This tutorial describes how to make star trail photos that can be used in time-lapse videos. I go through the process of making a Photoshop action to batch edit star photos and offer a link to those who don’t want to make their own action.

Links mentioned in this video:

I’ve decided to start posting some video tutorials in addition to my written posts because I usually look for videos when I’m looking for help online. I’m going to gradually update some of my more popular posts with videos to augment the existing material. You can subscribe to my new Youtube channel to avoid missing any updates or new tutorials. I made this video in response to a question about the star trails Photoshop action I made and uploaded a couple of years ago. If there are any other topics you’d like me to cover, please send me an email through my contact page.

I’m also looking for ideas on how best to use these videos. Do many people read these posts at work or someplace else where they can’t watch a video? If so, I’ll work to add text tutorials below the videos.

If you use this action, please post a link to your video below. It takes about a day for me to approve each one because of the amount of spam comments I get. Thanks for watching and please let me know if you have any suggestions.

My Camera

Canon EOS 60D
The 60D is an odd fit between the amateur 50D and the professional 7D. I’ve shot with the 7D and would have a hard time finding much of a difference. The controls aren’t quite as intuitive and the 60D isn’t quite as tough. However, the 60D offers a whole lot of camera for around $1000 (body only).

The 60D also has a few features that separate it from the lower models. For example, the 60D shoots full 1080p video, although I suspect even the Rebel line will do this soon. The best reason I can suggest a 60D is because this beautiful time-lapse by Randy Halverson was largely shot with one.

My Computer

Lenovo K330 77275DU Desktop
After years of faithful service, my Toshiba laptop began struggling under the weight of 1080p 30fps output from RAW files.  So far, this new computer has been great.  It has an Intel i7 processor, 12GB of DDR3 memory, 1.5TB of HDD space (which I’ve upped to 3Tb), and an ATI Radeon 6450 DX11 1GB.  This should sound a little impressive if you’re reading this in January 2012, or horribly, horribly outdated if you’re reading it in February 2012 or later.  I threw on a 24″ full HD LED monitor for a reasonable price that has more color shift than a true video or photo editor would accept, but it was half the price of better units.

I realize I’ll have to replace the video card at some point with a better model, but I wanted to make sure I had something with enough guts to upgrade it in the future.  I haven’t done much serious work with this computer so far, but I’ll try to give an update in a few months.

My Tripod

Targus Digital TG-6660TR
I bought this at Radioshack on sale for $20 a few years ago. It’s currently listed on Amazon for $30. I’m sure there are plenty of reasons why an expensive tripod is far superior, but I haven’t had any problems with this baby so far.

My Lens

Canon EF-S 18-55mm
This is the only lens I use for time-lapse. I’d like to buy something a little wider, but this lens has worked well for most landscape shots. The only reason I’d like a wider lens is to capture city scenes close up. I’ll probably rent a lens from borrowlenses.com and try it out before buying.

I bought this lens used for $100. I’d pay the extra money and buy a new one next time. I got lucky and was able to clean the dust from this lens but it’s possible it could have had a minor flaw I didn’t notice until after I had taken thousands of photographs.

My Filters

Zeikos Circular Polarizer Filter
$5.86 with free shipping. What else can I say? If you’re going to use one all the time and have the money, invest in something better. If you want to play with a polarizing filter, the entry level to acceptable filters has dropped to the price of a couple lattes. Make sure to buy the one that fits your lenses.

Polaroid Variable Range Neutral Density Filter
I bought this a few months ago and still haven’t used it. A variable ND filter lets you block as much light as you’d like. The advantage of this for time-lapse is that you can take four second exposures in broad daylight to avoid problems like birds flying through your shot and to give blurry effects to people on the street. You can also use one of these to increase the f-stop of portraits to blur the background, which is a popular effect in photography these days.

My Intervalometer

RS-60E3 Remote Timer
Just a few years ago a timer like this would have cost over $100. I bought this one for $16 and it works just as well. Although the Canon 60D isn’t listed as compatible, it definitely works. You may need a different type of intervalometer if you own a different camera.

