Time lapse settings
Discussion
Hi all, I have just ordered a dedicated time lapse camera (Binno TLC200) and plan to set it up to record the construction of our extension. When it arrives i'll have to figure out where to mount it, I want it to be somewhere up high and stable with a good view of the hole area. but all i have is a few small trees that are sprouting leaves rapidly, i think I will end up fixing a tall post to the shed and putting it on that.
But more importantly I wondered what would be the best settings to use to record the building works? I originally planned to take a picture a day manually and stitch them together, but since finding this camera it provides far more flexibility - and more importantly eliminates the highly likely risk of me forgetting!
If i were to set this camera to take a photo a day I imagine I will ended up with a very short jumpy video. but on the other hand would a picture an hour during the working day be too much (the camera can be set to only record between certain times or only in daylight - so I wont need to edit out nighttime or weekends)
Just wondered what your opinions are and if any of you have experience in time lapse and the results of different settings/frequencies
Thanks
But more importantly I wondered what would be the best settings to use to record the building works? I originally planned to take a picture a day manually and stitch them together, but since finding this camera it provides far more flexibility - and more importantly eliminates the highly likely risk of me forgetting!
If i were to set this camera to take a photo a day I imagine I will ended up with a very short jumpy video. but on the other hand would a picture an hour during the working day be too much (the camera can be set to only record between certain times or only in daylight - so I wont need to edit out nighttime or weekends)
Just wondered what your opinions are and if any of you have experience in time lapse and the results of different settings/frequencies
Thanks
I can help - I've shot a lot of time lapses, as I work as a cameraman / DOP.
Theres a lot of handy timelapse calculators online, but it helps to have an understanding of the maths involved, to help you plan out your shot. Bare with me, I promise not to go too nerdy.
Generally speaking, video in the UK, is played back at 25 frames per second.
So 1 frame a day, played back at 25fps, would, over 25 days, produce a film 1 second long. A little quick. Time between the photos is referred to as the interval. In that example just there, the interval is a day, which obviously is way too long.
So when planning a timelapse shot, theres 4 steps I tend to follow
1) How long will the event be? I always start with an estimate of how long the event will be. For example - if I was getting a timelapse of a studio set build, for example. I'd ask the set designer, how long do you think it'l take to build- 8 hours they might tell me. I tend to add on more time, in case things over-run. Things always overrun. Lets assume, for this example, that the set build will last 10 hours.
2) How long do I want my final video to be? This is just personal preference, and depends what you are using the video for. Are you making just one video clip? Or will it be used as part of an edit? Lets say for this example, I want one stand alone video clip, I might want the final shot to be 30 seconds long. That seems watchable without getting too boring.
3) Now how many photos would I need to take to make a 30 second video?
For a video running at 25 frames per second, a 30 second video would need 750 photos. (25 x 30)
4) And Finally - How long an interval would I need?
So I need to take 750 over those 10 hours of the set build.
10 hours = 36000 seconds. 36000 seconds divided by 750 photos = 48 seconds.
So if I take a photo every 48 seconds, over 10 hours I will get 750 photos which will give me a 30 sec clip. Simples!
In reality, I usually do much shorter intervals. Its very easy for software to speed up footage. (if speeding footage up by 200%, the software basically just cuts out every other frame.) - but you cant slow footage down if you haven't got the photos there.
So I tend to get way longer timelipse clips than I need. - That means I can speed it up if I need too, but also means I can have slower sections of timelapse where there might be interesting things happening. Also - super fast time lapses are a bit horrid to watch, as things can happen too quickly.
So I'd work out how long the construction will take, and go from there. I'd suggest an interval of 1 frame every 30 seconds, to give yourself nice long time lapses from every day - then you can cut out the dull parts where nothing happens, and have a nice interesting video of the construction.
For doing a long form timelapse as you'll be doing, make sure it has mains power, backup the footage regularly, and reschule a regular check of the camera to check its running, and that the lens is clean. - you don't want a week to go by and find that a spider has made a web across the lens.
Another thing to be aware of - I'm not sure how your timelapse camera works, but generally a good rule with lapses, if to have a longer shutter speed - so moving things have a bit of motion blur to them in each frame. This makes the video much more fluid and smooth to watch. Often cameras will require ND filters to achieve longer shutter speeds during the daytime. Also, if you are building a mount for the timelapse camera, make sure its very solid - any wobble from wind will be really noticeable in the final video.
hope some of that helps!
Edited to add - here's an example - https://vimeo.com/211530541
This was a timelapse of a little live shoot I worked on. It shows us going from empty studio, loading in equipment, doing the shoot and derigging. The video was just for future use on the companies showreel, and as a clip they might want to throw up on social media as a behind the scenes clip of the show. I had a small mirrorless camera with a timelapse controller, mounting above the studio door. I think it shot 1 frame every 5 seconds. (When theres people, people move fairly quickly, so with people I tend to go for much shorter intervals.). The camera was set to a shutter speed of 2 seconds. I then edited down the final video a lot in places, cutting out repetitive or boring bits, and hiding the edits with quick cross fades.
