mystik610 wrote in post #18311124
What he did is a little extreme in terms of exposing to the right, but generally that's the way you'd capture a shot if you want to preserve as much DR as possible. If you blow the highlights, they're lost as you don't have as much head-room to recover them in post. He could not have gotten that shot SOOC...not without blowing out the highlights. The extra DR you're able to extract in post is in the shadows, not the highlights, so the 'trick' is to expose for the highlights and recover the shadows in post. This often equates to an EC value of -2 or -3 when shooting outdoors like that, which means you need to push the shadows hard in post to get the foreground/subject detail you need.
When shooting to the right you need to use POSITIVE exposure compensation, not negative. This is so that you can push the brightest required highlight to the point where it is just short of clipping, moving all of the image data as far as possible to the righthand end of the histogram. You then bring the data back to the left so that it ends up in the correct relative position on the histogram during the RAW conversion process. This effectively allows you to bring up the brightness of the darker parts of the image, to fit in the available output DR, without having to actually raise the values from those recorded.
It is important to remember that some RAW converters are able to work with highlight details that other converters will always clip. I have always found that DPP3.x was one of the worst for working with highlight detail, on the other hand I find that Adobe's Process Version 2012, introduced with LR4 and then ACR in PSCS6 has been one of the better options for working with highlights. I understand that Phase one's C1 Pro is also very very good at highlights, but is a little too steep for my pockets. Because of this variation in ability to work with highlight detail across different software, the overexposure warnings shown in camera can be pretty worthless.
The issue with both the blinkies, and the histogram shown on the camera LCD is the fact that both are based on the fully processed in camera JPEG image. The in camera processing can make a huge difference, because if for example you chose a very high contrast picture style, that will tend to make highlights clip much sooner than if processed with a low contrast style. If you look at the full range standard picture styles, and other processing variables, it is possible to shift the clipping point of the processed image by up to two full stops. Even the most conservative in camera settings seem to indicate clipping on the LCD while the RAW data when processed in LR or ACR with PV2012 can still be very nearly a full stop from actually clipping.
Generally I keep my in camera settings to produce the lowest contrast image possible, using the Faithful picture style, with the other four parameters at their minimum settings, either -4 or 0. I then know that if I take my exposures so that the wanted highlights have just started to show blinkies my exposure will generally be as far to the right as possible, without clipping the actual RAW data. When using the camera's metering system this usually means running the exposure between +1 and +2 stops. In some situations I might go even higher, if I am dealing with a very large DR, where most of the detail in the images is in the deep shadows. Unfortunately this can mean having to allow some of the highlight detail that you would otherwise want to keep to clip, since there are hard limits to DR that simply cannot be overcome with current technology.
Alan