My Memory Card

Centon 32GB Class 10 SDHC
I don’t usually buy off brand memory cards, but this one was on sale at woot.com. Unfortunately, the sale is over so you’ll have to check back occasionally or buy one somewhere else. I used this one for the first time in December to capture the total lunar eclipse. It worked without a problem and was just big enough to capture the 30GB of RAW photos I took that night. I’ve also used it for HD video but had some trouble. It would occasionally stop recording after the first two seconds. I’ll buy a better, faster card soon and use this as a backup.

Odds and Ends

That’s it for my main equipment. I still use my Canon Powershot S3 IS, but it’s seen better days. I also have a telescope tripod with an equatorial mount that’s missing a motor but would be great for night time videos. Then there’s the random 2GB memory card floating around.

My major planned purchases include some sort of dolly system and maybe material for a crane. I have some motors and plan to make a cheap moving platform once I figure out a good way to control the motor speed. I’ve tried a 555 chip without luck and might just buy an Arduino microcontroller. Please let me know if you’ve had success with either method.

I began taking these photos a couple of weeks before Christmas but just finished the video this weekend. The workload finally became too much for my laptop so I had to upgrade to a new computer.

An interesting tidbit about these videos is they were all made without an intervalometer. Because they were night shots, I simply set my camera to low speed continuous mode and locked down the shutter. The average exposure is around five seconds. Here’s a brief description of the shots in order of appearance:

City Hall – They sometimes change the colors to fit the occasion in San Francisco. During the baseball playoffs last year, city hall stayed orange at night until after the Giants won the World Series. The people posing along the street are on their way to The Nutcracker at the War Memorial Opera House across from city hall, hence the music.

Cable Cars – Some of these are decorated with Christmas lights, which gave a cool effect when they were moving. I also like the large amount of holiday shoppers milling about.

Marina Boats – A couple boats had lights but I couldn’t get them to “pop” in the final edit. The Palace of Fine Art is in the background but barely visible. You can also see my shadow on the boat closest to the camera.

Fisherman’s Wharf Boats – These were more heavily decorated than the boats at the marina. The video by itself seems really frenetic, which is another reason I picked such an up tempo piece of music.

Skyline with Coit Tower – The only seasonal light in this shot is the bluish white one at the top of the Transamerica building. Still, it might be my favorite shot because of the different lights turning on and off in the houses and the Bank of America building.

Skyline from Pier 7 – I’ve been meaning to shoot from Pier 7 for some time, but this was my first time actually doing it. I would have used a wider lens if I had one, but I made sure to capture the seasonal outline lights on the Embarcadero Center buildings. While I was recording, I heard a burst of air and looked down to see a sea lion swim right below me. Somehow I forgot there might be condensation by the water and therefore didn’t bring a heater. There were no stores open nearby so I couldn’t buy a hand warmer and a rubber band for the lens. That’s why this shot slowly hazes over. I’ll make sure to try this shot again when I have the right equipment.

That’s it for now. I’m just glad I was able to post this video while it’s still winter.

I’ll keep this page updated as BorrowLenses.com comes out with new promotions and discount codes.

The current special is for 15% off new orders made by Jan. 31, 2012.

Click Here and Enter Code winter15 at Checkout

BorrowLenses.com ships to all fifty states and Puerto Rico. They also have several pick up locations in California and one in Nevada.

To get the discount, click the link above and enter the code into the Promotional Codes box at checkout. Renting through this link supports my website. I’ve used BorrowLenses.com in the past and had a good experience with them, so I don’t mind recommending their services.

Deshaker is a filter for the video editing program VirtualDub that reduces unwanted motion in your videos.  It can be used in any video with distracting camera movement but is especially well suited for time-lapse because the large photos mean you can crop your subject without reducing the video to less than full HD.  However, the following instructions work just as well for any other type of video. If you’ve never used either program, you might want to take a look at my VirtualDub for time-lapse tutorial before you begin.  This tutorial assumes you’ve already processed your photos into a video.  My only suggestion is to make the video slightly bigger than your imagined final version to allow room for cropping.  If you’re using a video shot live, this isn’t an option but the software can still help.  It just means your final video will be a bit smaller.

