Gus' plantation

2 Sharpening the Lightness channel 200:1:10 made the rocks in the Yellowstone image far too sharp, has he done this for effect or have I missed something?

3 Is it better to process a RAW image before converting it to photoshop or to open the image as shot and correct it all in Levels?

Thanks for this fascinating course, I have only now found time to start it so am running very late...

Stan

Thwack

Oct-18-2005, 02:34 PM

I don't think the dry grass came out quite the right color but the clouds are a definite improvement after playing with the LAB curves.

The grass looks a bit washed out in the posted version but it looked much better in Photoshop. Converting it to JPEG for posting seems to have washed it out a bit (I had the quality set to max).

A & B were just pulled in 16% on each end. L was pulled in less on each end and a slightly non-linear curve was used so I could move the lighter and darker portions slightly differently.

I forgot to take a screen shot of my L curve and I don't seem to be able to bring that curve back up afterwards (once that window is closed, there's no way to slightly tweak an existing curve???). :scratch

Oh yeah, I made a couple other edits while I was in there and forgot to save to a different file name before adjusting the LAB curves. For the "before" shot, I had to go back to the original which is lacking some cropping, reframing, and some other edits (no color adjustments though).

"Before":
http://thwack.smugmug.com/photos/40561352-M.jpg

"After":
http://thwack.smugmug.com/photos/40561171-M.jpg

pathfinder

Oct-18-2005, 06:54 PM

There's more than just LAB curves here, as the image has been cropped and the telephone pole has been cloned out. But the overall tonality of the second image is much better to my eye. I like the color in the water much better too.

DavidTO

Oct-18-2005, 07:26 PM

Yep, looks better.

What else you got? ;)

Thwack

Oct-19-2005, 09:21 AM

I have several pics from that river trip that I'm trying to clean up but I'm really struggling with the others. If I can make them presentable, I'll toss them up here (and I'll try to remember to save a copy before playing with the LAB curves so I don't have to post a 'before' picture that's lacking the reframing, cropping, etc). :)

Thaanks for the comments folks.

Tom K.

Nov-20-2005, 09:56 PM

I have been reading this thread with great interest. I have a problem that has been driving me up the wall for years.....and it's the fact that I am color blind. I can see colors but I have trouble differentiating between greens, reds, and browns when they are all mixed up in an image. So this LAB method could potentially help me with a "by the numbers" color correction routine........hopefully.

I have been applying the methods in the thread (book is in the mail on it's way as I write this). I have come up with several dramatic improvements to images that have been sitting on my hard drive for a year or two now. The only thing I really have to figure out is the adjustments to the Lightness channel. Channels a and b I am OK with. Creating that final curve in Lightness has been tough.....mainly due to my color blindness.

Anyhow....this thread is the reason I registered with this superb photograph forum. I fell upon it with a good old google search. Here's a split image I did using the basic chapter 1 method.

http://i.pbase.com/o2/88/528788/1/52557462.LAB9.jpg

pathfinder

Nov-21-2005, 03:02 AM

rutt

Nov-21-2005, 07:19 AM

.....and it's the fact that I am color blind. I can see colors but I have trouble differentiating between greens, reds, and browns when they are all mixed up in an image. So this LAB method could potentially help me with a "by the numbers" color correction routine........hopefully.
The second (more technical) half of Chapter 3 advances the theory that red/green color blindness really is magenta/green, i.e., A channel, color blindness. Dan has some interesting empirical evidence to back this up.

Creating that final curve in Lightness has been tough.....mainly due to my color blindness.

When I took Dan's class two weeks ago, he made us correct a set of images with our monitors set to greyscale, so we had to rely entirely on the numbers for color info. The results actually looked pretty good in color. In some ways correcting like this is liberating because it removes the temptation to look at the image so hard and long that one can no longer see it. If the numbers are right, well if one can't see the colors, that's the best one c an do.

Here is an approach for writing those pesky L curves which may help you. Turn off the visibility of the A and B layers (click on the eyeballs next to them in the layer palette.) Now you are looking at a B&W image. Adjust the L curve to make it look the best you can. And there is a by the numbers approach to that. Select light and dark points in the image, spots where you are willing to blow anything lighter or plug anything darker, respectively. Bring up the L curve and mouse over those points while left-clicking. You'll see a point on the curve representing the luminosity of the image at that point. Drag the endpoints of the curve close to thoee points (but not all the way there, you'll see why.) Now mouse over the areas of greatest interest in your shot (say a face in a portrait) and consider adding a point to the curve to make it steeper over those areas: perhaps just to the dark side and pulled up a bit or just to the light side and pulled down a bit.

What some concrete examples? Why not post a couple you are having trouble with and I'll show you. Higher resolution is much better and easier, so please provide that if possible.

Anyhow....this thread is the reason I registered with this superb photograph forum. I fell upon it with a good old google search.

On behalf of all the participants in the reading group, I'm flattered. Thanks.

Tom K.

Nov-21-2005, 08:12 AM

Here is an approach for writing those pesky L curves which may help you. Turn off the visibility of the A and B layers (click on the eyeballs next to them in the layer palette.) Now you are looking at a B&W image. Adjust the L curve to make it look the best you can. And there is a by the numbers approach to that. Select light and dark points in the image, spots where you are willing to blow anything lighter or plug anything darker, respectively. Bring up the L curve and mouse over those points while left-clicking. You'll see a point on the curve representing the luminosity of the image at that point. Drag the endpoints of the curve close to those points (but not all the way there, you'll see why.) Now mouse over the areas of greatest interest in your shot (say a face in a portrait) and consider adding a point to the curve to make it steeper over those areas: perhaps just to the dark side and pulled up a bit or just to the light side and pulled down a bit.

What some concrete examples? Why not post a couple you are having trouble with and I'll show you. Higher resolution is much better and easier, so please provide that if possible.

I am thankful for the quick response and extremely helpful info. The "L" curve method you described looks very promising. I have posted a very large original image along with my final rendition with a screen cap of the curve I used. I am quite sure the "L" curve could be improved upon as this is my first attempt at using your suggested method. Hopefully after practice (and with your assistance) I can get the hang of that "pesky" L curve technique. I use the USM settings of 200 > 1.0 > 10. Ten I resized down to 1200 x 800 with no further resharpening. As noted above I am color blind. My wife isn't and said that the final image looks good. That said.....she did not analyze the image closely.
http://i.pbase.com/o2/88/528788/1/52573247.watch_hill_lg.jpg

http://i.pbase.com/o2/88/528788/1/52573249.watch_hill_final.jpg

http://i.pbase.com/o2/88/528788/1/52573251.watch_hill_LAB_curves.jpg

I am so very thankful for the help.

rutt

Nov-21-2005, 09:25 AM

Not exactly. Your L curve should look more like this:

http://rutt.smugmug.com/photos/45421410-M.jpg

What does this curve do? The horizontal (x) axis represents input L values and the vertical (y) axis represents output L values. The brightest spot I could find in the picture was in the clouds and it only measured about L=80. By moving the leftmost endpoint of the curve inward by about 15, I specified that anything with L=85 or more should become L=100, i.e., as bright as possible. The darkest area I could find in the original was in the roots of the grass about 1/3 in from the left. It measured about 10. By moving the right endpoint of the curve inward by about 10, I specified that this spot and anything darker than it should be as dark as possible (L=0). This L curve completely loses all contrast for anything lighter than that patch of cloud or darker than that place in the roots. But that's OK, because there is nothing else that light or that dark. That's how I picked those values in the first place. With this curve active, when I mouse over the picture with the left mouse button pushed, the point on the curve moves through the whole range of the steep part. The sky and clouds occupy the upper 1/3, the sand and tops of the grass occupy the midtones, and the roots live in the lower 1/3 of the curve. Now we are using the entire range of luminosity. Should look a lot better. Does it?

