Thanks Marc.
For #1, #3, and #4 I used the LAB method. For #2 I used the Black & White conversion in CS3.
My LAB method starts by converting the image to LAB color and doing a basic curves adjustment. I get the best possible color image I can.
Then I go to channels and choose Lightness. I then go to Image and choose Grayscale. Once I have the Grayscale, if I didn't start with a copy of the Background I make one now, then go to Blend Modes and apply Multiply as your image will most likely be too light. Set Multiply to about 50% opacity, depending on the image.
You can stop here or do another Curves Adjustment Layer and even a Levels Adjustment Layer. And again apply opacity to those layers as is appropriate.
I will then, again depends on the image, use Draw with Light, as described by John Caponigro. For a Black and White image I make two layers one is for color Black and one is for color White. Each layer is set to Blend Mode Overlay. Then using between 10-20% opacity on a soft brush I paint black and white on the areas that need it. This brings out detail you wouldn't normally have. This is the simple version of how to do it. If you are interested there is a training course at Kelby Training. Draw with Light is a modern version of Dodge and Burn that is easily changed as you decide what's best for your image.
The above method is very personal as no two people will ever produce the same image.
I hope this helps. I would love to hear your method. I have used several and found LAB to be the best for me.
Last edited by Joe_M; 02-09-2008 at 02:05 PM.