Week 1 images

[mikemeister_admin]

I agree with Greg, keep your observations coming as they are making the class interesting.

The below is what I find myself doing too, the netural picker gets me in the ball park, and then I tweak it around until I am happy.

Have a good day, or is it night where you are!

Mandy

In this pass, I used RGB mode and used the same procedure as in the video.  I picked three different sample points on the belly and moved the neutral point around until all three sample points were very close to neutral.

[mikemeister_admin]

Here's my Lab duck.  The bill is yellow in Lab but orange in RGB.

[mikemeister_admin]

You guys are encouraging me to ask questions?  Be careful what you wish for?

OK, here's one: should the neutral point not be placed until after brightness and saturation are altered?  I think I noticed (may have been in RGB) one time that increasing brightness moved some of my sample points away from neutral.

[ggroess]

It depends on the color space, so for RGB yes. Remember, RGB can have multiple neutrals based on brightness.  Neutrals need not always be gray.  Black and White can be neutral.  In RGB any color with all three numbers being equal will be gray.  The higher the values the brighter the tone.  so 250,250,250 is neutral...

The answer to your earlier question about my settings is that setting the highlight and neutral to 254, 254, 254  and the Shadow to 4,4,4 keeps my CM from forcing every highlight to the highest values and every shadow to the lowest. This is so that when I print no matter how hard I push the highlights I should have some tonality in the pixels.

Hmmm the LAB should be a better correction overall...try it again and be picky about the neutral...remember LAB likes only 1 neutral.

Greg

[mikemeister_admin]

This time I picked three sample points and got them all to be a = 0, b = 0.
Also really pushed the saturation.
And fiddled with the L curve but pulling it up in the mid area.

How does it look?  Any more advice?

[ggroess]

Jay,

I think I want you to start again in LAB.  Your ducks world in LAB is very jaundiced...We could fix it, but I think doing so would confuse you more at this point.  I had to use some pretty strong moves to get the image back.

Set your neutral in a reasonable spot, Start by looking for the best neutral without marking it.  If you have not set up your hue clock to include clock and text as shown in shot1 please do so by setting the options as shown in shot2 for hue clock and text.  You can choose to leave the clock on RGB if that makes more sense to you but...In LAB a neutral is L (any value) A(0)B(0) it makes for a fast zero in on the neutral.

Hover the mouse around on the band until you see a decent neutral and then set it.  ten adjust the L curve.  Pay attention to the feathers in the darkest regions and see if you can open up those areas without blowing the rest of the image out of the water...keep the breast feathers in mind as you proceed.  Use "control" points on the L curve to bring the curve back to the middle like you would in a lizard tail correction.

Remember color is separate from Lightness in LAB.  make the neutral solid and you can do some really interesting stuff with LAB.

Hang in there it is confusing before it gets better...
Greg

[mikemeister_admin]

OK, I'll try Mr. Duck again.

Remember I pointed out that neutral points seem to be used in two different ways. It sounds to me that for this image you think I should use the neutral to HOLD a neutral point at neutral, unlike the Honest Abe video, where you are trying to SET a non-neutral point to neutral in order to get some sample points to desired values. Am I right?

"Control points" and "lizard tail correction" - are those things we will cover in upcoming lessons?  I don't remember reading about them in this lesson.

-Jay

[ggroess]

"Lizard Tail" is a technique for saving the shadows or highlights...They are in the special curves documents that I posted links to.  Control points are just that..points you add to the curve to keep parts of it from twisting out of control when you adjust.  If you place a neutral on an image and adjust one of the color curves the curve pivots around the neutral point.  Control points hold the curve line in place so you can adjust a very narrow part of the curve.

I think this might help...with the duck....

In LAB you are setting the neutral you have only 1 neutral and this holds a point as neutral.  In LAB it provides a point to pivot the A and B curves. It sets that point to L(x),0,0 that is the "new" center of the A and B curves. This is really important in the Man From Mars Technique we adapt from Dan Margolis later in the course.  In LAB Color is not connected to the L channel.

In RGB you can have a neutral based on the brightness.  There can be multiple neutrals.  In a By the numbers correction we leverage this to our advantage. 

To summarize Honest Abe...
In that image there were multiple areas where we can find an acceptable neutral.  Using the hue clocks for numeric checking and trying to find a balance that allows the image to look good was the goal. 

In the "Duck" there is an obvious choice for a neutral and you need only choose a point in that area that makes the rest of the image look good.  Using the hue clock can get you close or exactly on a neutral. If you find a neutral point in a area that should be neutral you have hit pay dirt. That point will hold the neutral no matter what you do to the rest of the color.

Many images have a point where the neutral should be obvious; Black Top, Silver, shadows on white surfaces etc. we are left guessing sometimes and our goal is to find a good balance point.  You commit to the neutral until you find that other corrections make a color cast show up.  Then you have to look at it again and decide if you need to move it or not.

Keep the gears turning...If you need more maybe we can arrange a skype call and I can talk to you more about this..

Greg

[mikemeister_admin]

Greg,

I don't think I want to spend any more time on the duck.  I'm no longer sure what I'm trying to accomplish, so I'm just spinning my wheels.  I'll wait until the next lesson.

-Jay

[ggroess]

I understand Jay...

I hope you have not spent too much time on it...
Here is a look at a "reasonable" correction. 

If you want to see the curves, save the ACV file to the desktop.  then in CM click the load button and select the ACV file.  It will load my curves file and you can look at what I did in LAB.

Greg