http://rutt.smugmug.com/photos/45421531-L.jpg
Full size (http://rutt.smugmug.com/photos/45421531-O.jpg)

jfriend

Nov-21-2005, 09:29 AM

Not exactly. Your L curve should look more like this:

http://rutt.smugmug.com/photos/45421410-M.jpg

What does this curve do? The horizontal (x) axis represents input L values and the vertical (y) axis represents output L values. The brightest spot I could find in the picture was in the clouds and it only measured about L=80. By moving the leftmost endpoint of the curve inward by about 15, I specified that anything with L=85 or more should become L=100, i.e., as bright as possible. The darkest area I could find in the original was in the roots of the grass about 1/3 in from the left. It measured about 10. By moving the right endpoint of the curve inward by about 10, I specified that this spot and anything darker than it should be as dark as possible (L=0). This L curve completely loses all contrast for anything lighter than that patch of cloud or darker than that place in the roots. But that's OK, because there is nothing else that light or that dark. That's how I picked those values in the first place. With this curve active, when I mouse over the picture with the left mouse button pushed, the point on the curve moves through the whole range of the steep part. The sky and clouds occupy the upper 1/3, the sand and tops of the grass occupy the midtones, and the roots live in the lower 1/3 of the curve. Now we are using the entire range of luminosity. Should look a lot better. Does it?

Full size (http://rutt.smugmug.com/photos/45421531-O.jpg)By way of making sure I what you're doing here, is this conceptually similar to "pushing in the ends" to the edges of the histogram in a levels adjustment layer?

rutt

Nov-21-2005, 09:32 AM

By way of making sure I what you're doing here, is this conceptually similar to "pushing in the ends" to the edges of the histogram in a levels adjustment layer?

Jeez, I don't know. I never use the levels histogram. Someone else will have to answer.

aporia

Nov-21-2005, 02:12 PM

By way of making sure I what you're doing here, is this conceptually similar to "pushing in the ends" to the edges of the histogram in a levels adjustment layer?
The concept seems similar to me. Margulis recommends changing the defaults of the pure white and pure black points permanently to 97L, 0A, 0B (white) and 6L, 0A, 0B to improve the auto adjustment commands in every colorspace.

Pushing or sliding in the ends in Levels or Curves also has a similar effect to using the eyedropper endpoint tools or adjusting the output numbers at the bottom of the histogram.

Now, if I can only figure out how to color correct by the numbers, I'll be home free. :D

DavidTO

Nov-21-2005, 04:34 PM

By way of making sure I what you're doing here, is this conceptually similar to "pushing in the ends" to the edges of the histogram in a levels adjustment layer?

If all you're moving is the ends, yes. Curves are much more powerful, but the ends would do the same thing as Levels.

rutt

Nov-21-2005, 05:18 PM

DavidTO

Nov-21-2005, 05:20 PM

You know, the odd way I learned to use PS shows. I learned it from reading Dan's Professional Photoshop. And one thing he hates is levels. He's a downright snob about it, and I guess I caught the disease from him.

Levels are good for simple moves. Easier than curves, but not much. If you know you're just going to set your black, gray and white points, or change your output levels, they're fine. Just not as elegant or versatile as curves.

pathfinder

Nov-21-2005, 06:36 PM

By way of making sure I what you're doing here, is this conceptually similar to "pushing in the ends" to the edges of the histogram in a levels adjustment layer?

Yes, I think it is. Rutt is finding the darkest spot and the lightest spot and using them to anchor a curve's outer limits. Pushing the sliders in frm the ends on the histogram is accomplishing the same thing without the ability to vary the slope of the curve, just the two endpoints.

Tom K.

Nov-21-2005, 08:16 PM

Should look a lot better. Does it?

http://rutt.smugmug.com/photos/45421531-L.jpg
Full size (http://rutt.smugmug.com/photos/45421531-O.jpg)

Bravo!!! Thank you for this educational experience. I'm getting this slowly but surely. You rendition is just perfect. I am curious as too how much sharpening you did to the final large image. Many thanks for this extremely generous and helpful guide you have provided. I'll be practicing this LAB technique until I can do it with confidence. I'll be the first to admit that make take a while.......but.......I'm determined to get there.

rutt

Nov-21-2005, 08:33 PM

I am curious as too how much sharpening you did to the final large image.

No sharpening at all in the interest of not complicating the lesson. It looks sharper because it has good contrast. Now if you go ahead and sharpen it carefully, it will look even better. But you'll need a light hand. Grass like that has a lot of fine detail and contrast and is easy to oversharpen. Might try the light/dark halos technique. (http://www.dgrin.com/showthread.php?t=9739)

Tom K.

Nov-21-2005, 09:22 PM

No sharpening at all in the interest of not complicating the lesson. It looks sharper because it has good contrast. Now if you go ahead and sharpen it carefully, it will look even better. But you'll need a light hand. Grass like that has a lot of fine detail and contrast and is easy to oversharpen. Might try the light/dark halos technique. (http://www.dgrin.com/showthread.php?t=9739)
Oh yes. I have been reading that sharpening method and gave it a try. It's the defacto technique IMO and although I'll need much more experience with it..........once mastered will be my sharpening method of choice.

I am still waiting for the Canyon Conundrum book to arrive from Amazon.com. I blew it when I choose free shipping. That extended my delivery time several days. :(

Tom K.

Nov-22-2005, 10:48 PM

rutt

Nov-23-2005, 08:20 AM

I took another image and gave it a go. In fact I have been going through image after image and applying the techniques in this thread. The fact that I'm color blind saps some of my confidence because in reality I just don't know whether the colors are accurate or not. My eyes lie to me. I have an image here that I wanted to get the contrast in the sky much more pronounced than I ended up with. It's the L channel that I believe is making me ineffective on this shot. If I may be so bold I am posting a link to the original photo and the downsized final image I finished with. I really wanted to improve the contrast in the overcast sky. I just couldn't figure out how to do it. Unfortunately I forgot to do screen caps of the curves I applied in a b and L channels.

I still have a lot to learn. Book should be incoming next week.

Getting detail in the sky while maintaining daylight below in this shot is well beyond the scope of Chapter 1 technique. I'll go through it, mostly as a plug for later chapters and to help you and others see the rewards for hanging in there.

Before I do, I want to reiterate somethng. You've come to the right place if you want to learn how to color correct in spite of color blindness; but, you've come to the wrong place if you want to use it as an excuse. The calibrated monitor, color management people will be very sorry for you.

Your color blindness doesn't cut it as an excuse for not being able write L curves. You still see contrast just fine, it's the A channel that you have trouble with. Do this: go to the Channels Palette and click on the L channel. Now you will see a B&W. Adjust the L channel to make it look good. No excuses. As for the color blindness, you've come to the right place. Learn the by-the-numbers technique and you'll be able to do a good job in spite of your disability.

