New tools of LR and ACR

[Nicolas Claris]

I have been first very impressed when LR (Light Room) and ACR 4 came out, but I would like to introduce a discussion about the power and the + and - of 2 of their new tools that seems to be now really fashioned to a lot of us (in the mean time I've returned back to my preferred raw processor C1 V3.7)

1/ Clarity (this is a copy of a post I made earlier in a "sharpening thread")

Yes, Clarity "is a great and useful rendering tweak"
but as all tools it must be carefully used, and checked viewing at 100% in the raw processor, as it very easily produces halos when overused (as many contrasting and/or sharpening techniques)…

I do prefer to adjust midtone contrast and sharpening, after all WB/color/resize process just before converting to output space and switching from 16 to 8 bits…

2/ High Light recovery
It is a very nice and usefull tool, but I can see posted here in OPF a lot of pictures where this function has been overused, yes some highlights are back, but the overall tone of the image is flat and colors moreless washed-out. They look like bad HDR renderings…

I come to the conclusion, that it is always better to achieve a good shot, in camera than in post-processing, the "adjust to the right" law is still alive!

As usual, it seems that all the powertools can and do help photographers but must be used with a lot of caution, looking carefully what has change in the ALL image before pushing the "OK" button as we usually are focused on one little part of the image when looking for a local enhancement…

[Michael Fontana]

>that it is always better to achieve a good shot, in camera than in post-processing, ......<

Nicolas, there isn't a substitute, there isn't really a alternative to a good capture!
I only can place emphaisis on your point.

The older rule from ole analog days hasn't gone; sh*** gets in--> sh*** gets out...

meanwhile today, we had some helpers - you mentioned some - to improve some shortcomings_
I personally don't mind these helpers, as todays, we have to shoot in light-situations, when we would have left the film cam in the box. But I look at them as helpers, for critical situations, nothing more, nothing less; there' no wonder.

Off course, one can work arround for hours with a single image, ending up with a flat, dullooking clone.
Is it interesting? and still not the same...

[Nicolas Claris]

Thanks for your reply Michael,
coming from a so meticulous photographer as you, your point is really important!

I do to use some of these helpers (RAW is a great helper!), but wanted to warn newbies to theses technologies to tell that Clarity and highlight/shadow recovery can't do miracles but can destroy a file when overused… yes the old adage is still true…

[Michael Fontana]

As you mentioned highlight/shadow recovery; I first thought on it beeing in PS:

I use it rarely, but when: alwith on a backgrund layer, and often I reduce its impact by reducing the opacity of that layer to about 70%.
This 70% is interesting, as I very often use it in editing; giving a 2 nd look to a tweak, 70% of the first edit looks in almost every sitaution better.

Do you know if there's a significantt difference between LR/ACRs SH-recovery, and the one in PS?
Andrew might tell...

[Nicolas Claris]

Good question!
I find the PS version more powerful as you have, when more options checkbox is checked, more sliders for a more accurate result, you may also work in the same panel on the midtone contrast and color saturation to prevent evident dull looking image.
If this complete feature were implemented in LR and ACR that would be much better and powerful to use because it would be before converting raw…

Otherwise, working on a layer is the best advice to give!

[Andrew Rodney]

Highlight recovery in LR isn't going to bring back anything that's clipped. Its just a tool that makes bringing down that portion of the tone curve easier than say using curves (which can be used after Highlight recovery). If you clip important detail you wish to render, no tool will bring that back. I find I use this slider a lot because I expose to the right for the linear data. The default rendering is almost always too light. I have a "Normalize" preset that automatically sets most of this for most images but I tend to still use the slider and maybe (far less so) curves in LR. Its also useful to attempt to render images in the order the tools are specified (top down, left to right). You don't HAVE to do this, but its usually far more effective than bouncing around to curves first, then the basic controls.

[Michael Fontana]

Thanks Andrew

I can see the point, where ETTR is corrected by highlight-recovery within LR.

As we've the discussion now: Does ETTR make sense in these cases, when the scene's DR is higher than the cam's one; therefore bracket exposures are the way to go? Sometimes when developing for HDR, I feel that a conversion with a slightly higher contrast - compared to a singleshot-conversion - leads to better results im the tonemapped file.

Apart from the linear space at the RAW-state, vs the 16bitdata in PS, what is the difference between therecover-tecnique's in LR vs PS?

[Andrew Rodney]

Bracket is great when you can do it. Often you can't.

No matter the scene DR, you've got linear encoded data, half of that falls within the first stop.

There are plenty of disadvantages to ETTR, many I have discussed in an upcoming piece in Digital Photo Pro mag.

The major and huge difference between LR and PS is one uses linear encoded, scene referred data. The other uses baked pixels. There's so much LESS you can do in the later. Oh, having metadata edits is nice too but one could consider adjustment layers the same.

[Michael Fontana]

Originally Posted by Andrew Rodney

........No matter the scene DR, you've got linear encoded data, half of that falls within the first stop.......

I agree; still, the 96bit data - "space" could be big enough, to keep the brightest white of the overexposed shot (no clipping in 96 bit); in that case, I think a highlight-recovery at the RAW-stage might "flatten" the final images contrast. It's rather out of experience, than real science ;-)

Simpler: "half of that falls within the first stop." --> will still be in 96bit-data?