Sometimes when I post a timelapse shot I’m asked “why not just speed up the video?”. I see that question so often I decided to write an article on that topic – so I can just paste them a link. Writing the same answer for a dozen times can be frustrating 😉
There are pros and cons of both techniques, but in most cases shooting a real timelapse with still photographs would be a better choice. Much better 😉
Interested in Timelapse, Hyperlapse or Stop-Motion? Feel free to join the Facebook Group, where you can find BTS content of my tutorials: https://www.facebook.com/groups/timelapse.hyperlapse/
First and main thing here is the quality. Let’s assume that you’re editing the video in full HD resolution, which is aboout 2 Megapixels. DSLR/DSLM cameras shoot usually at least 12 mpix stills (and more often 18-24mpix). Downscaling that big image even to 4k video (~8mpix) will increase the amount of details in your shots and let you use the full potential of your camera 😉
Speaking of resolution – having bigger image than your final video allows you also to reframe or zoom your shot without quality loss. It can be useful for some heavy stabilization or more complex transition, like in my Panta Rhei video.
The last thing related to quality is the ability to shoot RAW. There are tons of advantages of shooting RAW. I know that sometimes people argue whether shoot JPG or RAW, however if you want to shoot timelapse professionally, you’ll shoot RAW 😉 So, one of the advantages of shooting RAW is using the full dynamic range that your camera provides. Of course in video you can get close to that shooting some very flat LOG profile, but now every camera has that option and it’s more complicated.
Long exposure shots
For me one of the biggest advantages of shooting ‘real timelapses’ is that I can stick with the 180 degrees rule with my shutter speed. If you’re not familiar with that rule, check out this article: Cinematic Motion Blur – 180° Rule . The basic idea is to blur things that are moving in the frame (people, cars, water etc.). Adding a motion blur to the timelapse really makes the difference. It’s so important that I wrote specific, in-depth tutorial on this subject – Should you use ND filters? Motion Blur in Timelapse. The other important thing here is that long exposure = better low light 😉 You don’t have to raise up the ISO that much, because you can set longer shutter speed.
The amount of data
The amount of collected data can be important for some people. I think it’s more a beginner’s problem, because experienced timelapsers are used to store hundreds or thousands of GB 😉 It’s a little bit tricky topic here – it depends on the interval (the speed of the final video) and whether you shoot RAW or JPG. To simplify, I’ll talk just about RAW data and I’ll compare that to 100Mbps video (the bitrate of GH4).
Speeding up video is basically deleting frames. For example if you speed up your video 200%, the software deletes every second frame (or in different words it shows only every second frame). So, making a timelapse from video is just collecting the big amount of data, that you don’t need. The graphs above shows the amount of data that is collected to get a final clip with specific duration. The horizontal axis shows the duration of the final video, the vertical axis is the amount of the data collected. For those who are not very good at math – more to the left – longer final video and higher the line goes, more data is collected.
I’ve already mentioned that timelapse from video is just deleting frames. So, faster video equals more data to delete. As you can see in the graphs, the amount of data collected in video depends on the desired interval (in speeding up video you don’t set an interval, but it refers to a speed of the final clip). From this graph you should learn one thing – for longer shoots it’s definitely better to shoot stills 😉 Not only you get better quality, but also less data.
I believe that there is no perfect solution, no matter what it is. Everything has some pros and cons, so I feel obliged to mention also when it can be better to shoot video 😉
First of all, beginners often ask which interval should they choose for certain shot. There isn’t one correct answer for that, such knowledge comes with experience. You’ve got to know how the final video would look like and adjust the interval to the specific effect you want to achieve. In video, you choose the speed of the clip in post production and you can modify that as much as you like.