OK. So what about that sky? The best you can do with Chapter 1 technique is just to make the L curve very steep in the highlights where the sky lives:

http://rutt.smugmug.com/photos/45691499-S.jpg

That gets you to here:

http://rutt.smugmug.com/photos/45691489-L.jpg

Nice sky, but the foreground is now way too dark. You need a way to treat the sky and foreground separately.

The easiest way is to take use the highlight half of the shadow/highlight adjustment before converting to LAB. It will bring back sky detail very easily and then you can go on and correct according to the Chapter 1 recipe:

http://rutt.smugmug.com/photos/45695998-M.jpg

http://rutt.smugmug.com/photos/45695976-S.jpghttp://rutt.smugmug.com/photos/45695980-S.jpghttp://rutt.smugmug.com/photos/45695986-S.jpg

with this result (before sharpening)

http://rutt.smugmug.com/photos/45696083-L.jpg

Well, a little Chapter 4 crept in there because I think there is a slight yellow cast in the highlights on the ground. But forget I said that.

You can get better control over the treatment of ground and sky by using the Chapter 8 technique of working on them in separate layers and then using blending options and perhaps a little layer masking to apply what you want to the parts of the image you want.

I started in LAB mode by making a duplicate of your image and then applying the L curve shown above to bring out the sky detail. Of course, this made the foreground too dark. So I opened the Blending options dialog for the duplicate layer (double click on the layer, carful not to click on its name.)
And moved the blend if slider for the L channel and the lower layer to exclude all but the lightest parts of the image from the blend:

http://rutt.smugmug.com/photos/45691496-M.jpg

After this, I added a layer mask and zapped a few annoying light spots in the foregrounds that were showing the blend. Then following the chapter 1 (well, a little chapter 4 here again) recipe to get this:

http://rutt.smugmug.com/photos/45691448-L.jpg

After lighten/darken USM, I got this:

http://rutt.smugmug.com/photos/45691543-L.jpg

rutt

Nov-23-2005, 08:26 AM

BTW, where was that picture taken. Looks so familiar.

Tom K.

Nov-23-2005, 08:41 PM

BTW, where was that picture taken. Looks so familiar. The photo was taken in Norwich, Connecticut. Your edit is exquisite.

Many thanks for the time and effort you have taken with my images and this advice you are providing me is solid gold. No more excuses from me. Full steam ahead.

Still waiting for the book to arrive (that's not an excuse ;) ).

vangogh

Jan-09-2006, 09:09 AM

edgework

Jan-09-2006, 10:21 AM

Hi

I've just read the tutorial for curve work & working with LAB colour & just had a couple of queries. I work as a designer & use photoshop CS2 everyday. I work mainly in CMYK or RGB if certain effects such as lens flare or whatever are being used. How do the comments for adjusting curves in LAB apply to CMYK or RGB? I normally adjust curves directly in CMYK, ususally on an adjustment layer. Should I change everything to LAB & then change back to CMYK? If so, how does the continual converting affect the final image for print?

Many thanks
Nicola
Iconic Creative
http://iconiccreative.smugmug.com

Keep reading. Good things await you.

I am an image retoucher working primarily in a prepress setting and so CMYK is the coin of my realm. However, since diving into the lab techniques so ably discussed in these various threads, and reading Dan Margulis' latest book (the source of these threads) my use of CMYK has drastically diminished—at the end of a job, primarily to fine tune skin color or work with the black plate. For all major contrast moves, color cast removal, color enhancement and sharpening moves (definitely sharpening moves!), I work exclusively in LAB, converting back at the end.

Most of the images I use come in RGB, sometimes a CMYK scan. Either way, moving to LAB involves no loss of color since LAB encompasses both spaces, with hefty room to spare. Keep in mind, also, that LAB is Photoshop's mother tongue. RGB and CMYK are conveniences arranged for the sake of the user, but even then, Photoshop is working translations on the fly based on the profile setup for each space. Internally, the colors are always in LAB.

There can often be a loss of color after a series of moves in LAB, owing, again, to the possibility of creating colors in in that space that far exceed the CMYK and RGB gamuts. It's crucial that you do not convert directly from LAB to CMYK. There are numerous situations where you suffer a serious hit to shadow detail, as well as unexpected and unwanted color shifts. But remember that whatever you see on your monitor, even in LAB, is already within the RGB gamut, since that's all the monitor has available. A quick stop into RGB on your way to CMYK will eliminate the possiblity of truly bizarre shifts.

Even so, I have managed miraculous color transformations in LAB when starting with a dull, drab CMYK image, and while I cannot always move the full scope of the effect back to CMYK, I can produce a result that would never have been possible in CMYK alone.

Essentially, the techniques here are rewriting the conventional wisdom of what is involved in normal image enhancement. If you're not familiar with the inner workings of LAB, be prepared to treat curves, at least as they relate to color, like something from another planet. But the ideas are really quite simple, just different. Check out the various threads linked from here: http://www.dgrin.com/showthread.php?t=18203 and enjoy!

Thusie

Jan-10-2006, 02:31 AM

Thank you Rutt for your time and the wonderful tutorial, it makes some sense now.

rutt

Jan-10-2006, 03:36 AM

Thank you Rutt for your time and the wonderful tutorial, it makes some sense now.

It will make lots more sense if you read on a bit starting here: http://dgrin.com/showthread.php?t=18308

Thusie

Jan-10-2006, 05:08 AM

I do have a dumb question. Is there anyway to have the a & b boxes up on your monitor at the same time? Or is there some reason they shouldn't be:dunno

OsirisPhoto

Jan-10-2006, 05:47 AM

I do have a dumb question. Is there anyway to have the a & b boxes up on your monitor at the same time? Or is there some reason they shouldn't be:dunno

Don't think it is possible, unfortunately. :scratch

Would be nice, though. Just like being able to adjust the a & b curves symmetrically in one go - I'd like to see that possible.

DavidTO

Jan-10-2006, 05:50 AM

Don't think it is possible, unfortunately. :scratch

Would be nice, though. Just like being able to adjust the a & b curves symmetrically in one go - I'd like to see that possible.

That is, in a way. Set your curves and then save them. Then all you have to do is load them. I"ve saved a bunch of options:

LAB05
LAB10
LA-5B10

etc.

OsirisPhoto

Jan-10-2006, 06:28 AM

Well, maybe not perfect.. but much better than the original posted on the whipping post here (http://www.dgrin.com/showthread.php?t=25537), and before & after here (http://www.dgrin.com/showthread.php?t=20074&page=2).)

http://www.hyperchamber.com/images/pull.jpg

'After' version on before & after used basic symmetrical steepening and USM. After applying these LAB curves (http://www.hyperchamber.com/images/stonehaven2623.acv)+ some USM, this is what I have.

http://www.hyperchamber.com/images/pull4.jpg

If you look at the a & b curves, they're more than just symmetrically steepend. Don't know if this is the correct way to look at things, but while roaming around the image with the colour sampling eyedropper looking for areas on the curves to adjust, I noticed that much of curve seemed 'unused', hence the freedom to allow the shape of the curves. :dunno

I think it works. From what I have read, the whole point of adding POP and using LAB curve adjustment to do it, is to have a resulting image that takes the photographer right back there.

rutt

Jan-10-2006, 07:56 AM

Well, maybe not perfect.. but much better than the original posted on the whipping post.

Give yourself a pat on the back for being adventurous! But, please, can you explain what your were trying to do when you varied the basic recipe and how the curves you wrote achieved it?

