This photo is taken 35 minutes later than the last ones I published from Nyhavn, Copenhagen. This is the Opera House of Copenhagen, Denmark, and I have been down there a few mornings to try to get a good shot, but the sunrise has always turned into … a grey one. But this time I was rewarded with a grand ‘almost’ sunrise. I never actually saw the sun, it went behind the clouds before it got up. I saw the sunshine on the buildings behind me, but I never saw the sun. But I did get a wonderful display of colors!
I do like Long Exposure photography. It really fascinates me, what the long exposure does to both water and clouds. And on a morning like this, when the water wasn’t particular smooth, and the Opera house didn’t make a good reflection in the water, a long exposure is a perfect solution. I took out my 10 stop B+W screw on filter and placed that on my 16-35mm lens. A lens I have purchased for this sole purpose, to be able to attach filters on it.
What I did was to first shoot the 7 HDR shots, and then I screwed the filter on the lens, and did a 58 seconds exposure (using a cable release). My plan was to use the water, and maybe the clouds from the long exposure, and mix it with the HDR shot. As it turned out, I must have moved the camera ever so slightly when I put on the filter, because the images doesn’t fit on top of each other. That is a risk.
Then I thought of something I learned from another photographer ‘Photography is always a compromise‘! I looked at the long exposure. It really was an almost perfectly exposed photo. Only a few blown outs in the darks and lights, and some of it even in the corner, which I knew I would crop away anyway:
The histogram tells me a lot. I have clicked on the two triangles in to top left and right corner. They show where the blown outs are. The blue splashes show where there is clipping in the dark areas, and the red splashes shows where there are clippings in the highlight areas.
But there is more important information in the histogram. The triangles that you click, also changes colors. The left one is white, which means that all color channels (red, green and blue) are clipping (blown out), in some dark areas. That is OK, because I going to crop that away.
The other triangle is red, which indicates that only the red color channel is clipping (has blown outs), which leaves me with greens and blues. I can live with that, because it will end up giving me yellows in my sunrise, and that is certainly an acceptable color. Had I lost two colors, the triangle would change color (red + green clipping would give a yellow triangle). Had it been the green channel that had clipped, the triangle would be green, and so on and so forth.
When all channels are blown out, the triangle turns white. And if that is the case, then you have absolutely no information in an area. This you can’t use, unless it’s really small.
With this histogram I’m good, and I can proceed to post-process this photo, as a single exposure.
I still really wanted to do the HDR process on my photo. And the histogram really is perfect for that. It’s very well distributed and has a lot of information on all light levels (the height if the histogram from left to right tells me that). In other words it is just perfect for HDR.
Usually you take more exposures, to get more information across the histogram of the combined photo and avoiding the clipping of the histogram. In this case I have managed both, within a single exposure.
There is a luring discussion hidden here. Is it really an HDR photography, if only I use one exposure? Well, it’s not that well defined really. What does make it look HDR’ish, is the tone mapping of your HDR software, not the fact that you shot 3 bracketed shots of e.g. -2, 0 and +2.
I shot this with a Nikon D800, which has got the highest dynamic range of all cameras. Had I shot it with an entry level Canon camera like the 700D, I would have blown the highlights for sure, and I would have needed a set of -2, 0 and +2. Why? Because the canon only covers 11.2 EVS (according to DXO mark) and the Nikon covers 14.4 EVS, this makes the Nikon 3,2 EVS (exposure value steps) better. With a -2 and +2 exposure added to Canon, it will effectively through, 3 exposures get 15.2 EVS (I have described this in more detail in the HDR section of this article).
Clearly 14.4 EVS was just (!) enough, but enough. Does this make it less HDR, than if I had shot it with the Canon and had needed to do the 3 shots to get the same result? I don’t so.
I did one thing to this photo, before putting it into Photomatix Pro. I adjusted the color balance. There is a color cast from my B+W filter, and I have to compensate for that. There are many ways of doing that, in this case I used the tone splitting.
My purpose was not to get it to look exactly as it looked that morning, but to get something that I liked. I ended up with these settings:
Split toning is a fun way of adjusting the colors of the image. I could have used the White Balance too, but I get different results using the split toning, because it’s a different tool. You can do some pretty awesome things with it. Many of the “Instagram” vintage looking photos are split toned.
This made it look more blue hour-ish and the orange was present that morning, and less and I liked the less wild colors. It’s still pretty dark, but the HDR tone mapping will fix that.
What Photomatix Pro does, is to raise the shadows and do it’s Tone Mapping Magic. I did a little layer blending in Photoshop, with the Lightroom image and the Photomatix Image, but not that much. I pretty much liked what I got from Photomatix Pro, directly.
Just for fun, here’s the 0-exposure of the 7 shot HDR series: