Making Time Lapse Motion Capture of the Night Sky

In previous blog posts, I’ve written about taking still pictures of the night sky, and how to take time exposures in daylight. Now, let’s look at the ten basic steps for capturing the motion of the stars:

1.       Leave town. I typically drive 1 to 2 hours north of the city to get away from its ‘light dome’.

2.       Make sure you have a sturdy tripod for your DSLR.

3.       Use as wide-angle a lens as you can get. For a full frame sensor, consider using anywhere from 14mm to 18mm focal length. For APS-C sensors (ie. Canon Rebel or equivalent), a 10 to 12 mm focal length will work. Camera brand lenses can be outrageously expensive, so consider much cheaper but still good third party brands like Rokinon or Tokina.

4.       Consider getting an intervalometer, or at least a wired shutter release for your camera. Either one will cost you less than $20 on eBay.

Intervalometer

5.       If the air temperature is expected to drop below the dew point, be ready to clear condensation off your lens. You could use a microfibre cloth between shots, buy a battery operated fan at the dollar store to keep the air moving across the front of the lens, or look into some form of lens warmer.

6.       Set your ISO to about 3200 and your shooting mode to Manual. Since you want to keep exposure to less than 30 seconds to prevent the stars from creating noticeable trails on each frame you shoot, select a value between 20 and 30 seconds and set aperture to maximum (ie. f/2.8). I use f/4 with success, but it might force the choice of 30 seconds over 20 seconds. You may also want to use a fixed White Balance mode instead of Auto, but I generally have not found Auto to be a problem.

Normal Infinity Focus Setting

7.       If your lens has a focus scale, set it to the vertical bar near the infinity mark, but not at the infinity mark. Take a test shot and zoom in on the playback to see if you have optimal focus. If not, tweak it very slightly and re-check. The actual infinity focus point varies slightly with focal length, so leave the lens at your intended focal length when setting the focus point, and only zoom in on the playback to verify it.

8.       If using an intervalometer, set your camera to Bulb (may be found in Manual mode settings or may be a separate mode depending on the camera). Set the intervalometer shutter  ‘on’ time to 20 to 30 seconds (depending on your test shots), set its interval between shutter actuations to something like 5 seconds (to allow you to wipe condensation off the lens between shots), and set the number of frames to about 150. I normally shoot with RAW+JPEG, but you may want to consider doing such a long sequence in JPEG only to conserve memory card space. Elapsed time will be close to an hour and a half, so bring a coffee flask!

9.       If using a simple wired shutter release, set the camera to burst or continuous mode instead of single shot mode. You will not use Bulb mode in this case. Instead, select either 20 or 30 seconds for your shutter speed in Manual mode. When ready to start, use the lock on the wired release to hold the shutter button on. The camera will take one shot immediately after the other with no breather in between. This will get the sequence over with more quickly than when using an intervalometer with a 5 second breather, but you won’t have an opportunity to clean off condensation between shots. Depending on the camera, you may also have to manually keep track of the number of shots taken.

10.   Find an interesting object on earth that you can anchor your shots to. Put it in the lowe part of your frame, about a third of the way in from the edge. You can have it silhouette against the sky, or use a flashlight to do a little light painting. In a sequence, however, you may want to forget the light painting because it is difficult to make the object look uniformly lit in all 150 or so frames. Now, consider putting Polaris somewhere in your shots, so you can see the stars rotate around it in your final sequence.

That takes care of the capture portion. Now for the post-production steps:

1.       I import all 150 frames (images) into Lightroom. Then, if corrections for exposure, white balance, etc. are needed, you can make the changes to one frame and sync the changes to all 150. Similarly, I have set up a custom crop mode in Lightroom for 16x9, so that the resultant video will frame properly on television. Again, I can crop one and sync-crop all the others.

2.       A sidebar on ‘stuck’ pixels: You are going to notice that there will be bright spots on all frames, which don’t move with the stars. These are generally due to the long exposures and the number of such exposures, which result in heating up of the sensor. If these are left alone, it will look weird once you put all the stars in motion, since stuck pixels don’t move. You could use the camera’s long exposure noise reduction feature to mitigate this, but you will be out there for three hours instead of one and a half hours capturing the 150 frames. My approach is to use the clone brush in Lightroom on one frame and sync to all other frames.This is the most time-intensive and frustrating part of the process due to thenumber of stuck pixels my camera produces, and the way you clone them out may adversely affect other frames once you sync. There is a bit of an art to it, and I could write a whole treatise on it. But not now.

3.       Once I’ve cleaned up all the frames, I export them from Lightroom as TIFF files. At this point, you may want to decide if you simply want to make a video clip of the stars rotating around Polaris, or create progressively longer and longer star trails as they rotate. For the latter, you could use StarStax. This software creates another set of files in which the star trails get progressively longer. This is done by ticking a checkbox in the settings menu to force it to save a file every time it adds another file to the sequence. If you don’t check the box, it will only create one file that consists of all 150 frames forming a single long trail for each star.