I just applied the basic recipe to your original, with 10% steepening of the A and B curves. My L curve looks a lot different than yours. I found a good shadow near the crotch of the guy on the viewer's right. What I didn't find was a good highlight. The lightest thing I found was the brightest patch of the arm of the second guy from the viewer's right. So I moved the light endpoint of the L curve toward the center so it was just to the left of this point. I added a bit of steepness through the mid tones with those two interior points. I didn't sharpen, as that's not really the issue here.

Here is what I got:

http://rutt.smugmug.com/photos/51848853-O.jpg

Here are the curves I used:

http://rutt.smugmug.com/photos/51848832-S.jpghttp://rutt.smugmug.com/photos/51848841-S.jpghttp://rutt.smugmug.com/photos/51848844-S.jpg

Compare my result side by side with yours. Here are the things I think are better about mine.
The shadows didn't plug (lose detail), especially in crotch area.
The grass is a greener than in the original, but more believably so.
Generally there is better detail throughout the men and their clothes.
That tag hanging from the rope is bright, but a believable color.

Note: My favorite way to compare before/afters like this is to load them both into photohop and copy one into a layer of the other. Then you just click on the layer eyeball to toggle back and forth and you'll see them in place. If this tip doesn't make sense, speak up and I'll go into detail.

Anyway, I don't mean to discourage experimentation. In fact you have skipped ahead a couple or 10 chapters here. But you should learn to crawl before you run with this one. Get the basic recipe just as I described it in the tutorial to work for you first and then start working through the more subsequent chapters and more advanced techniques here: http://www.dgrin.com/showthread.php?t=18203

jfriend

Jan-10-2006, 01:25 PM

Give yourself a pat on the back for being adventurous!

I just applied the basic recipe to your original, with 10% steepening of the A and B curves. My L curve looks a lot different than yours. I found a good shadow near the crotch of the guy on the viewer's right. What I didn't find was a good highlight. The lightest thing I found was the brightest patch of the arm of the second guy from the viewer's right. So I moved the light endpoint of the L curve toward the center so it was just to the left of this point. I added a bit of steepness through the mid tones with those two interior points. I didn't sharpen, as that's not really the issue here.
First, I'd like to point out that the original posting for this image is in aRGB so that's one reason why it looks so washed out in all non-color-aware browsers. Just moving it to sRGB helps quite a bit.

Second, I agree with the Rutt's basic philosphy of sticking with the recipe unless you are sure you are making it better.

I did basically the same thing as Rutt (followed the basic chapter 1 LAB color enhancement recipe of pusing in the A and B curves 10% symmetrically on each end), but I added the following things afterwards (just because this is a fun image to play with and learn from):
A steeper L-curve, but used blend-if and curve shape to keep from making the blown highlight areas worse.
A small amount of shadow correction in shadow highlights to bring a little detail back in the darker parts of the blue jeans.
A separate A-curve for just the faces (using a mask) to pull some magenta out of the faces while not affecting the arms.
Slightly brightened the faces with an L-curve on the same face mask.
Fixed blown highlights on the arms with both the impossible color fix and a little cloning
Tried to fix some of the blown highlights on the shoulders (though I was less successful here)
No sharpening added.Here's what I ended up with:
http://jfriend.smugmug.com/photos/51879309-L.jpg

mercphoto

Jan-10-2006, 06:26 PM

Another thread in the whipping post that I started led me down the path of trying to add color punch through the LAB color space for the first time. I'm looking for advice on whether it is an improvement to what I am doing now so I'm posting three photos here: the original, what I normally do, and what I am thinking of doing.

Specifically, the color punch phase is done in LAB with curves. Lightness channel sets the black point to 9 (probably too high), and then the A and B channels brings the white point down to 104 and brings the black point up high enough to get it to pass through a middle neutral gray once again. Then back to RGB for the sharpening pass.

Thanks in advance for any thumbs up or down.

http://mercphoto.smugmug.com/photos/51911704-M.jpg

What I usually do for color and sharpness:
http://mercphoto.smugmug.com/photos/51911853-M.jpg

What I am thinking doing from now on:
http://mercphoto.smugmug.com/photos/51912040-M.jpg

DavidTO

Jan-10-2006, 06:54 PM

Can you post your curves? That would help. I need to see them.

Yes, your black point on the L looks to be too far, as you've plugged up the shadows.

The colors look better, IMO, in LAB, but it's hard to be sure when you've plugged up the shadows....

Also, I'm moving this to the Chapter 1 discussion....

rutt

Jan-10-2006, 07:32 PM

I didn't follow your description of your LAB curves. A picture is worth a thousand words; perhaps you could post a screen shot of the curves you are talking about.

And once you have the image in LAB, why not sharpen the L channel there? Sharpening in RGB will often introduce unwanted color artifacts which can't happen in LAB. My sharpening tutorial shows this. See: http://dgrin.smugmug.com/gallery/1104535/1/51379922/Large

Anyway, this image has a complexity which makes the simplest LAB recipe not work quite right. Those reds are very magenta, almost out of gamut as it is. So just blindly steepening the A curve won't work here. And there is a blue cast in the 3/4 tones and shadows. You can either deal with that first with RGB curves or at the very end in CMYK.

So I'm very curious to see what you actually did.

mercphoto

Jan-10-2006, 07:45 PM

I didn't follow your description of your LAB curves. A picture is worth a thousand words; perhaps you could post a screen shot of the curves you are talking about.

http://mercphoto.smugmug.com/photos/51922694-S.jpg
http://mercphoto.smugmug.com/photos/51922699-S.jpg

And once you have the image in LAB, why not sharpen the L channel there?
Because I'm sharpening with a high pass filter, not with USM.

Anyway, this image has a complexity which makes the simplest LAB recipe not work quite right. Those reds are very magenta, almost out of gamut as it is. So just blindly steepening the A curve won't work here.
Dang. :( Looking for a standard recipe for this type of shot. Does not need to be ideal, just needs to be good, and also have little risk of doing anything strange to a photo. Its what got me to my current way of doing things.

rutt

Jan-10-2006, 08:27 PM

Well, for starters, you could try an L curve more like this:

http://rutt.smugmug.com/photos/51919745-S.jpg

(Note the reversed axis, with lightness to the left.)

Next I'd steepen just the B curve and not the A curve for this one because of that extereme magenta. That's going to make it a little orange, but better than not doing it. There are many better things, but this one is simple. I steepened the B curve by 10% on each end, used the L curve above, and USM'ed in L and got this:

http://rutt.smugmug.com/photos/51919789-M.jpg

mercphoto

Jan-10-2006, 08:32 PM

Next I'd steepen just the B curve and not the A curve for this one because of that extereme magenta.
Thanks for the help, but that particular statement makes me want to abandon this for my racing photos. Reason is simple: I need a standard recipe. Can't hand-edit 1,000+ photos in that way. But its a good exercise to learn from nonetheless.

I think what I will do is simply add a shadows adjustment to what I normally do. That does seem to make some difference, to adjust the shadows to about 5, as you did as well.

OsirisPhoto

Jan-11-2006, 04:31 AM

Give yourself a pat on the back for being adventurous! But, please, can you explain what your were trying to do when you varied the basic recipe and how the curves you wrote achieved it?

The more I fettled with the curves, the more I came to like the 'punch' of the highly saturated colours. It was a very bright day, and the high contrast image really reminds me of the day. Although TBH, your version does have more detail and I appreciate that it is a better image and more natural. :thumb

It was fun to manipulate points inside the A & B curves, just as with the L channel, by using the colour sampler eyedropper to pick areas where I wanted to exagerate the colour.

