cgardner
Offline
• • • • •
Upload & Sell: Off
|
Doug — Nice to see you finally get past your intellectual property reservations and actually post a photo, the first of yours I've ever seen. I don't say that in a mean spirited way either.
Like I do with any photo here to better understand what is happening in it technically I opened it in Levels in CS5 and applied the "Alt Slider" test by holding down Alt/Opt key and clicking on the H and S slider triangles to reveal clipping in this final 8-bit JPG step in the workflow and comparing it to the highlights and shadows in the image.
I know your preference for naturally light and you've used it quite skillfully here. You exposed the highlights as I would at capture letting only the brightest clip. They don't show in the Levels histogram, but the Alt/Opt trick for analysis show just what clipped. In the shadow as with any natural light shot you get what the ambient skylight fill and the DR if the sensor wind up giving you. The histogram piled up on the left as it is indicates scene exceeds sensor as would be expected. In lighting like that even more loss of shadow detail wouldn't be unusual.
The "pop" and 3D realism in lighting like this, especially on the water drops, comes from rendering clean, sharp 255 highlights where appropriate while retaining 254 - 230 gradients around them on objects like the leaves. While the highlight clipping is where I'd put it at capture as form of "don't loose solid detail" insurance, I think you can push them a bit more in PP and get more sparkle in the catchlights. I pushed them to 216, equivalent to about a stop more exposure to the point where the leaves were clipping in two channels. I did this by eye moving the H slider back and forth, looking at the image and finding what I thought had more "pop" and 3D realism in the highlights.
There isn't much actual clipping in the image as captured but that's not the whole story technically. I look at the content and what tone it is based on seeing similar content in person (my baseline for comparison of normal looking or not). Things like the black cage supporting the plant I'd expect to be very dark so they don't seem abnormal. But the shaded green stems are dark to the point I'm missing the detail I'd expect to see.
In a technical sense, a histogram piled up on the left as much as it with scene content that isn't almost black as here indicates the 3/4 tones — detail between 12% reflectance center of the histogram and shadows — was rendered by the limited sensor range darker than seen by eye, not in the wide "see it all at once" POV, but the brain's impression after scanning the scene, focusing on the sparkle on the water drops, the separately on some eye catching detail in the shadows.
The way to make a digital image lighter in the 3/4 tones isn't to alter the shadow slider in Levels but to adjust the middle one left. Adjusting the middle slider back and forth darker / lighter found 1.08 looked more "normal" to me. Not noon day in the sun normal but tomato stem in the shadows normal, compensating for the loss due to sensor DR and scene contrast. I you look at the sliders in the screen shot you'll notice that after first moving the highlight slider to clip speculars more, then moving middle slider left to make the mid tones and 3/4 tones just a bit lighter the middle slider pointer winds nearly dead center in the middle of the adjusted end point sliders where it is an a full range image that looks seen by eye normal without any tweeking out of camera. It's a indication to me, that in the technical sense the adjustment did in fact "normalize" the image. I also lowered output in shadows a bit which does make them a bit lighter than 000 in the darkest areas.
In the edit below I added high pass sharpening selectively to the sharply focused tomatoes only to restore sharp specular "pop" the AA filter of the camera kills. I do this and the adjustment of tonal range on my images separately on each size and type of output. I'd adjust the range and sharpening differently for small JPG vs large and print vs screen so don't assume the adjustment are universal to all images. I'm just suggesting a analysis / adjustment workflow you might want to try in the slim odds you might actually find something I suggest helpful, even with the circles and arrows. Better arrows in an edit than in someone else's back, no?
Off to church to give thanks for all things, like those beautiful and tasty as those tomatoes 
Chuck
|