4.       Now to put things in motion. There are a number of software packages that can accomplish this, but I use my Corel VideoStudio video editor. In that software, you simply right-click on an empty timeline and select the time lapse option(‘Insert Photo for Time-lapse/Strobe’). You then tell it where your frames are and how long you want to make the clip (I generally start with about 10 seconds), and the frames load onto the timeline. If you want to have the stars rotate without making trails, use the files you exported out of Lightroom. On the other hand, if you want the stars to create progressively longer star trails, use the files created by StarStax.

5.       Now you can create a finished video (Share tab in Corel). I prefer to save mine in .mp4 format. Have a look at my two samples on YouTube. One is without StarStax (Milky Way video) and the other is with StarStax (observatory video).  I hope these inspire you to try out this fun technique.

TRY TIME EXPOSURES FOR ETHEREAL EFFECTS

When we think about taking long exposures with a camera, night time images usually come to mind. Only dark conditions generally call for slow shutter speeds, right?

True-but there is a way to take time exposures in broad daylight...with surprising results.

So the first question that comes to mind is: how do you take a long exposure in bright light? Won’t the camera try to set a fast shutter speed to ensure that the exposure is correct, regardless of how small the aperture is set?

Then there’s the other question: why would you want to take long exposures in daytime, anyway?

OK - a long time exposure (keeping the camera’s shutter open for a long time – from seconds to minutes) will occur when:

• Light level is very low
• A very small aperture is chosen, and
• ISO is set to the lowest value that the camera will allow.

So, to answer the first question, you need to simulate low light conditions by giving your camera dark sunglasses. The device that will do this is an inexpensive thread-on piece of glass called a neutral density (ND) filter. These filters come in a wide variety of grades and are either specified in ‘stops’ or ‘times’.  They are also available as variable grade versions, often referred to as Fader NDs. Be sure to buy one that has the same filter thread diameter as your lens.

A one-stop (or two times) ND filter, with aperture and ISO remaining the same, will require that you force the shutter to stay open twice as long to get the same exposure as you would without the filter. Similarly, a two-stop filter will require slowing the shutter by four times, since a stop means a factor of two in exposure. For the kind of images I’m proposing we take, you will need a ten-stop filter in order to get exposures that are minutes long. This is equivalent to increasing exposure by a factor of 1024 times, but is usually referred to as a 1000 times filter.

Why minutes? Well, in answer to the second question above, anything moving (water, clouds) will take on a foggy (in the case of waves) or streaky (in the case of clouds) look. As long as you have stationary elements in the scene, such as rocks, ground and a tree trunk in the sample shots shown in this post, the viewer is confronted with a paradox that doesn’t appear ‘natural’. This makes for a more interesting image than if you had captured the same scene with a shutter speed of, say, 1/250^th of a second. While these images were shot in colour, I converted them to black and white, since they tend to look more dramatic that way.

And now for the practical considerations. You will need:

• A stable tripod
• A remote shutter release that plugs into your camera, since you can’t hold down the shutter button on your camera for minutes without jarring it. You can also get an intervalometer, which allows you to set a specific time interval without having to manually time the shutter.
• A camera with a Manual mode that will allow you to set the shutter to ‘Bulb’, or one that has a dedicated Bulb mode.
• A day with calm winds, so that your tripod doesn’t get buffeted during the exposure.

Look for a scene that has a combination of moving and stationary components. Getting the right exposure is definitely going to take some experimentation. As a starting point, set your camera’s ISO to the lowest value it provides. This is typically 100. Do NOT use Auto ISO. In Manual / Bulb mode, set the aperture to a value anywhere in the range of f/16 to f/22. This will depend on whether it’s an overcast or sunny day. You should get exposures in the range of 2 to 4 minutes. Use the histogram to confirm good exposure and tweak the time (or aperture) as needed to get perfect exposure or desired motion effect.

With digital SLRs, a practical consideration is the heating of the sensor during long exposures. While it won’t damage the camera, this heating tends to amplify what are often referred to as ‘stuck pixels’. Every camera has them, and they appear as unexpected bright spots on the final image.

To combat this, DSLRs generally have a long exposure noise reduction setting which can mitigate this annoyance. Unfortunately, it requires that the camera will make you wait twice as long to see your image, since it takes as long again as the exposure time to process out the stuck pixels. This could result in an 8 minute wait until you take the next shot. While this technique is effective, I prefer to take my chances and fix the spots after in either Lightroom or Photoshop.