Note: My favorite way to compare before/afters like this is to load them both into photohop and copy one into a layer of the other. Then you just click on the layer eyeball to toggle back and forth and you'll see them in place. If this tip doesn't make sense, speak up and I'll go into detail.

Got that, cheers :thumb

Anyway, I don't mean to discourage experimentation. In fact you have skipped ahead a couple or 10 chapters here. But you should learn to crawl before you run with this one. Get the basic recipe just as I described it in the tutorial to work for you first and then start working through the more subsequent chapters and more advanced techniques here: http://www.dgrin.com/showthread.php?t=18203

Fair point about starting with the basics. I'm going to have a final go at this image from the start, using the basic method and your curves as a starting point, with maybe just a little tweaking... just a little :wink

OsirisPhoto

Jan-11-2006, 04:51 AM

First, I'd like to point out that the original posting for this image is in aRGB so that's one reason why it looks so washed out in all non-color-aware browsers. Just moving it to sRGB helps quite a bit.

Thanks for the tip :thumb I think I need some color-aware eyes :rofl

I did basically the same thing as Rutt (followed the basic chapter 1 LAB color enhancement recipe of pusing in the A and B curves 10% symmetrically on each end), but I added the following things afterwards (just because this is a fun image to play with and learn from):

A steeper L-curve, but used blend-if and curve shape to keep from making the blown highlight areas worse.
A small amount of shadow correction in shadow highlights to bring a little detail back in the darker parts of the blue jeans.
A separate A-curve for just the faces (using a mask) to pull some magenta out of the faces while not affecting the arms.
Slightly brightened the faces with an L-curve on the same face mask.
Fixed blown highlights on the arms with both the impossible color fix and a little cloning
Tried to fix some of the blown highlights on the shoulders (though I was less successful here)
No sharpening added.Your version certainly has some significant improvements over the original, particularly the fine details (blades of grass, boots, jeans, etc).

I think your right that the blown highlights on the arm / shoulder need fixing - I would normally blend a couple of different exposure compensated versions from RAW (still to try with this image, so may well try your method instead).

I must admit though, I prefer the faces a little magenta... this day was a (very rare) hot, hot, day, and these guys were really going for it. Like a row of grunting beetroot! :D

rutt

Jan-11-2006, 04:59 AM

I must admit though, I prefer the faces a little magenta... this day was a (very rare) hot, hot, day, and these guys were really going for it. Like a row of grunting beetroot! :D

Yeah that rule about yellow and magenta balance in faces, I've come around to the conclusion that you have to be very careful about it in unusual cercumstances, such as this one. When it's cold, when people exert themselves (as here), you need a very light hand here. If the color balance is correct otherwise, maybe it's time to resist the temptation to make a mask to turn up the yellow in faces when there really is a reason for the faces to be hot colored.

In this particular case, there's additional evidence that leaving the magenta in the faces is the right thing. The flesh of the people in the background does seem OK.

DavidTO

Jan-11-2006, 05:53 AM

Yeah that rule about yellow and magenta balance in faces,

What I remember (that backs up what rutt's saying here) is that you are supposed to measure a representative area. An area that represents what you think would be normal skin tones. That means that people can have patches of red (like a birthmark) and you can ignore it. It also means that people exerting themselves are not representative of normal skin tone at all...but the people in the background are.

Thusie

Jan-11-2006, 08:42 AM

I hope for some help here. I have been workin off and on with this shot all morning using the basic LAB and USM in Lightness. Yes it's a squirrel:D but they seem to be what I do best and my minds eye knows what they should look like, most of the time. This shot drove me nuts! I appreciate your time.

http://Thusie.smugmug.com/photos/51972728-L.jpg

http://Thusie.smugmug.com/photos/51972327-S.jpg

http://Thusie.smugmug.com/photos/51972343-S.jpg

http://Thusie.smugmug.com/photos/51972352-S.jpg

http://Thusie.smugmug.com/photos/51972312-L.jpg

rutt

Jan-11-2006, 09:12 AM

I hope for some help here.

Looks a lot better to me, but maybe you don't like the orange that has crept into the background and coat? Try more conservative A+B steepening. Try 5%, 10%.

rutt

Jan-11-2006, 09:13 AM

Looks a lot better to me, but maybe you don't like the orange that has crept into the background and coat? Try more conservative A+B steepening. Try 5%, 10%.

Oh, and I just noticed that your A and B curves aren't steepened by the same amount. Is there a reason for that, or just oversight.

jfriend

Jan-11-2006, 09:18 AM

Looks a lot better to me, but maybe you don't like the orange that has crept into the background and coat? Try more conservative A+B steepening. Try 5%, 10%.

When I played with this one, I liked that with some steeper A and B curves, you could start to see some of the color in the squirrel, but clearly the background went nuts and really detracted. At first I started thinking about a mask for the background (you can find a channel that, when steepened makes an OK mask), but then I realized that you can just use the BlendIf sliders on the L channel on the steepened A and B curves because there's clear tonal separation between the squirrel and the background (the background is pretty much completely brighter than the squirrel). So, you can draw color out of the squirrel without overdoing the background.

I've got to learn that lesson that before I go to a mask, try blendif.

Sorry, I can't post a result right now, but thought I'd share this idea.

Thusie

Jan-11-2006, 10:09 AM

jfriend

Jan-11-2006, 12:02 PM

John,

I'm real basic or real beginner with all this, blending will be for another time me thinks:): Yes I also like the squirrel with the steeper curves but the orange, yeow!

Rutt what I was trying to do was keep keep the brown in the coat w/o the popping orange so tried setting the a&b a bit different. Didn't work. Plus I think I looked at it way too long
So here tis with a&b both set to 5%, I like this much better. Used the (lower) eye for a black point and the lightest spot I could find for the white, that made working with the lightness curve much easier.

I think this is better?

How does this one work for you? This is a drastically steeper A and B (30% on each end) where the effect has been mostly blocked from the background so it's just the squirrel that gets the benefit. It's also got an L curve that tries to bring out more detail in the tones of the squirrel. If you're interested, I can explain how I did this, but probably can't do it until tonight.

http://jfriend.smugmug.com/photos/51988603-L.jpg

Thusie

Jan-11-2006, 01:50 PM

John,

If you have the time I would certainly appreciate the 'how to', I think it would come in quite handy.

Thanks!

rutt

Jan-11-2006, 02:37 PM

I think this is better?

http://Thusie.smugmug.com/photos/51979611-S.jpg

I love him now. You might be just a little braver with that L curve to try to lighten up a bit and get more detail in the fur. Try pulling down the quatertones a touch (that means making a point on the curve about 1/4 of the way over and pulling it downward just a little. You want the curve a bit steepper in the area of the fur. You might have to make another point near the right hand side and pull it upward.

There are all kinds of ways to limit the color the A and B curves introduce when they get too steep. John really really wants to show you how to do that, but the best way to learn is either to plow forward through the reading group posts (http://www.dgrin.com/showthread.php?t=18203) and/or (preferably) buy Dan's LAB book. (http://dgrin.smugmug.com/gallery/1085484)

ginger_55

Jan-11-2006, 03:33 PM

"Blend if" sliders on the L curve??? To keep the background out of the mix????

g (It looks very good, Blend if (?) )

Thusie

Jan-11-2006, 03:39 PM

I love him now. You might be just a little braver with that L curve to try to lighten up a bit and get more detail in the fur.

