I got sidetracked from another discussion I was having, and decided to find out if adjustment layers behave differently depending on whether you adjust their opacity or the strength of the adjustment itself.
I suspected there would be some difference in results, but I was very surprised at how
much difference there is.
So, the experiment went like this: I opened an image and duplicated it, so that two copies of the same image were sitting side-by-side on my screen.
Then, I'd apply the same adjustment layer to both images ... still looking exactly the same, of course.
Then, on the first image I'd reduce the opacity of the adjustment layer to 50%; on the second image I'd go into the adjustment layer and reduce the strength of it by 50% (eg a Contrast adjusment of 40 would be reduced to 20).
To read that description, you might expect that both images would still look exactly the same -
not so! In some cases, the difference was remarkable.
Here's an example, using a Photo Filter adjustment layer. On this one, I used the Warming Filter (85) at 50% density, and left the layer at 100% opacity:
Here, I set the density to 100%, but the opacity at 50%:
I'm sure you can see the difference, especially in the sky.
So, I found that the strength and opacity aren't equivalent for Photo Filter. Nor are they equivalent for Levels, Hue/Saturation or Selective Color. Channel Mixer was very close, but not quite the same.
Curves is peculiar. The endpoints don't give the same results when comparing strength vs opacity; but a midtone adjustment is identical! (Well, a modest one, anyway. I didn't test a really strong one that induces clipping).
Brightness/Contrast is also interesting. The Contrast slider gives very different results (strength vs opacity), but the Brightness slider gives
almost the same. (I didn't dwell on this one too much, though, because B/C is a completely useless tool, and should never be used.)
The only adjustment layer that seems to give completely reliable results no matter whether you adjust the strength or the opacity is Color Balance. I haven't figured out why.
In all the cases where there was a difference in the effect, I'm not saying that one looks better than the other. They just produce different outcomes, that's all, as the above examples prove.
Ok, so what's the point of all this???
Well, for you, dear reader, possibly nothing. For me, not much, because I usually get my adjustments right the first time, and have little need to re-adjust them later, either by strength or opacity. But the differences I saw tonight will make me extra careful about my re-adjustments in the future.
The reason I wrote this post is that I know there are some people who habitually make their adjustment layers
over-strong, then use opacity to reduce the effect to a satisfactory point. I'll be encouraging those people to re-assess their workflow, to see if they're truly getting the best results.
Here, I set the density to 100%, but the opacity at 50%:
I'm sure you can see the difference, especially in the sky.
So, I found that the strength and opacity aren't equivalent for Photo Filter. Nor are they equivalent for Levels, Hue/Saturation or Selective Color. Channel Mixer was very close, but not quite the same.
Curves is peculiar. The endpoints don't give the same results when comparing strength vs opacity; but a midtone adjustment is identical! (Well, a modest one, anyway. I didn't test a really strong one that induces clipping).
Brightness/Contrast is also interesting. The Contrast slider gives very different results (strength vs opacity), but the Brightness slider gives almost the same. (I didn't dwell on this one too much, though, because B/C is a completely useless tool, and should never be used.)
The only adjustment layer that seems to give completely reliable results no matter whether you adjust the strength or the opacity is Color Balance. I haven't figured out why.
In all the cases where there was a difference in the effect, I'm not saying that one looks better than the other. They just produce different outcomes, that's all, as the above examples prove.
Ok, so what's the point of all this???
Well, for you, dear reader, possibly nothing. For me, not much, because I usually get my adjustments right the first time, and have little need to re-adjust them later, either by strength or opacity. But the differences I saw tonight will make me extra careful about my re-adjustments in the future.
The reason I wrote this post is that I know there are some people who habitually make their adjustment layers over-strong, then use opacity to reduce the effect to a satisfactory point. I'll be encouraging those people to re-assess their workflow, to see if they're truly getting the best results.