While shooting, I also staged a few bits, like the lights going off and on in sequence, as I thought it would look more interesting than just all lights off / on at once.
Theres a lot of handy timelapse calculators online, but it helps to have an understanding of the maths involved, to help you plan out your shot. Bare with me, I promise not to go too nerdy.
Generally speaking, video in the UK, is played back at 25 frames per second.
So 1 frame a day, played back at 25fps, would, over 25 days, produce a film 1 second long. A little quick. Time between the photos is referred to as the interval. In that example just there, the interval is a day, which obviously is way too long.
So when planning a timelapse shot, theres 4 steps I tend to follow
1) How long will the event be? I always start with an estimate of how long the event will be. For example - if I was getting a timelapse of a studio set build, for example. I'd ask the set designer, how long do you think it'l take to build- 8 hours they might tell me. I tend to add on more time, in case things over-run. Things always overrun. Lets assume, for this example, that the set build will last 10 hours.
2) How long do I want my final video to be? This is just personal preference, and depends what you are using the video for. Are you making just one video clip? Or will it be used as part of an edit? Lets say for this example, I want one stand alone video clip, I might want the final shot to be 30 seconds long. That seems watchable without getting too boring.
3) Now how many photos would I need to take to make a 30 second video?
For a video running at 25 frames per second, a 30 second video would need 750 photos. (25 x 30)
4) And Finally - How long an interval would I need?
So I need to take 750 over those 10 hours of the set build.
10 hours = 36000 seconds. 36000 seconds divided by 750 photos = 48 seconds.
So if I take a photo every 48 seconds, over 10 hours I will get 750 photos which will give me a 30 sec clip. Simples!
In reality, I usually do much shorter intervals. Its very easy for software to speed up footage. (if speeding footage up by 200%, the software basically just cuts out every other frame.) - but you cant slow footage down if you haven't got the photos there.
So I tend to get way longer timelipse clips than I need. - That means I can speed it up if I need too, but also means I can have slower sections of timelapse where there might be interesting things happening. Also - super fast time lapses are a bit horrid to watch, as things can happen too quickly.
So I'd work out how long the construction will take, and go from there. I'd suggest an interval of 1 frame every 30 seconds, to give yourself nice long time lapses from every day - then you can cut out the dull parts where nothing happens, and have a nice interesting video of the construction.
For doing a long form timelapse as you'll be doing, make sure it has mains power, backup the footage regularly, and reschule a regular check of the camera to check its running, and that the lens is clean. - you don't want a week to go by and find that a spider has made a web across the lens.
Another thing to be aware of - I'm not sure how your timelapse camera works, but generally a good rule with lapses, if to have a longer shutter speed - so moving things have a bit of motion blur to them in each frame. This makes the video much more fluid and smooth to watch. Often cameras will require ND filters to achieve longer shutter speeds during the daytime. Also, if you are building a mount for the timelapse camera, make sure its very solid - any wobble from wind will be really noticeable in the final video.
hope some of that helps!
Edited to add - here's an example - https://vimeo.com/211530541
This was a timelapse of a little live shoot I worked on. It shows us going from empty studio, loading in equipment, doing the shoot and derigging. The video was just for future use on the companies showreel, and as a clip they might want to throw up on social media as a behind the scenes clip of the show. I had a small mirrorless camera with a timelapse controller, mounting above the studio door. I think it shot 1 frame every 5 seconds. (When theres people, people move fairly quickly, so with people I tend to go for much shorter intervals.). The camera was set to a shutter speed of 2 seconds. I then edited down the final video a lot in places, cutting out repetitive or boring bits, and hiding the edits with quick cross fades.
While shooting, I also staged a few bits, like the lights going off and on in sequence, as I thought it would look more interesting than just all lights off / on at once.
Edited by Fordo on Tuesday 4th April 19:15
Flipfloptrader said:
Excellent reply!
I'm very much a beginner and apart from shooting the odd picture on my phone, I don't have a lot of experience.
How does a GoPro that can shoot in say 60fps affect all of the above? Do you need more shots?
Yes - as esstenailly you're filming more than double the amount of frames a second.I'm very much a beginner and apart from shooting the odd picture on my phone, I don't have a lot of experience.
How does a GoPro that can shoot in say 60fps affect all of the above? Do you need more shots?
If you're filming at 60fps on the gopro, and want to edit the project at 60fps, and are cutting a timelapse in with that footage, then your timelapse will need to be played back at 60fps too.
So the maths would be a little different. -
1) How long will the event be? This hasn't changed - 10 hours
2) How long do I want my final video to be? This hasn't changed - id still like it to be 30 seconds when played back.
3) Now how many photos would I need to take to make a 30 second video?
This is the part that changes the formula - we want a 30 second video played at 60 frames per second.
At 60 frames per second, a 30 second video would need 1500 photos. (25 x 60)
4) And Finally - How long an interval would I need?
More frames means the interval will need to be shorter
I now need to take 1500 photos over those 10 hours of the set build.
10 hours = 36000 seconds. 36000 seconds divided by 1500 photos = 24 seconds.
So if I take a photo every 24 seconds, over 10 hours I will get 1500 photos which will give me a 30 sec clip when played back at 60fps.