There are all kinds of ways to limit the color the A and B curves introduce when they get too steep. John really really wants to show you how to do that, but the best way to learn is either to plow forward through the reading group posts (http://www.dgrin.com/showthread.php?t=18203) and/or (preferably) buy Dan's LAB book. (http://dgrin.smugmug.com/gallery/1085484)

Thank you Rutt and I will work on the L curve. I have been reading the group posts and printing some things. For some reason I have always done better with something in hand than reading on the net:dunno We have a start, small one but a start.:):

rutt

Jan-11-2006, 04:30 PM

"Blend if" sliders on the L curve??? To keep the background out of the mix????

g (It looks very good, Blend if (?) )

Ginger, it's a more advanceced technique. Pathfinder describes it in his Chapter 7 summary (http://www.dgrin.com/showthread.php?t=20433). I covered it right at the top of my Chapter 8 summary. And since I happen to know for absolute sure that you have this book and haven't read it, I'll give you a page reference: 157-159 and, better, 174-177.

Next time I want to hide money from you, I'll put it in a book.

Anyway, you don't need this fancy stuff in order to use this simple recipe and get great results. Learn to simple A+B steepening and learn how to write good L curves and you'll be able to improve lots and lots of your shots and you'll be ready to go on to more advanced stuff. Only Ginger has no excuse.

Phil_L

Mar-22-2007, 03:11 AM

I tried the tutorial as an experiment, since I am planning to buy the book.

The results were totally amazing in relation to the time spent tweaking the pics.

:thumb :lust

It is probably worth mentioning in the tut that you can do all the tweaks on adjustment layers, if you want to be able to go back for adjustments!??

I am still plowing through Barry Hanes book PS Artistry.

Will these threads remain open so I can come and play at a later Date?

rutt

Mar-22-2007, 04:19 AM

This thread will remain open as long as dgrin does.

I tried the tutorial as an experiment, since I am planning to buy the book.

The results were totally amazing in relation to the time spent tweaking the pics.

:thumb :lust

It is probably worth mentioning in the tut that you can do all the tweaks on adjustment layers, if you want to be able to go back for adjustments!??

I am still plowing through Barry Hanes book PS Artistry.

Will these threads remain open so I can come and play at a later Date?

Phil_L

Apr-28-2007, 02:43 AM

I have taken the quick and dirty tour of the first six chapters of the book and I am completely blown away by the raw (:D ) power availible in L*A*B*.

Here are two examples of reworking using curves steepening:

1) Snap shot of a house in the Swiss ski resort Andermatt.
Dates back to before the Napoleonic wars and today maintained in original shape by the town council!
Camera 3.0 MP Oly P&S.
Chickened out of taking my DSLR skiing.:dunno

Before

http://phil.smugmug.com/photos/147519786-L.jpg

After steepening curves. Fairly agressive with a and b.

http://phil.smugmug.com/photos/147519715-L.jpg

Tell me if you think it is overcooked.

Yes I know there are blown highlights and blocked shadows.:D

2) From the summer palace of August the Strong of Saxony (also King elect of Poland) on the outskirts of Dresden in Germany.
Camera D70 with AF-S 12-24 f4.

Before

http://phil.smugmug.com/photos/147520214-L.jpg

After the steepening of curves

http://phil.smugmug.com/photos/147520543-L.jpg

Actually tweaked the sky on a seperate layer but still.
Could probably have been more agressive with the steepening?

Sorry dont know how to do screen shots! :dunno

Going to review what I have covered before attacking the rest of the book.

For those who havent tried it yet:
If you have a reasonable understanding of curves and PS you can get a lot out of this book and be up and running, without having to digest the whole thing .
Dont be put off by the for geeks only rep of the book!

fwiw

JimW

Jun-10-2007, 07:25 AM

I just got the LAB book (Canyon Conundrum), and so far have read chapter one, as well as Rutt’s excellent synopsis. I know I’m about a year and a half late, but better late than never. After years on the road doing color correction for the printing industry, I’m retiring from press okays (mostly – still have one client), and concentrating on photography. So I’m familiar with cmyk, and starting to learn rgb color correction, when I happened to notice this section of the forum. I’m used to fashion clients going berserk if the seafoam green shirt isn’t perfect, so the lunacy of spending ridiculous amounts of time & effort getting an image just right seems completely normal to me, and I love it.

I started practicing on a recent wedding I shot 2 weeks ago in St Augustine, Florida. This wedding was as close to shooting blind as I’ve ever come. It was in the afternoon, on the beach, and there wasn’t a cloud in the sky until the very end. I’ve spent much of my life in dark pressrooms, and live in a small dark NYC apartment, so for me this wedding was like shooting on the surface of the sun. I shot with my dark shades on, and still couldn’t see a thing. I kept saying to myself, “just don’t overexpose”. So I overcompensated a tad. :D I think the ambient exposures are okay, but I failed to get enough flash into the subjects on all these “formals” in the sand dunes. I think I’ll have to retire my Lumiquest bounce card thing, as most of the flash never reached the subjects. Oops.

This is the final image of the day, and the sun was behind some thin clouds. First one shows the raw settings with no adjustment.

http://http://jimwhitaker.smugmug.com/photos/161276193-O.jpghttp://jimwhitaker.smugmug.com/photos/161276193-O.jpg
http://jimwhitaker.smugmug.com/photos/161276201-L.jpg

Then I processed the image through RAW twice, once for the sky (dark) and a second time for everything else. Here are the RAW settings for the “light” version (everything except sky).
http://jimwhitaker.smugmug.com/photos/161282855-L.jpg
I’ve never had to add so much exposure in RAW before, but it appears to have worked.

Then I merged them in PS and masked around the grass and dunes to let the darker sky show through. Next I brightened the sand & people slightly with an rgb curve. Then I used jfriend’s method of lightening on an overlay layer for the flesh only (which I just found in the “dark eyes” thread). Then I curved the sky layer only to add contrast to the clouds & sky, and (still in rgb) got this:

http://jimwhitaker.smugmug.com/photos/161276176-O.jpg
http://jimwhitaker.smugmug.com/photos/161276203-L.jpg

Then I flattened, took it into LAB, and applied the following curves:
http://jimwhitaker.smugmug.com/photos/161299921-O.jpg

and got this.

http://jimwhitaker.smugmug.com/photos/161276169-O.jpg

The clouds are neutral grey, although blue is a little high compared to red & green, but, coming from the printing world I see that as a good thing (insurance for on-press issues). My original goals were to brighten the sand & people a lot, while showing good value in the sky. I paid particular attention to the flesh tones. In this case, I don’t care about being accurate to the scene. The scene was brutally bright and difficult. I just want a nice photo for them. I’m very pleasantly surprised with the amount of improvement I was able to get, and so far have been very impressed with LAB results. In most cases, the LAB versions look almost as if they were (dare I say?) ... backlit (at least as compared to the rgb version). I capture in Adobe RGB, and should note that when I converted these to srgb for uploading, some of the warmth disappeared.

All comments and critiques are most welcome. I want to learn, and hope to keep reading and following Rutt’s synopses of the chapters. I really want to add LAB processing to my toolbox of processing tricks, and am enthused by the support here. Thanks very much for this opportunity. (Have I made this post too long?)