The Software

The newest Deshaker software might require a VirtualDub update.  Both are completely free to use for Windows systems.  Deshaker only works with 32-bit PCs for the moment.

VirtualDub software link

Deshaker software link

Extract and Run the Filter

Unzip the file and copy the Deshaker.vdf file into the VirtualDub plugin folder.  VirtualDub doesn’t have an installer, so you might have to search for the folder if it’s been a while since you downloaded it. The plugin folder should appear as soon as you open the main VirtualDub folder. Once the filter’s in the right place, start VirtualDub and your filter should be available.

Open the file you want to deshake and right click the videos to make both the input and output fit in the screen.  Click Video>Filters and then click “Add.”  The Deshaker filter should be visible in the top half of the list.

Click “OK” and you’ll come to the window shown below.  Deshaker takes two passes, so we’ll begin by accepting the default settings on the first pass.  You can change this later if you have trouble or select “rolling shutter” if you took the video with certain video cameras.  For time-lapse videos made from photos, this option is unnecessary.  Click “OK” to return to the main window.

At the bottom left of the screen are three tiny buttons:  Stop, Play input, and Play output.  Make sure the video is at the first frame and hit Play output.  This is the play button with the almost invisible “O” under it.  The filter should start comparing frames and deciding which way the camera was moving and how to correct it.  You should see an output screen covered with white and red dots that shift along with the subject.

Yawning Lemur

 Compress and Run the Second Pass

Select a compression codec even if the video was originally compressed.  After running your video through this filter, the video file will be far bigger than before.  I use the Xvid codecalready on my computer.  Select Video>Compression and pick your favorite. Once the first pass finishes, go back to Video>Filters and double click on the already selected Deshaker filter.  In the upper left hand corner, click on “Pass 2.”  From here you have two main choices.  You can choose the default mode, which will keep most of your video intact but give you moving black bars around the edges, or you can select one of the “no borders” options.  I recommend choosing one of the “no borders” choices if you want a polished video.

The downside is the video has to be cropped in order to keep the moving edges from showing.  This is why this filter is ideal for time-lapse, you can make your original video larger than HDTV’s 1920×1080 so you probably won’t have to compromise on the final quality.  By contrast, a video originally shot for HDTV has to be cropped below the true 1080 level.  If this is confusing, save a copy of both so you can compare.

Save and Refine if Needed

The second pass will run without rewinding the video or playing it a second time.  You can save it soon after you click “OK” in the Deshaker filter.  You can also tweak how the second pass looks without running the first pass again if you’re not happy. I’m almost always satisfied with the default settings, but I should mention that some subjects are more difficult for the filter than others.  For example, when I made my walking time-lapse video, I used the Deshaker filter extensively to smooth the very rough cuts between photos.  However, I found that the filter was less effective with fewer photos and useless against the spinning fountain shot.  For the latter, I had to manually line up the shots in Photoshop.

Wrap Up

I might make a video tutorial for this if I ever get over disliking the sound of my recorded voice.  If you think one would be helpful, please let me know in the comments or send me an email through my contact form.

I shot this video from San Francisco on Saturday morning. This is the last total lunar eclipse until 2014. The moon started going dark around 4:30am and stayed in the Earth’s shadow until just before it hit the horizon on the west coast. A brief outline of my workflow is below.

Photos – Taken with a Canon 60D with a Canon EFS 18-55mm lens set to 18mm. I used Av mode because the light changed from full moon to eclipse to daybreak. I used manual focus and disabled the image stabilizer. I had to use ISO 800 to keep the exposures under six seconds in order to keep the motion smooth without blurring the moon or stars.

