civilized ku # 628 ~ radiant light - the answers

Jefferson Memorial ~ Washington, DC • click to embiggenQuestions, re: picture making have been asked so herein is the first tech answer entry.

That said, I must point out that my tech answers will not be a step-by-step how-to. They will more along the lines of an overview that should give you, the interested parties, a big headstart on the path to figuring it out for yourself. After all, according to the teach a man to fish dictum, and the words of Brain (from, Monty Python's The Life of Brian) who stated to the true-believer throng - "Look, you've got it all wrong. You don't need to follow me. You don't need to follow anybody! You've got to think for yourselves! You're all individuals! ... You've all got to work it out for yourselves."

So, here goes ....

The questions about the picture posted in picture window # 53 ~ radiant light all referred to the HDR technique, which for the most part is a technique generated by an HDR bit of software, although PS does have its own Merge to HDR function. In as much as the aforementioned picture was NOT processed with any HDR software / plugin, I tend to not think of it as an HDR picture. Although ... the pictured scene did, indeed, have a very wide - light to dark - dynamic range.

As mentioned in the picture window # 53 ~ radiant light entry, to obtain the results that you see, it was necessary to make 5 separate exposures - no, I don't think 3 exposure would have covered the required dynamic range - that spanned the range of light from extreme highlight detail to extreme shadow detail.

The exposures were fairly evenly bracketed in 2/3 stop increments. The camera was, of course, tripod mounted in order to ensure exact registration in all 5 exposures which makes blending elements of separate exposures a much easier proposition than having to deal with registration issues.

After converting all 5 exposure files from RAW to .psd format - identical initial WB, overall color adjustments and early stage sharpening were applied to each individual exposure file in the RAW conversion stage, I opened them in PS and went to work on the blending.

I started with the best overall file for the interior exposure (the lightest overall exsposure bracket) as my "base" layer / master file. I then worked my way "up" the bracket chain (that is, from lightest to darkest) using the Polygonal Lasso Tool (feathered to 100 pixels) to select only the parts of each bracketed file that I deemed appropriate for each segment of the exposure range.

After making a feathered selection on an individual file, I dragged that selection to my Master file - using the Move Tool while holding down the Shift key, which automatically places the selection (on a new layer) in precise registration over the same section of the picture. I did this action on all 4 bracketed files - the 5th bracket was already in the master file as the base layer.

At that point I had a good idea of what the blended picture would look like. However, each individual layer required some individual attention with minor contrast adjustments and, most importantly, adjustments to the feathered edges of each selection.

The contrast adjustments were made using the PS Curves function (the Levels function is for use only by simpletons) and results were determined according to the it looks just right method.

The feathered edge adjustments were made using the Eraser Tool to erase or further soften the areas where a given layer blended with the layer (or layers) beneath it. This is done to make the blend look smooth and natural, not hard or abrupt. The Eraser Tool size used to perform the blending depended upon the the detail on given layer - the general rule is to use a larger softer edged eraser for areas of broad detail (like a section of white wall) and a much smaller harder edged brush to blend smaller finer detailed elements together*.

FYI, and this very important to understand, there is no one-size-fits-all / by-the-numbers way of performing all of the above. Each and every picture needs its own content-driven solutions.

As an example, I mentioned that I used the Polygonal Lasso Tool with a feather of 100 pixels - that feather is based upon the fact a 100 pixel feathered edge works very well for me my 300 dpi files. And even so, on any given Lasso selection I make, I will change the feather level dependent upon a variety of factors (see general eraser rule above).

So, if you are really keen upon obtaining blended-exposure results like that of the radiant light picture, the best thing to do is to start messing around using the overview described here. Unless you are a first-class idiot or an absolute rank beginner, you will most likely work it out for yourself by thinking for yourself.

That said, feel free to ask more questions about the preceding.

And, BTW, to answer a related but separate question from Andrew:

Do you find Photoshop essential? Along the way I lost my serial numbers and basically let my copies of Photoshop "expire," but I have been reconsidering it.

As far as I am concerned, there is no life in picture making without Photoshop. I know of no other software other than PS that allows a picture maker to exercise so much control over picture processing / editing. It is, quite simply, the best and most complete image processing software available.

PS is not perfect - it's expensive. It's a bit of a memory hog as a result of packing so many capabilities under its hood. One could also say that it packs way too many capabilities - with all of their attendant complexity - relative to the needs of most picture makers. Nor am I a fan of its RAW conversion results and methodology (not to mention its companion software, Lightroom). Many find it a daunting bit of software to learn, no doubt due to its aforementioned capability-rich complexity.

However, all of that said, I can't imagine picture making life without it.

*some PS users add unnecessary complication (and more demands on RAM memory) to this whole step by creating and using layer masks to blend layers. There is nothing inherently wrong with this approach and it does offer quite a number of advantages such as the ability to use a variety of brush sizes on a single layer.

It is also a non-destructive blending technique in as much as the layer itself is never changed (erasing is destructive because it erases away parts of a layer). If you don't like the blending created by layer masking, you just discard the layer mask, make a new one, and start again. Or, you can simply "paint", at any time, hidden portions of a layer back in or "paint" other portions out with relative ease.

Layer masks can be saved as part of a picture file. A fact which makes them available for future changes/editing.