One of the things I really want to learn to do better is astrophotography. I would love to see what I could do with a full frame sensor DSLR camera, but for now I will work to get all I can out of my trusty Canon 60D crop sensor and Tokina 11-16mm f/2.8 lens. I recently ran across a tip from Justin Ng that offered an alternative approach to shooting the stars from a location with heavy light pollution. He suggested using the concept of ETTR (expose to the right) explained very well here by Martin Bailey.
ETTR isn't new to me, I actually use it in all the other areas of my photography because I have done the experiments and believe that it is the best way to get the most out of the digital sensor inside your camera. It's a math thing. However, I hadn't tried to use ETTR in my astro photography, mainly because every tutorial I had seen up to this point when I ran across the post by Justin recommended using the maximum shutter speed you can get before the stars will start to go from circles to ellipses.
The way to determine this magic shutter speed is by following something called the "500" rule where you divide 500 by the focal length you are shooting at. If you are shooting at 16mm then you divide 500 by 16, which is 31.25 and round that down to 30 seconds as a max - better to do something like 25 seconds to be safe. (NOTE: I used to think you needed to factor in the crop sensor and multiply the 16mm focal length times 1.6 for the Canon crop factor, but have recently learned that is not necessary as it is the optics in the lens that matter here rather than the sensor size).
I have tried shooting the stars a number of times using the 500 rule and opening my shutter as long as I could to try and get a good shot of the stars. You can see some of my earlier attempts that turned out pretty well here and here. Both of those shots came when I was out at scout camps, which put me a little further away from the significant light pollution any decent sized city will produce. However, I could wait to experiment with this tip as soon as I saw it, so I decided to try it that night out on my back deck.
The 4 shotsThe 4 shots showing what the exposure looked like straight out of the camera along with the histograms from LightRoom I live in a suburb of Salt Lake City, Utah called Herriman, which means taking a shot in the southwestern sky away from the "big city" there isn't as much light pollution. There are also nice little foothills that are visible from my backyard deck here, so I thought that would add something more interesting to the shot. I waited until it was well past sunset and was disappointed to find there were a lot of clouds out that night. I was hoping that I might be able to find a little of the Milky Way galaxy in the shot but the cloud cover made sure that wouldn't happen. I still wanted to try out the tip, so I got the tripod setup and used my remote shutter release to dial in a few different shutter speeds to see which one would give me the ETTR without over-exposing and losing detail.
In Justin's tip he said with his full frame sensor he was using ISO 6400 and a shutter speed of 9 seconds. So setup my gear with I used an aperture of 2.8, a 16mm focal length, and ISO 6400. I was really nervous setting at 6400 since my crop body doesn't deal as well with noise as the full frame big brothers. But I was experimenting, so I gave it a try. Knowing my crop isn't quite as sensitive to light I decided to try 7, 9, 11, and 12 second exposures just to see what that got me in my location with my gear. The results straight out of the camera were not impressive, but that is frequently how it goes when you use the ETTR technique.
LR Basic PanelThe settings I liked for the 7 second shutter in LightRoom for the Basic Panel I copied the photos to my computer, then imported them into LightRoom 5.4 (5.5 was released at this point, but I have found it to be VERY slow). I looked at all of the exposures and noted that the 7 second exposure was the one that was ETTR without being over exposed. The histograms of the other three shots were slightly overexposed to the point that detail was being lost in the brightest spots of the shots. It made me wish I had taken one at 5 seconds just to see how that would have gone. I overestimated how big an impact the crop sensor would be vs. the full frame that Justin was using.
You can see all of the settings I used in the Basic panel, the biggest ones being the exposure (had to lower it almost 2 full stops) and the highlights and shadows. You should also notice that the histogram has changed significantly with most of the data actually showing on the left hand side, even to the point that there is a little bit of the darker spots in the picture being "clipped." The arrow in the upper left of the histogram is LightRoom telling me that I have altered the overall exposure of the picture to the point where some of the data in the blue channel of the photo is now no longer distinguishable from the darkest of the data in that channel. Meaning in the blue channel there are now pixels that used to be very slightly different from each other that are now the same, the slightly lighter pixel is as dark as the one that was slightly darker.