By way of a summary, here are the steps I take:

1. Set up the camera on the tripod and compose the scene with the chosen lens.
2. Ensure ISO is at its lowest setting, and choose a small aperture to help force a long exposure and to give me the depth of field I want.
3. Plug in the remote shutter release or intervalometer (and set it to the exposure time I want).
4. Thread on the ten-stop ND filter, being careful not to disturb focal length or focus settings.
5. Cover up the eyepiece if the sun is behind me, to prevent light leak onto the image (it has happened to me!)
6. Check my image and histogram, rinse and repeat until I get the right exposure and effect.

It’s a slow process that requires patience. At the end of it though, you will get images that will be set apart from the ordinary.

New Digital SLR Photography Book for 2014

"Digital SLR Photography Demystified" is my new venture in the world of photography instruction.

Originally conceived as a reference for students taking my Digital Photography for Beginners course, it also serves as a standalone guide for those who prefer self-instruction. Written in the same way as my easy-to-understand teaching style, it includes ten practice exercises to reinforce the concepts presented.

The content goes beyond the "Beginners" course by covering panoramic, high dynamic range (HDR) and night photography. Also included is a comprehensive Glossary of Terms and a Quick Solution Guide.

While no one book can explain the functions of all camera makes, a generic model is used to help the reader locate a button or menu item on their own camera, using industry-standard icons where applicable.

Currently, the book is available as a 96 page paperback from Lulu.com for $14.95 CDN + shipping (slightly less in $US). In the coming weeks, it will also become available on Amazon and Barnes & Noble websites under the ISBN 978-1-312-03258-3.

Use the QR code below to review the book on Lulu.com.

PERFECTING THE PANORAMIC

It’s getting easier to make panoramics with today’s cameras. Some will actually do the stitching together for you without the need for software. But unless you take control of the process, you may be selling yourself short on results.

Here is a checklist you can follow in preparation for shooting your panoramic. It may look a little daunting, but it will become old hat with a little practice:



• Arrive Early   In order to get through this checklist, you’ll need the extra setup time before the light changes!
• Use a Tripod   A panoramic is comprised of a series of adjacent photographs stitched together in software to form a single wide-angle image. In the stitching process, the software does its best to match adjacent images seamlessly, but there is usually a bit of twisting required to match them perfectly. This requires cropping the final panoramic. Unless the individual images are shot with the camera absolutely level, the twisting and subsequent cropping may get to the point where much of the scene is lost.
• Level the Tripod and Camera  I have a bubble level on the base of my tripod which I use to get the legs levelled up first. I also use a bubble level that slips onto the camera’s hot shoe, which I use to do fine levelling adjustments. The reason for using both is that uneven legs may give me the impression that I have a level camera at one position, using the hot shoe level. But as I pan, I would likely see the level drift off centre. Once you have tweaked both the legs and the pan/tilt head, try panning through the intended range of your panoramic to ensure you have optimized both.
• Set the Camera to Manual (M) mode   Imagine you’re shooting a panoramic made up of photos that have dramatically different light content. In any mode but Manual, each photo would be exposed differently because the light metering system built into your camera tries to set the exposure automatically to an average 18% grey level. In this scenario, the panoramic wouldn’t look right because the inherently darker scenes would be forced to look lighter compared to the others, resulting in a patchy-looking result. Sure, you could try matching the exposures in software before stitching, but it’s easier to get it right in the camera. I pan the entire scene first, using the digital meter to set a shutter speed/aperture combination that averages a normal exposure. As a result, some shots will be over-exposed slightly, others under. As long as the histograms for all photos don’t show any clipping at the black or white points, you’re OK.

  

  The camera's digital light meter. Zero is optimal exposure.

  

 

 

 

 

 

 