Jim

DavidTO

Jun-10-2007, 09:58 AM

Hey, Jim, nice work! Glad you found the LAB stuff here, it's really powerful, and a hoot, to boot!

I merged your thread with this one, so that everyone can stay on track with Chapter 1 and follow along with your work, as well!

JimW

Jun-19-2007, 11:04 AM

Thanks David. It helps me to write it down and post work on each chapter, and maybe it’ll help someone else too. The results I'm getting so far are great, but it leaves me wanting a better understanding, so I'm determined to keep going.

I’ve studied chapter two and understand that:

Positive numbers indicate warm colors, negative numbers indicate cool colors.
Zero is no color, a neutral.
The L is slightly lighter, and higher in midtone contrast, than if we converted to grayscale.
On the L scale, 100 is light and 0 is dark.
(It helped me to know that the scale of 0-100 for the L channel still represents 256 levels, and that Photoshop just reports it as 0-100 for our ease of use.)

Skies almost always fall to the green, not the magenta, side of blue.
Grass, and stuff that grows, is always to the yellow, never the blue, side of green.

In rgb, to determine if something will be neutral, all three channels must be compared. In LAB, there is a fixed neutrality inherent in each channel. If the object is at 0 on the A channel, we don’t have to compare it to any other channels, it will be neutral in that channel.

For practice, I processed this file as rgb with the following curve:

http://jimwhitaker.smugmug.com/photos/164265429-Th.jpg

http://jimwhitaker.smugmug.com/photos/164265494-L.jpg

Then I processed as LAB with the following curves:

http://jimwhitaker.smugmug.com/photos/161299921-S.jpg

http://jimwhitaker.smugmug.com/photos/164265459-L.jpg

Question: For the rgb version, I flattened, reduced size from 300 dpi to 72 dpi, then sharpened using unsharp mask, at (85, 1.0, 4). For the LAB version, I sharpened the Lightness channel in LAB using unsharp mask at (200, 1.0, 8 as the author suggests), then flattened, converted to rgb, reduced size to 72 dpi.
Why does the rgb look sharp but the LAB doesn’t? (Maybe I should wait for the sharpening chapter?)

Comments and/or help is welcome. Thanks, Jim.

canoflan

Jun-19-2007, 11:41 AM

JimW

Jun-19-2007, 12:58 PM

I'm not going from 300 to 72 and back, just from 300 to 72, in order to post online here with a smaller file. I don't think I really completely understand the dpi issue yet.

It's not from the book. I just thought I should reduce the size because monitors only show 72 anyways.

canoflan

Jun-19-2007, 01:19 PM

I understand. Just another case of me overthinking things.

jfriend

Jun-19-2007, 03:37 PM

Question: For the rgb version, I flattened, reduced size from 300 dpi to 72 dpi, then sharpened using unsharp mask, at (85, 1.0, 4). For the LAB version, I sharpened the Lightness channel in LAB using unsharp mask at (200, 1.0, 8 as the author suggests), then flattened, converted to rgb, reduced size to 72 dpi.
Why does the rgb look sharp but the LAB doesn’t? (Maybe I should wait for the sharpening chapter?)

Comments and/or help is welcome. Thanks, Jim.
If you post a link to the original image (probably in a Smugmug gallery), we can take a look at it for ourselves.

Perceived sharpness comes from a variety of operations. I actually find that the most pronounced contributor to sharpness is contrast. In most cases, higher contrast things will look sharper. Without seeing the original myself, I can't tell for sure, but I think you may be seeing a contrast difference between these two images, particularly in the tones that cover the bricks.

Some things that affect sharpness are:
Contrast (as I said above)
Your sharpening process (I've found I have to sharpen the L-channel much stronger than an RGB sharpening operation)
Your resizing process before viewing on the web
Your sharpening process after resizing for the webNow, on resizing for the web, it is a lot easier to completely forget about dpi. The web works in pixels, not dpi and not in inches. What you want to do is to resize your image for the web to a specific pixel dimension. It doesn't matter at all what the labelled dpi or image dimensions are on the web. Those are just stamped attributes on the image that have nothing to do with the actual pixels in the image and have nothing to do with how a browser displays them.

For a little reading material on resizing for web display, here are couple links:

http://www.cambridgeincolour.com/tutorials/image-resize-for-web.htm
http://www.infinitee-designs.com/Save-for-the-Web.htm

For posting here, I don't resize the images myself. Instead, I upload the original to a non-public Smugmug gallery that I use just for posting images in forums, I let Smugmug generate the various sizes and I then link in either the -M or the -L version here. When Smugmug does the creation of those sizes, it not only resizes, but also adds a bit of sharpening and it saves me lots of time to just let them do the work for me automatically.

JimW

Jun-20-2007, 06:06 AM

Yet another very helpful post John. Thanks for taking the time.

If you post a link to the original image (probably in a Smugmug gallery), we can take a look at it for ourselves Since I have been posting 72 dpi only, this wouldn’t help, right? By “original image”, I think you mean hi res (300 dpi)?

I actually find that the most pronounced contributor to sharpness is contrast. In most cases, higher contrast things will look sharper I understand and agree.

I've found I have to sharpen the L-channel much stronger than an RGB sharpening operation I’m just beginning to notice this, too.

Your sharpening process after resizing for the web Yes, I think that’s the key, especially after reading the first of your links. On the LAB version, I sharpened in LAB, then converted back to rgb, then downsized BUT did not resharpen after downsizing. I think I should have. But, at least in the first two chapters, he didn’t mention this.

It doesn't matter at all what the labelled dpi or image dimensions are on the web. Those are just stamped attributes on the image that have nothing to do with the actual pixels in the image and have nothing to do with how a browser displays them.
Hmm, it’s taking a while for that to sink in. So, if I upload an image to my galleries (non-public, as you said), for posting here, it makes NO difference whether I upload 300 dpi versus 72 dpi? However, it must make a difference on sharpening, since a 72 dpi image seems to show sharpening much quicker than a 300 dpi image?

let Smugmug generate the various sizes and I then link in either the -M or the -L version here. When Smugmug does the creation of those sizes, it not only resizes, but also adds a bit of sharpening So you must get used to knowing how much SM will sharpen, and you can adjust your sharpening accordingly?

Really appreciate the help. Jim

jfriend

Jun-20-2007, 07:25 AM

Yet another very helpful post John. Thanks for taking the time.

Since I have been posting 72 dpi only, this wouldn’t help, right? By “original image”, I think you mean hi res (300 dpi)?

Hmm, it’s taking a while for that to sink in. So, if I upload an image to my galleries (non-public, as you said), for posting here, it makes NO difference whether I upload 300 dpi versus 72 dpi? However, it must make a difference on sharpening, since a 72 dpi image seems to show sharpening much quicker than a 300 dpi image?

So you must get used to knowing how much SM will sharpen, and you can adjust your sharpening accordingly?

By "original image", I meant a full resolution image (the original pixels with no resizing) that you haven't yet done post processing on. Then, we could see for ourselves what influences sharpness the most in that image.

For the web, ALL that matters in this regard is how many pixels the image has. An image with 12 megapixels will be a lot different than an image with 2 megapixels, particularly in relation to sharpening as you have noticed.