Processing – All photos were shot in RAW, which meant 1300 photos took around fourteen hours to process. I chose a temperature that kept the moon white in the beginning but let it transition to red at the height of the eclipse. I used the Digital Photo Professional software shipped with the Canon 60D to convert the RAW files to jpegs. The size of the original RAW photos were around 30GB.

Video Processing – I used VirtualDub with MSU Deflicker to correct for minor light changes caused by shooting in Av mode. Frames were compiled at 25fps and compressed with Xvid.

There’s a bit of noise in the video because of the high ISO, but my attempts to eliminate the noise with Photoshop lead to more banding than I’d like. Both versions were around the same level of quality, so I went with the less processed version. If you have any other questions, please let me know.

Click here to go to the Youtube video

I thought I’d get away from my usual time-lapse tutorials to post something I hope anyone interested in online video will appreciate. One of my non-time-lapse Youtube videos just hit a million views. I learned a few lessons along the way that have already increased my total view counts. At least a couple of the lessons should apply to your videos as well.

Youtube estimates ten hours of video are uploaded to its servers every minute. They also say the average video gets fifty views. How can a sixteen second video stand out from the deluge of mediocrity? It needs a good subject, a good photo, and a little help.

How it Began

In the summer of 2009, I found a really big black widow spider living just outside my front door. They’re native to central California where I was living, but this black widow was by far the biggest I’d ever seen. Instead of simply smashing it, I decided I wanted to take a few photos and maybe even make a video. The short video came out well, so I put it on Youtube in case anyone else thought it was as interesting as I did. The video sat there for about two years gathering a few thousand views but not really taking off until August of 2011. Here’s what happened:

Make a Good Video

The video I made was good. Not great, but good. More importantly, it was better than most black widow videos available. You can clearly see the red hourglass shape on her abdomen and her skin almost glistens.

It’s possible a better quality video will displace mine but quality alone doesn’t win on Youtube. I think some of the appeal is that the video was shot by some guy dumb enough to catch a black widow in a jar at his home. Professional videos remove some of the danger and act to distance the viewer from the situation. Some film makers have picked up on this and used the faux-documentary horror genre with great results in movies such as The Blair Witch Project, Paranormal Activity, and Cloverfield. Higher quality doesn’t always mean more popular.

Wow! Aww! Eww! – The Importance of Emotion

If I show someone the photo of the black widow and they say, “eww,” I know they’ll ask to see the video. This seems counter-intuitive but I know from experience it works.

Very few things make adults say, “wow,” but I believe trying to elicit that response should be one of the main goals of video makers. For the purposes of popularity, people also respond to gross and cute things; I’m not sure which one is more popular. The viral video of the baby panda sneezing was very popular, but movie studios bank on gross things to consistently bring people to the theaters. “Slither” would probably draw a more reliable audience than a movie named “Cuddly.”

It’s All About the Thumbnail

If you take one thing from this article, make sure it’s picking the right thumbnail to represent your video. I don’t think nearly as many people would watch my video if that perfect red hourglass wasn’t visible in the thumbnail. Youtube automatically selects a frame from your video to act as the photo people see when they search or when it appears on a list at the side of another video. This used to be taken from predictable spots but they’ve since changed to make it more random. The only thing you can do, short of a couple questionable techniques, is to make sure your video is high quality throughout or apply to Youtube’s partners program. If you’re unhappy with the default thumbnail, you can select one of two other thumbnails taken from your video by clicking “Edit info” and selecting “Video Thumbnail.”

I’m not positive how Youtube picks which videos to feature in the sidebar but I’d imagine a video that gets a high percentage of clicks will appear more often.  In essence, you’re competing with several other thumbnails for a user’s attention.  If you’re the second most interesting thumbnail, the user might never return to give your video a chance, even if it’s the best.

Picking the Right Title

I chose “Huge Black Widow Spider,” because I wanted to tell people why they should watch it (it’s huge) and what it is (a black widow). Then I added the word “spider” so it would appear in broader searches. Choosing a good title should be familiar to anyone who knows about search engine optimization, or SEO. There are thousands of articles on how to write good titles, so I’ll just talk about this one.