I almost never use 60 fps. Uk TV plays back at 25fps, and film in the cinema is 24fps. They have a certain visual cadence to them. - while 60fps is smoother when the camera is waved around, people tend to associate that look with TV sitcoms (as that kind of TV used interlaced, which essentially blends frames and makes motion seem smoother.).
But that said - 60fps might be better for personal videos, and you can wave the camera around and it won't look jerky. At 25fps / 24fps, its at the lower end of what works to fool our brains into a moving image, and theres certain camera moves that can look jerky. At film school, your taught to not pan the camera at certain speed, to avoid this.
It's quote odd really - in theory, faster frames per second should technically always be better to the eye. But a lot of people react negatively to films shot at a higher fps. The hobbit was shot at 48fps - people said it seemed fake and TV like. Theres a lot of phycology in fooling our brain. Perhaps we just associate the 24 / 25 fps 'look' with film, so anything that differs feels 'cheap' to us. Or perhaps faster frames per second makes it 'too real' and therefore makes us uncomfortable - the uncanny valley effect'. Anyway, I digress a bit!
One thing to remember about filming things at 60fps - It'll use more double the amount of data that 25fps will. You're capturing more than twice the amount of frames per second - so more images, which take up more space.
You can also use higher frame per second filming for slow motion. If I'm editing a project at 25fps, and i film at 60fps on a gopro, or any other camera, if I put that 60fps footage on a 25fps timeline, the clip will playback slower - essentially more than half speed slow motion.
Essentially that whats going on for slow motion in any form - a camera will be capturing a higher frame per second - 50fps, 60fps, 100fps or even 1000fps - and then they are just played back at a slower fps - say 25.
Sorry, lots of info there, I could talk about this stuff for days
Edited by Fordo on Wednesday 5th April 09:27
Thank you for your very informative post Fordo, it’s all very helpful and interesting. One of the reasons PH is great as there are experts on virtually anything hanging around its various forum subsections.
For the duration of the project, the builder has said that it will take till the end of the year, he will be working with just him and a labourer for most of it so it will be a little slower than if he had a large workforce blitzing the work. (works for us too as there will be slightly longer periods between paying for each stage)
Obviously, there will be a point where all the external works is complete so there will be nothing to see from the outside.
As for the video length, I have no idea! Long enough that it shows the building going up nicely, but not long enough to bore people to sleep! I think your comment on more photos is better than too few as it can be sped up, very good advice there. The motion blur shown on your example video is a good effect too, tracking shows that I should receive the camera tomorrow so I will be able to look at the specifics available to me when I set it up.
The camera is self contained and has a claimed battery life of 120days – though I will probably be checking it weekly to make sure it is ok and to review and save the time lapse from that week. And as you say I need to fabricate a sturdy mounting system for it, that will be this weekends project.
For the duration of the project, the builder has said that it will take till the end of the year, he will be working with just him and a labourer for most of it so it will be a little slower than if he had a large workforce blitzing the work. (works for us too as there will be slightly longer periods between paying for each stage)
Obviously, there will be a point where all the external works is complete so there will be nothing to see from the outside.
As for the video length, I have no idea! Long enough that it shows the building going up nicely, but not long enough to bore people to sleep! I think your comment on more photos is better than too few as it can be sped up, very good advice there. The motion blur shown on your example video is a good effect too, tracking shows that I should receive the camera tomorrow so I will be able to look at the specifics available to me when I set it up.
The camera is self contained and has a claimed battery life of 120days – though I will probably be checking it weekly to make sure it is ok and to review and save the time lapse from that week. And as you say I need to fabricate a sturdy mounting system for it, that will be this weekends project.
Well, I ended up doing the obvious thing and have downloaded a PDF of the cameras instruction manual. (Clicky)
here is a screenshot of the settings and available options of the camera:
It says that for a construction project they recommend a time interval of "30 - 60 mins" and a frame rate of "15 - 30 fps"
So I have never been great at Maths but if I assume all the external work is complete by the end of October, that's 29 weeks which is approximately 145 working days.If I record between 8:00 and 18:00 that's 10 hours per day. So with the above i'd use:
10hrs / Time Interval = Frames per day
Frames per day * 145 = Frames for the whole project
Frames for the whole project / FPS = Video length
Therefore if the intervals are set at half hour, I capture 20 images a day. If my video playback is set to 20fps my completed video of the whole construction would last 2mins25.
here is a screenshot of the settings and available options of the camera:
It says that for a construction project they recommend a time interval of "30 - 60 mins" and a frame rate of "15 - 30 fps"
So I have never been great at Maths but if I assume all the external work is complete by the end of October, that's 29 weeks which is approximately 145 working days.If I record between 8:00 and 18:00 that's 10 hours per day. So with the above i'd use:
10hrs / Time Interval = Frames per day
Frames per day * 145 = Frames for the whole project
Frames for the whole project / FPS = Video length
Therefore if the intervals are set at half hour, I capture 20 images a day. If my video playback is set to 20fps my completed video of the whole construction would last 2mins25.
Gassing Station | Photography & Video | Top of Page | What's New | My Stuff