Collecting additional data also allows you to use the frame blending option to simulate motion blur within the frame. For me the blur isn’t that pleasing as shot in camera, but it should be mentioned here ;)Last but not least – you can set ‘intervals’ shorter than 1 sec, which is the usual intervalometer limit. For me it is the real advantage that I can use it to get a desired effect in my videos. Sometimes 1 sec is too much 😉
Low FPS video
Some cameras have ability to shoot video with lower framerate than 24 frames per second. In Canon cameras there is a FPS override option in Magic Lantern hack, where you can go up to 1 fps. Panasonic GH4 has Variable Frame Rate option (the same you use for slow-mo, goes up to 2fps). Sony a6500 has that option too. Even your camera, like GH4 or a6500 shoots 4k, the low fps video would be ‘just’ Full HD. I’m really happy that in GH5 producers add that function for 4k video, too. I would test that for you if I have a chance to get my hands on that camera 😉
Again, someone can ask “Why? Can’t I just speed up the regular 25/30 fps video?”. You won’t get better quality here, however you can get a beautiful motion blur. Because it’s a video mode, you can choose even 360 shutter angle (shutter angle explanation article), which is not possible with still phtographs (the camera must have some time to write the picture to the card). For example, if you shoot 2 fps video, you can choose up to 0.5″ shutter speed. In normal video it’s up to 1/24 (at 24 fps). This is very nice feature which I use occasionally with quite good results 🙂 Especially I like to use that with handheld hyperlapses with my gimbal. You can read about that technique here: Handheld Hyperlapse with a Gimbal [VIDEO TUTORIAL].
So, here’s the summary of all things that I think are worth mentioning in this article. Do you have some other pros and cons on both techniques? Share that in the comment or in my Facebook group: https://www.facebook.com/groups/timelapse.hyperlapse/
Timelapse with still photographs:
- Higher resolution – better quality and ability to reframe the shot in post
- RAW data – more flexible in post and in most cases wider dynamic range
- Long exposure shooting – ability to add much more motion blur & low light shooting.
- Less data (for longer interval shoots) – you’re taking only the frames you need
Speeded up video:
- Choosing the speed (and duration) if the clip in post – useful if you can’t predict the duration of the shoot
- Intervals can be shorter than 1 sec
- Less data (for shorter shoots)
Thanks for staying with me for another article 😉 Now, if someone asks you why you’re shooting stills, you can just paste them a link to this post. I hope it would let people understand, that timelapse is a great and very powerfull technique, it’s not just a speeded up video 😉 To get notification about new tutorials on my blog, click the little red bell in the bottom left corner. You won’t regret it!
My name is Jason Mills. I have a couple of questions on time lapses. I plan on shooting the solar eclipse with the gopro hero 4 silver. What do you think the best settings are for this event regarding the following:
1) Shutter rate auto (for when it goes light to dark) or selected
3) White balence auto or selected
4) Shooting color vs flat
5) Iso min & max
Thank you so much for your posts online! They help a lot. ?
I’m not experienced with go pro, so I will answer what I can:
Of course it would be best to shoot it in RAW, especially because of the exposure changes. In a dslr I would use holy grail technique, so I would manually change the shutter/aperture. Not sure if it’s possible with gopro, maybe auto shutter would be a good idea. Definitely the exposure settings must be changed if you want to get when everything goes dark.
As I checked, solar eclipse duration is about 3h, but the total eclipse is just for two minutes. If you’ve got enough space on the memory card I would go with an interval o 5-10 sec and then speed up in post everything before and after total eclipse. Of course it gives you lots of pictures, but you’ve got to decide how you want the final clip to looks like 😉
I wouldn’t use auto white balance shooting jpg. It can ruin the shot.
For jpg probably better to shoot flat.
if you use a tripod it would be better to change just the shutter, not the iso.
Good lick in shooting 😉 In my region the eclipse is not visible so I’m waiting to watch some nice timelapses for you guys 😉
Are you sure that time-remaping deletes frames ?
I usually slow down video so that 3 hours takes just few seconds. I wonder how would “timelapse” of 1s interval looks like