Would you rather see a “pretty big spider,” a “rather large spider,” or a “huge spider”? Use extremes whenever possible but keep it honest. I would love to call this video “World’s Largest Spider,” but that isn’t true. False claims may draw more views, but people will leave your video as soon as they figure out it’s a scam. Youtube doesn’t seem to favor videos with lots of “thumbs up” votes right now, but a change in their algorithm could easily shoot a misleading video to the bottom of the results in the future.

The other reason a misleading title can work against you is ads won’t appear for several seconds if you decide to monetize your video. The ads on my spider video don’t pop up for ten seconds, so a bad video with millions of views still won’t earn money.

Comments – Keep ‘Em Coming

I think I originally included a sentence or two under the video telling where I found the spider but not much else. As the video gained in popularity, people started leaving comments asking what I did with the spider. I responded to some of these questions, and still do, but I eventually included a line beneath the video about how I released the spider afterward. Without meaning to, I created controversy and split viewers into two camps–most chided me for not destroying it, but some thanked me. Despite how they felt, it encouraged viewers to take a few seconds and respond. If the viewer was respectful and asked another question, I took the time to respond. This often brought them back to the video and got others to join in the conversation.

The Tipping Point

If you’re searching for the best of anything online, you probably use a small number of websites that aggregate the best of the best and present it in an easy to browse format. These sites have more influence than most individuals in promoting something new and interesting. In my case, the maven who spread my video ran an Arabic language channel on Youtube. One of the advantages of a video without words is that it transcends languages and gets more play in non-English-speaking countries. I might try to push this with future videos by trying to hand deliver them to those with large circles of influence in whatever category was appropriate for the video.

I’ve seen this phenomenon happen in the time-lapse videos I watch. For example, after the last lunar eclipse, I searched and found a few excellent videos. Some, however, had fifty views while others had fifty thousand views within the first day. The only difference is that the popular videos were featured on a couple of high traffic sites, which led other sites to feature the video. At some point, the video reached a tipping point where it shot up exponentially. If you have a good video, try bringing it to the attention of someone with influence who can promote it to others. It takes some work and practice to approach strangers like this and get what you want.

Linking

Another good way to bring views is to hitch yourself to someone else’s wagon by linking your video as a response. You’ll have to be picky because you can only do this once for each video. I’d suggest finding a video similar to yours but not obviously better. You want people to be hungry for your content when they see it. Find a video that has a large number of views, a healthy percentage of “likes,” and a bunch of comments. Because your video will appear under the host video, you’ll need to make sure people care enoguh to scroll down. After you post your video as a response, the other video’s owner must accept or reject it. Don’t get discouraged if the first few people say “no.”

Monetization

At around 30K views, Youtube sent me an email asking if I’d like to put ads on this video. It was really easy and fun to see the money slowly trickle in.

We all know advertisements can be annoying, so I’d suggest waiting until you get several thousand hits before even considering ads. If a video is good enough or becomes popular, people will be more willing to accept ads. You can submit your profile for Youtube’s partners program or simply wait until they contact you. Google is good at making money and won’t let a video sit idle if they think it might be headed for a bunch of hits. Read this New York Times article for more information on how people make money on Youtube.

Conclusion

I didn’t set out to make a million hit video, but I’m giving serious thought on how to make another. If you have a video you think could make it big, share a link in the comments. Are there other “rules” I haven’t thought of?

Today’s post features a video of New York City shot by Benjamin JEHAN from France who goes by Titou on Vimeo. JEHAN is new to timelapse, but he did a good job with this video and is getting a good number of views. He agreed to share how he made this video with timelapseblog readers.

Tell me about this video. Why did you make it and what equipment did you use?

This is the first timelapse of a collection called LIFE, which consists of making videos of different places in the world to see life as it is….

I used a Sony Alpha 33 camera, with a 18-55mm lens, and a Vanguard MAK 203 tripod.

