I don’t like tripods. I do own a gorillapod, which is a bit like a tripod, except it’s small, light and unobtrusive. I can use it to clamp my camera to anything, at any angle, and it doesn’t limit me – no, that’s what a tripod does.
Consequently, the majority of my night photography is taken and high ISO, large aperture, and hand held. With a 50mm prime lens, opened wide to f1.8 and an ISO of 1600 I can take photos at night, in an urban street lit environment quite easily. Trouble is, f1.8 gives practically no depth of field, so the chances are that 90% of the photo will be out of focus – not a great look. Even at f5.6, street lights will look like blobs of light, so to get a nice starburst effect from the lights I need to use a small aperture of f22.
Another issue to deal with, is the camera metering. Cameras try to get the brightness of the image an average of a light grey colour. Precisely what I don’t want to get out of my night shots. I want dark blacks, with bright points of light. Getting this right usually takes a bit of fiddling around. I start by dialling in under-exposure of 2 stops, and see what works. I’ll try a bit more, a bit less, and usually end up with a manual setting of more than 2 stops under. At this stage I’m handholding the camera, and the aim is to get the exposure right. Now, I’m quite steady holding a camera, but to get this level of exposure at f22 will normally need something around a few seconds – I’m not that steady.
So, in the absence of a tripod, or that incredibly useful gorillapod, I look around for a suitable wall. Fortunately it’s cold, so I’ve got some gloves with me, which make a very useful rest for my camera to keep it stable, and stop it getting scratched by the wall. Unfortunately it’s cold, so I need to wear some gloves, but I’ve only one pair with me, and they are underneath the camera, so I have to work fast before I get frostbite. I may be exaggerating a touch there, but you get the idea.
One more techy point – if you’re going to try this, find out how to use mirror lock up if you are doing this with an SLR. When you take a photo with an SLR, the mirror moves out of the way, and in doing so causes vibrations – not what you want on a 2 second exposure. Mirror lock up works so that the first press of the shutter button moves the mirror, you then wait a few seconds for the vibration to die down, and then press the shutter a second time to take the photo. Now, it is really, really important that you remember you’ve switched on mirror lock up – I often forget and then experience a horrible moment when I think my camera’s stopped working, until I remember. One other point, use a cable release, no point going to all this trouble and then pressing the shutter with a big clumsy thumb. If you don’t have a cable release, just use the camera’s timer release, that works just as well as long as timing isn’t vital.
I always check the histogram after taking a photo in near darkness, it’s the best way of checking you have the exposure about right, as looking at the small review screen in the dark isn’t going to help much. The graph should normally show the bars largely in the middle of the scale, but with night photography, much of the image will inevitably be black – that’s fine, but just make sure the whole image isn’t black. You can recover some detail in editing, but black is black is black. Even after checking, I’m never sure, so there’s no harm in taking a few extra shots at different exposures, you never know which one is going to turn out best.
Having taken the image, I then adjust the exposure slightly in my raw editing software – I use GIMP and UFRAW, they are free, open source and do everything I want, but use what you are familiar with. I don’t adjust the exposure much, the result of excessive tweaking will be noise which usually looks like lots of red pixels scattered all over the image, not a great look. I sometimes have to do this, and then I convert to black and white, as I think I can just about get away it, but even then it is not ideal.
Once in GIMP, I avoid any aggressive as it creates noise, I rely on a technique which involves layers and blend modes. I’ll not explain all he technicalities here, it’s interesting, but not that interesting. All you need to know is the effect it has, which is simple. I use a layer with blend mode set to Overlay to darken the image, and layers set to Screen to lighten the image. My starting point is three or four Screen layers with an Overlay layer on top. I then add or hide layers depending on the effect I get. I can experiment here, extra layers, change the opacity, try a screen layer on top, anything until I get an effect I like.
On this particular image I used five Screen layers, with one set to 50% opacity – so effectively four and half layers, then two Overlay layers at the top to darken and saturate the colours. Normally one Overlay will do the trick, but on this shot the colours became deeply saturated and the blacks so dark – I like the effect and that’s what I settled on. The great thing with this technique is you can tweak to your heart’s content without noticeably adding noise.
Once I’m happy with the effect, I just flatten the layers, sharpen gently, and save as a jpeg file. Oh, and then share it on this blog, and hope that people like it!