For this picture I actually want the sky to be as dark as I can get it without having the stars disappear. It is something you have to tune to your taste and what you were trying to get as you captured the shot. I usually tune things using the LightRoom sliders with the objective of making the photo look like I remember the scene with my own eyes. However, this time I wanted to make the scene be enhanced from what I could see with my own eyes. I wanted the stars to really pop out and there was enough light pollution that I really couldn't see any but a few of the brightest stars. So I played with the sliders until I could get that result. The interesting thing was that I was able to get that here with these settings better than I could in my previous attempts when I had the ISO lower and the shutter speed much slower (30 seconds). So, Justin's tip seems like it is something I want to continue doing because I am rarely far enough away from the city to get shots without quite a bit of light pollution.
LR Detail PanelCrop sensors aren't well known for producing good shots at high ISO like 6400 without too much noise, but it works in this scenario I also wanted to point out the results of my concern with noise. I can't remember a time I have ever shot my Canon 60D at ISO 6400 in the past. I have shot at ISO 3200 very rarely, mostly only in previous attempts at astro photography. Generally the highest acceptable ISO I will use with this crop sensor is 1600, which I frequently decide is just too noisy to really use even then. So I was really curious to see how this would turn out when I cranked that ISO all the way up to 6400. But you know what, it worked for this type of shot.
I guess the thing to think about regarding noise is not just that at a certain ISO level your shots won't be good because they are too noisy. Instead, I need to remember that the noise really shows through most when you are RAISING the exposure, highlights, and shadows sliders. In other words, every time you are going to INCREASE the exposure in some way that is when the noise will become so bad that the shot really can't be used.
Actually, when I think about it now, this is the entire premise of ETTR. ETTR is playing to the strengths of the digital sensor. If you can just slightly over expose the shot, you can very easily bring down the exposure and actually end up with a better quality shot than if you had tried to get it properly exposed according to the light meter in the camera. That may not be true if you used an external light meter, something I have never done. But there are many scenes where the internal light meter of the camera is simply going to be wrong. Shooting the moon, you really need to go about 2 stops lower than the camera thinks you should to have a good exposure. Same thing here with astro photography. In this specif shot the clouds were really fooling the camera's light meter.
Even though the noise didn't show through badly like I was worried it would, there was still plenty to deal with. You can see the setting I landed on as I processed this 7 second shutter shot in LightRoom. I did still crank up the luminance noise reduction slider quite a ways. But the color noise reduction slider is actually on a pretty normal setting for a RAW file, which makes me happy that I didn't have to go to extreme lengths there in order to make the sky look very even with regard to color.
At this point I was very happy with the shot just processing it in LightRoom. But if you read through Justin's tip he didn't end there. He took his photo into PhotoShop and used the HDR Tone Mapping feature to give his shot just a tiny bit more contrast in a different way from what LightRoom can do (the contrast slider combined with the highlights and shadows can often emulate HDR pretty well in LightRoom, but it isn't exactly the same as doing some actual HDR tone mapping). I tried that out, but frankly I am not good enough with PhotoShop to use the HDR Tone Mapping function. Instead I used the HDR processing software from HDR Soft called PhotoMatix Pro.
Photomatix Pro SettingsI decided I like the way that PhotoMatix Pro did the Tone Mapping with the Details Enhancer method the best, along with all the settings you see here Even though this isn't a group of bracketed shots exactly, although I do have 4 shots at different exposures, I wasn't going after a shot of mixed exposures in this photo. I wanted to make the shot a little bit more than what I saw with my own eyes, but not so much that I wanted to mix different exposures. So, I exported the one 7 second shutter exposure into PhotoMatix Pro and played around with the many options available there by going through my presets and finding one that was the closest to what I wanted. Then I changed the sliders a little bit from there until it looked like I wanted.
Again, I was actually very happy with the shot just through the processing done in LightRoom, but taking it through PhotoMatix Pro just gave it a little more pop. This little experiment from my backyard deck went far better than I thought it would. I didn't think shooting at ISO 6400 would go well on my crop sensor, or that increasing the ISO and shortening the shutter speed would have any real impact on the exposure vs. what you get with ISO at 3200 and a shutter speed of 30 seconds. I am very pleased with how the shorter shutter time wasn't even close to having the stars turn into star trails rather than nice little circles. So I think I will continue trying this. Maybe I'll do some of both methods next time I am further away from the city just to see how that goes and learn a little more of how I prefer to get the shots I envision.
I hope you enjoyed this "How I Got the Shot" blog post. Keep watching the blog here as well as my Google Plus feed if you are interested in learning with me as I do more experimenting in the future. While you are here please check out my library of stock landscape photos which are for sale in my store for both royalty free and single use downloads as stock photos.