The dpi rating is actually just a metadata tag on the image (or derived from some other metadata tags). It can be changed from 72 dpi to 300 dpi without changing the pixels of the image one bit and thus without changing the display on the web one bit. For the web, you should learn to ignore the dpi value and start to focus only on how many pixels you have. As it turns out, the workflow you were using to change dpi was also changing the number of pixels so it "seemed" like the right thing to do, but it is not the best thing to look at.

That's kind of like watching the tachometer in your car to see how fast you're going. It's sometimes an indication of how fast you're going (and sometimes not because of gear changes), but watching the speedometer is a much more direct way to keep track of your speed that is never misleading. For the web, keep your eye on the number of pixels, not the dpi rating.

On Smugmug, I sharpen my high-res versions the way I like them and Smugmug does a good job of preserving that "look" when they create smaller, downrezzed versions of the images. There are many different methods and opinions on sharpening so this is just how I do it - there are lots of other viewpoints too.

JimW

Jun-20-2007, 09:19 AM

It (dpi) can be changed from 72 dpi to 300 dpi without changing the pixels of the image one bit Ok, I think I get that. If you change dpi from 72 to 300, you still have the same number of pixels, and the same file size, but now they will make a smaller image size. Yes, Ok.

For the web, you should learn to ignore the dpi value and start to focus only on how many pixels you haveI'm getting it now.

As it turns out, the workflow you were using to change dpi was also changing the number of pixels so it "seemed" like the right thing to do, but it is not the best thing to look at.I'm afraid I lost you there.

Here is the way I was taught to change a 300 dpi image for uploading to the web:

Open an “original” image, full res with all original pixels
Open Image/Image size
Uncheck resample
Change resolution to 72
Re-check resample
Change pixel width (or height-whichever is bigger) to 800 px.
Click OK.

I have always done this when preparing images for uploading or for emailing. Now I’m beginning to suspect that I’m not doing it right, or that it’s unnecessary.

On Smugmug, I sharpen my high-res versions the way I like them Wait, please. The way you like them for what, print OR web? In my experience, they are totally different. Are they the same for you?

I jumped into Photoshop learning curves and masks to get work done, mostly inkjet prints, and now I realize I should go back and learn the basics. I’ve got big gaps in my postprocessing knowledge, which are now driving me crazy.

:bash

jfriend

Jun-20-2007, 03:31 PM

Ok, I think I get that. If you change dpi from 72 to 300, you still have the same number of pixels, and the same file size, but now they will make a smaller image size. Yes, Ok.
Displaying an image in Smugmug or elsewhere on the web has absolutely nothing to do with the dpi. You are right that if you keep the pixels the same and change the dpi from 72 to 300, the labeled image size will go down, but that's just a label on the image and hardly anything these days uses that value so I just ignore it.

What that labeled size was made for was printing, but ALL software that I use for printing lets me tell it how big I want the output print to be which is a specific instruction to ignore the labeled size on the image. That's why it's largely irrelevant.

I'm afraid I lost you there.

Here is the way I was taught to change a 300 dpi image for uploading to the web:

Open an “original” image, full res with all original pixels
Open Image/Image size
Uncheck resample
Change resolution to 72
Re-check resample
Change pixel width (or height-whichever is bigger) to 800 px.
Click OK.

I have always done this when preparing images for uploading or for emailing. Now I’m beginning to suspect that I’m not doing it right, or that it’s unnecessary.
OK, I thought you were changing the dpi with resampling ON. Since you are separately changing the pixel width, you can just drop the steps that change the resolution. They are not needed for any web uses or printing uses where the printer is told what size print to make.

Wait, please. The way you like them for what, print OR web? In my experience, they are totally different. Are they the same for you?

If you use Smugmug for both web viewing and printing, then you are, by definition, using the same high res image as the source for both. As such, you have to strike a balance in your sharpening so it can meet both needs. I am not one of those people who spends a lot of time optimizing the sharpening of an image for a particular purpose. I tend to do general purpose sharpening that seems to work fine for both prints and web. I realize there are other opinions on this topic, but I don't try to get that complicated or that optimized.

JimW

Jun-20-2007, 06:20 PM

John, I understand and sincerely appreciate your valuable help. I have learned.

Jim

Myer

Jul-05-2007, 05:53 AM

After reading the book, LAB has become my colorspace of choice for my enhancements.

My standard process has become:

"S" curve area of significance in L
Slope A and B between 10-15%. I find more often looks artificial.
Highlight/Shadow to bring back dark areas (Shadow) and remove haze and glare (Highlight).
Sharpen - USM
= = = = = =
I have found that the most bang appears to be in the Highlight area. Removing the glare seems to be the part that adds a lot of crispness to the end result.
= = = = = =
I'm considering moving this step to the start and then continue as required once the glare is gone.

Comments and experiences would be most welcome.

StevenL

Mar-07-2008, 12:44 PM

I'm new here, so please forgive me if I screw up my reply... an old thread anyway, so may go unnoticed

Perceived sharpness comes from a variety of operations. I actually find that the most pronounced contributor to sharpness is contrast. In most cases, higher contrast things will look sharper.

My understanding of some of the things John brought up here are a little different, so I thought I might post them to try to stimulate further discussion, and help better my understanding.

Concerning the quote above, the way I learned it, contrast is not just the most pronounced contributor to perceived sharpness, it's the ONLY contributor. The only things that change from one sharpening technique to the other are the way the edges of contrast are selected, and how the contrast is applied.

Some things that affect sharpness are:
Your sharpening process (I've found I have to sharpen the L-channel much stronger than an RGB sharpening operation)On this one (of four) points John listed, I wonder if the need for more sharpening on the L channel in LAB mode, compared to sharpening in the RGB mode, is related to the fact that you're not sharpening (increasing the contrast of) the colors, but only the luminosity. When I do high pass sharpening in RGB mode I always use two clipping layers, one a levels adjustment layer to vary the amount of sharpening (contrast) of the high pass layer, and sandwiched in between those two a hue/sat layer set to desaturate 100%, to avoid any color fringing or halos from the sharpening.

Now, on resizing for the web, it is a lot easier to completely forget about dpi. The web works in pixels, not dpi and not in inches. What you want to do is to resize your image for the web to a specific pixel dimension. It doesn't matter at all what the labelled dpi or image dimensions are on the web. Those are just stamped attributes on the image that have nothing to do with the actual pixels in the image and have nothing to do with how a browser displays them.

Here I think there was confusion about dpi, ppi, and resizing of an image. I think John sorted it out nicely in subsequent posts, but for my own clarification I'll relate my process for posting to the web. What this is about isn't really resizing the viewing size, but actually resampling the image. For print, the aforementioned 300 ppi, or so, is preferred, but for monitor/web viewing I always resample (downsample, as in throw out all the extra data the monitor can't use) the image to 72 ppi to keep the file size as small as possible while retaining as much of the image quality as I can with the conversion to the sRBG color space, and with the losses inherent in jpeg compression. Unlike John, I find significant differences are required in how I sharpen for print output as compared to web viewing output, and save different version of each file destined for both kinds of output.

For posting here, I don't resize the images myself. Instead, I upload the original to a non-public Smugmug gallery that I use just for posting images in forums, I let Smugmug generate the various sizes and I then link in either the -M or the -L version here. When Smugmug does the creation of those sizes, it not only resizes, but also adds a bit of sharpening and it saves me lots of time to just let them do the work for me automatically.
I suspect what this means is Smugmug is doing the downsampling and resizing for you. Since I do my own web output sharpening, I'm not sure I'd like any further sharpening done to my posted images... is there any way around that?