•  Set the Camera ISO setting to a fixed value  Some cameras have Auto ISO setting. Per the last point about exposure, ensure you choose a fixed value (eg. 200) rather than Auto.
• Set the Camera to manual focus  Most SLR lenses have a switch allowing you to disengage the motor that automatically focuses your lens before you take a shot. For panoramics, set it to manual focus (MF). The reason for this is that as you pan to take each of the shots, the camera may refocus on something close in or far away, especially if there’s an object in one of the shots that is prominent. That could result in uneven focus across the panoramic. Pick one focus setting that works for the whole panoramic and leave it there.
• Take the Camera out of Auto White Balance  Similar to the argument in the last two points above, you don’t want the camera to change the colour balance as you pan because of changing content. Choose one of the presets instead of Auto. For instance, if shooting on an overcast day, use the Cloudy Bright setting. If shooting an indoor scene under fluorescent lights, use the Fluorescent setting. If you have access to a photo editing program that allows white balance adjustment, don’t worry about what preset you choose, because you can correct all the shots by the same amount if the preset you chose results in wonky colour.
• Overlap the Photos  Panoramic software uses artificial intelligence to look for similarities between adjacent shots then join those features seamlessly. In order to provide those reference points, you need to provide overlap in content between adjacent shots. I typically overlap by 25%. If I start my panoramic on the left end then take subsequent shots by panning left to right, this means that my second photo contains the rightmost 25% of the first photo. I’ve made marks on my tripod head that tell me how far to pan before taking the next shot while allowing 25% overlap. I have marks for both landscape format and portrait formats. But, of course, those marks only work for one focal length, which brings me to the next point….
• Use a “Normal” focal length  For a consumer digital SLR with an APS-C sensor, normal focal length is about 33mm. Anything shorter is considered wide angle. The wider you go, the more barrel or pincushion distortion will be introduced. This may make it harder for the stitching software to match the photos without twisting the photos to counter this effect. The more twisting, the more cropping of the final panoramic you’re likely to have to do. Of course, you could use a longer focal length without this concern, but remember that depth of field reduces with focal length.  I use a 35mm prime lens for panoramics, adapted from an old film camera.
• Choose the Right Content  I like shooting sky panoramics. I found out the hard way that sometimes the stitching software will not work properly with some content – particularly soft images with few hard edges. Some sunset clouds fit that description. Sometimes, it helps to include some hard edge detail (like treetops) to assist in the stitching. You can always crop it out later.

• Choose the Right Software  I find that the ‘premier’ photo editing package gives too many choices when creating panoramics. You have to try them all to find out which one works best, but my experience is that I generally don’t like any of the results. On the other hand, a cheap and cheerful package sometimes works better and with less fuss. I particularly like ArcSoft’s Panorama Maker, which came bundled free with my point-and-shoot camera.

 

 

3 overlapping images stitched together, before cropping. Note how the software has to distort the images to make them match up.



As you can see, the process of shooting panoramics like a pro is very manual. If you’re shooting a scene with moving clouds, set up marks on your tripod so you can rapidly make all your shots before the clouds change position and shape. And finally, resist the temptation to use more than 3 or 4 shots in a panoramic, because you’ll end up with a long, skinny picture sitting in the middle of your print that will look silly.

Meter this.

We’ve all seen the portrait photographer holding a white-domed object next to the subject’s face to take a light reading. So why is that being done? Why not just let the camera do its automatic exposure thing and take the picture?



 

It’s a good question in this digital age, where we can instantly review our shots to see if we need to re-shoot with a little exposure compensation. After all, the histogram in the camera tells us just about everything we need to know about the exposure in the shot – is it under- or over-exposed and is the exposure evenly distributed over the entire tonal range?

The short answer is that the handheld meter is mostly used to measure the light falling on the subject from studio strobes or flashes mounted off-camera.  In this scenario, the camera doesn’t typically have the ability to control the flash output, as it does in ETTL mode when the flash is mounted on the camera’s hot shoe.  Hence, studio photographers usually work with the camera in manual exposure mode, where they manually set the shutter speed to a value that will synchronize with the flash and the aperture to a value that gives the correct exposure. Alternatively, they can use the  meter reading to set the brightness of each flash in a multi-light setup to give a particular effect while ensuring a proper exposure.

Here, I’ve used the words ‘light falling on the subject’, meaning incident light. Hence, the incident light handheld meter measures the ambient light and ignores how much light the subject is likely to reflect.  The metering system in our camera, on the other hand, does rely on the light reflected from the subject to determine exposure.

Which is more accurate? The incident meter is, because it is calibrated so that it will give camera settings to make an 18% grey card look like 18% grey (middle mark on the histogram) when photographed.  The reflected meter, however, relies on the subject's tone averaging out to 18% grey in order for the exposure to be accurate. If the subject is overly dark or bright (such as a snow scene), the reflected meter gets fooled because it tells the camera to expose as if the subject is mid-grey. The incident meter would read the light falling on the snow rather than the light reflected from it, thereby resulting in the snow looking like bright white snow rather than grey slush.

I started out stating that the incident meter is used in the studio to measure the light from strobes or flashes, but obviously it has a function in outdoor ambient (continuous) light photography as well. Some photographers work in manual exposure mode all the time and therefore rely on an incident meter for camera settings.

So why bother? Part of the answer may be a throwback to the film days, where you never knew until it was developed whether you got the exposure right or not. An incident meter (and a little exposure  bracketing) gave you the confidence that you got the shot. That way of working may still influence the photographer who has made the transition to digital, partly because it’s good practice.

The bottom line, though, is that anyone seriously considering turning pro using studio lighting setups should consider working with a combination flash/continuous incident meter, if for no other reason than to save time. The alternative is to shoot, tweak, shoot and then tweak some more.

Combination meters can be a bit of an investment, so they’re not for everyone.  If you’ve got the spare change, though, a meter can help you understand the exposure mystery a little better and hit the bullseye every time.