I made this video over about two days, but I prepared for it over ten days in NYC. I upped the colors a little and changed some others parameters in Lightroom. I used LRTimelapse for deflickering and then made sequences with 25 images per second with Quick Time Pro. Afterward, I used Adobe Premiere to make the film and After Effects to create titles.

Can you tell me about using LRTimelapse? Was it easy to use? How much does it cost?

LRTimelapse is a free application. I use it to deflicker my sequences. Because sometimes, when you shoot during the day, you can have some clouds and you have a flicker in the final sequence. You need to change little parameters to have smooth transitions between frames.

Why did you use 25 fps?

I used 25 fps because it’s the standard in France. I calculated during the shoot that 25 frames will correspond to one second of my final sequence. It’s very important when you shoot your timelapse to visualize in your mind, as best as possible, what you want and what the sequence will look like in the end.

How did you get the moving shots? Did you rotate the camera or do it with software?

Normally, professionals make moves with motion control rigs, but I don’t have the right equipment. I’m just beginning to make timelapse videos, so I just made the moves in post production with After Effects. My resolution was bigger than 1920×1080, so I could move my virtual camera within my scene.

Did you use any filters?

I didn’t use any filters for this timelapse. I just corrected my sequences with Lightroom to have more beautiful colors, but for my next timelapse I will use an ND filter… But I won’t talk more about it now. You will see it when I finish it! :-)

What settings did you use?

I only use manual mode. It depends on the sequence, but I also use manual white balance, with ISO 100, and f/22 to make a good image. My shots are every two seconds to have beautiful results with clouds, but with cars and people too.

What settings did you use for the final shot showing evening turning into night? Did you change the settings as the light changed?

No, this sequence was really difficult. I tried to change my settings but I didn’t have good results, so I made another sequence with ISO 200 and I just waited for night. You can try it and after 2-3 tests, you will be successful, I’m sure.

What would you do differently next time?

I will have a new camera. I don’t know yet if I will buy a Canon 60D or a Sony Alpha 65, and I will have a 50mm 1.4 lens. My next timelapse will be about another subject… I will make it in a beautiful city….with great color and harmony between history, nature and technology!

I compiled this video from publicly available NASA photos. This is similar to the webcam time-lapses I’ve made before, except much easier.

I tried to pump up the excitement by using classical music performed by the US Air Force (because it’s public domain) and by including a couple of wipe transitions (a la George Lucas).

I posted a video by Luke Shepard a few months ago featuring Paris at night. I’ve finally gotten around to posting a response to that video. This video was made with my old PowerShot so the quality isn’t great and I haven’t gotten down the technique well enough for a tutorial, but I think it offers a neat new way to make time-lapse videos.

I made the video by taking photographs on a tripod and focusing on a single point to keep it in the same place. I’d move the tripod a few inches, manually snap a photo, and then move it again. I then lined up the shots in VirtualDub using a free deshaker filter. I’ll post a deshaker filter tutorial sometime soon.

The first shot is coming out of a tunnel at the Conservatory of Flowers in Golden Gate Park. I tried to line up the finial at the top of the building as my focus point.

The next shot was made by walking away from the Francis Scott Key statue. I really like how this turned out. The clouds look good and the pan-out gives a sense of scale.

The fountain was the hardest to line up. VirtualDub couldn’t align the images because the background and lighting changed so drastically. I used Photoshop instead and lined up each photo with the one before it. Even this slow method didn’t give the result I wanted. I focused on lining up the lion’s head but the rest of the fountain wobbles a bit in the video.

Notre Dame? Oh yeah, we’ve got one of those. This shot was of Grace Cathedral on Nob Hill in San Francisco. There weren’t any clouds the day I went, so I cut this shot pretty short. If I had a wider lens, I would have included traffic and pedestrians to give a sense of time passing.

The final shot is of the Golden Gate Bridge. I used a polarizing filter, which really brought out the colors. It almost looks like HDR. I also like the lenticular clouds over the Marin Headlands.