Frozen waterfall

[DanK]

I've been working on a few shot from this setting and am finding them very difficult, so reactions or suggestions would be welcome. For those who know New England, this is Waconah Falls in the Berkshires.

I did quite a bit of work on this but am not satisfied yet. I pulled down the very bright snow a bit, added a lot of texture to the ice, lightened the rocks by creating a color mask, did some dodging and burning, applied midtone contrast both globally and locally, the latter with luminosity mask, changed WB, and sharped with smart sharpen. I probably did other things I'm forgetting.

I linked to a fairly large copy, so you can see it better by double-clicking to get into the light box.

[Geoff F]

The only thing I wonder about is a slight crop from the left side. Lose that uninteresting bit of a small tree and possibly a similar amount from the top to concentrate more on the actual snow and ice with less distraction?

[DanK]

I tried another edit:

[Polar01]

Dan I played just a little with the image, I found no true black in the image, so I added black, lowered both the whites and highlights this brought some detail out in the snow, then just warmed it up a little. That is just me.

Cheers: Al

[DanK]

Al,

Thanks. This is helpful. These are things I have been working on. I've tried to bring out detail without darkening the image too much, but I agree that pulling some of the snow down more could help. Re temperature: I've been all over the map. I have two versions that are warmer than this. I'm trying to avoid artificially cold snow tones while retaining the actual feeling of the location, which had cold deep shadow lighting. I'll play with that more.

Dan

[Round Tuit]

Dan,

I too find cold ice difficult to capture. I have had some luck selectively applying a generous amount of local contrast enhancement (clarity) to relatively clear ice. It brings out the shine. You would need to avoid applying it to snow or to the whitish ice.

I would also consider opening up the midtones.

[DanK]

Andre,

thanks. I'll play with both. I did to local contrast adjustment, but globally. I should do it locally, as you suggest. I tried opening the midtones in a couple of different ways with unsatisfactory results, but I'll go back to that as well.

Re local contrast: I use three tools, texture and clarity in LR/ACR and unsharp mask with a huge radius in photoshop. Texture and clarity appear to be very different, although I haven't seen clear documentation from Adobe. What I have gleaned is that texture is higher-frequency than clarity and that clarity includes a hefty does of midtone contrast. The result is that even a modest amount of clarity can look overcooked, while one can often go to very high levels on the texture slider. I've been meaning to experiment with different radiuses in USM to see if I can approximate clarity without the midtone contrast boost.

Dan

[Round Tuit]

Andre,

thanks. I'll play with both. I did to local contrast adjustment, but globally. I should do it locally, as you suggest. I tried opening the midtones in a couple of different ways with unsatisfactory results, but I'll go back to that as well.

Re local contrast: I use three tools, texture and clarity in LR/ACR and unsharp mask with a huge radius in photoshop. Texture and clarity appear to be very different, although I haven't seen clear documentation from Adobe. What I have gleaned is that texture is higher-frequency than clarity and that clarity includes a hefty does of midtone contrast. The result is that even a modest amount of clarity can look overcooked, while one can often go to very high levels on the texture slider. I've been meaning to experiment with different radiuses in USM to see if I can approximate clarity without the midtone contrast boost.

Dan

I use the same three tools. I find that usm with high radius is closer to clarity than texture. I have not figured out how to mimic texture in photoshop.

[DanK]

I’m guessing a slightly smaller radius


Sent from my iPad using Tapatalk

[Round Tuit]

Possibly but a radius as low as 10 still creates a significant boost in midtone contrast.

[DanK]

Possibly but a radius as low as 10 still creates a significant boost in midtone contrast.

Try doing it on a layer with a luminosity blend. All tonality adjustments in RGB space with a normal blend change midtone contrast. A luminosity blend avoids that; it's like working in the L dimension in Lab space, but without all the fuss and bother. So if I'm right, changing the radius with a luminosity blend will only change the frequencies affected. My logic is that a smaller radius should affect higher frequencies, hence more like "texture".