|
Work-a-Long Image:
A work-a-long Example1 image, IMG_7796_1200.jpg, is available. It is an 800x1200 pixel, 351 KB image file down-sized from a largest/highest quality Canon Rebel XT [EOS-350D] in-camera RAW to JPEG conversion file. The JPEG download file was re-sized using the default bi-cubic method settings, slightly sharpened, and then saved as a *.jpg file at 98% quality. The in-camera image parameters [contrast, sharpness, saturation, and tone] were set to 'normal'. Other camera settinges were: AWB; Center-weighted Averaging metering; and Aperature priority.
Methodology:
Starting with a near 'full range' and near preference midtone brightness/colored image, select a histogram 'dip tone' between peaks and create a 'contrast mask' and a mask of its 'transition zone'. Using the 'starting image' and the 'contrast mask', create a preference highlight tone range ['dip tone' to white (100%)] image and a preference shadow tone range [black (0%) to 'dip tone'] image. Blend the 'starting image', 'highlights preference image' and 'shadows preference image' using the StackImages transform with the Density masks 'in-active' and the created 'transition and contrast' masks as the transform's Amount masks. Adjust the Amount masks' white and black sliders to obtain preference tone/color and to minimize halo while monitoring Preview.
The two masks utilized are simplier to construct then the 'complex tone range masks' described in the 3Tone Range Adjustment Procedures yet when utilized with the 'starting image' and preferentially altered image versions in the StackImages transform will provide virtually seamless [no halos nor tone reversals nor artifacts] blending between two aggressively tone/color altered tone ranges. The end result is a very photo-realistic image.
Step1a: Open or click on the 'starting image' and open the Mask Tool - Brightness Curve. Select 'Stair Step' curve. Move the [100,100] control point left to the 65% tone and 'lower-left' Apply. 'lower-left' Apply a Blur = 52 and click OK, creating Mask1.Comments: (1) To better select the 'dip tone', one could select 'High Histogram Resolution' from the OPT-ions and (2) the Blur radius is selected based upon the largest image pixel dimension of a 2:3 or 3:4 aspect image such that raduis = approx. 0.043 x largest pixel dimension, i.e., radius = .043 x 1200 = 51.6. This is not a hard rule but seems to provide sufficient blur for most images to prevent halo's after preference StackImages blending, yet sufficiently seperates the two tone ranges.
Step1b: Click on Mask1 and open the Mask Tool - Brightness Curve. 'lower-left' Apply a Smooth curve = [0,0], [50,100], [100,0]. Click OK, creating Mask2.
Step2: Click on the 'starting image' and open the ColorCurves transform; select Mask1; white = 100; black = 0; and then form and OK, HSL,L Smooth curve = [0,0], [50,35], [100,100] and HSL,S Smooth curve = [0,0], [10,10], [50,40], [100,100], creating the 'Step2 image' [preference 'dip tone' to 100% tone/colors]. Comment: The HSV color space model could also be used if prefered although the curves may differ.
Step3: Click on the 'starting image' and open the ColorCurves transform; select Mask1; white = 0; black = 100; and then form and OK, HSV,V Smooth curve = [0,0], [30,30], [60,85], [80,100], [100,100] and HSV,S Smooth curve = [0,0], [10,10], [60,75], [100,100], creating the 'Step3 image' [preference 0% to 'dip tone'/colors]. Comment: The RGB BrightnessCurve transform could also be used if prefered although the curve will differ with hue shifts in addition to changes in tone and saturation.
Step4: Click on the 'starting image' and open the StackImages transform. Select images, masks, and sliders as illustrated in 'Step4 screen view A'. Adjust Amount sliders to minimize halos and to obtain preference tone/color; perhaps as illustrated in 'Step4 screen view B'. Click OK, creating the 'Step4 image'.
One could continue to refine the 'Step4 image' using favored Dodge/Burn techniques perhaps those as suggested/linked in the DL-C message board posting here and the subsequent postings in this thread. Or perhaps a 38% amount of SoftLight filtering of the 'Step4 image' with itself.
In place of the suggested Step2 and Step3 images, one could use preferentially adjusted actual camera minus/plus bracketed exposure images or simulated EC adjusted RAW conversion images that have near 'full range'.
For some images resulting from the suggested StackImages-2Tone Range Technique, there is a need for additional refinement in the shadows... so a further 0% to 10 or 20% shadow tone range enhancement could be done... This occurs when the 'preference shadow tone range [black (0%) to 'dip tone'] image' used in the StackImages transform has been enhanced for a much larger tone range and contains significant 'deep' shadow image areas.
The 'work-along image' used for illustration [Example1] was optimized for a 0 to 65% tone range ['dip' tone = 65%] image version [Step3]. As it did not contain significant 'deep' shadow image areas, no additional tone adjustments were made or preferred.
A possible final image version that includes: additional dodge/burn to further emphasize the dappling sun spots; slight noise reduction in the sky/cloud image area; a slight 'warming' midtone color balance adjustment; and AdvancedSharpen-ing:
You are welcome to see other IMAGEs by DEN.
