How to shoot stop motion ?

Mohamed Osam · November 21, 2016

Hello ,

I would like to know how to shoot stop motion as it in this video technically .

Brian Drysdale · November 21, 2016

You basically shoot one frame at a time, moving the subject (live subject or a model or in a series of drawings) between shooting each frame, so as to create the impression of continuous motion. Some video cameras can do this or a stills camera is commonly used and then import the stills as an image sequence into a NLE,

https://en.wikipedia.org/wiki/Animation

https://en.wikipedia.org/wiki/Stop_motion

In this case you move the subject and camera between each frame. You'll have to work out the speed the subject will be moving at in order to work out the distance between each frame. The concern would be keeping the camera set on the subject so that they don't wobble around the frame, perhaps using a mark on the viewing screen as a reference. Motion tracking in post would assist with in this in post.

Edited November 21, 2016 by Brian Drysdale

Phil Connolly · November 28, 2016

Most professional stop motion animators use a software package called Dragon Frame. It allows you to control most of the settings on many dlsrs and gives you a live view of the image - with lots of framing guides and onion skin options. These allow you to see the previous frame(s) and check alignment/movement etc.

You can just do simple stop motion by taking stills on a camera and using a tape measure/marks on the monitor to line up shots as well. But the software makes it easier and gives you more tools.

I

Gregg MacPherson · November 29, 2016

Mohamed,

That was a delightful little film...

In the old days, when lots of artists played with film cameras, the effect you see here would be easily created with a windup 16mm Bolex. Fairly close to what is seen here anyway.

If you shoot actual single frames in a film camera, hand held, keeping steady (still), perhaps shooting two or three frames of each or some particular mment, you are very clse to what we see here.

Some call "stop motion" a process where you lock the camera off and incrementally alter the subject. The scene of the boy being beaten, becoming a crushed little sack, is basically being done that way.

With film work processes being basically lost to most people now, the digital option may be to shoot, even at 24fps, select frames at an interval, and replicate some, occasionally. It is a disturbing thing to me, that digital, begins by finding the means to replicate the look of film, then, as in this technical case, achieves smething like single frame and stop motion, by a clumsy paraphrase of proccess.

Are there any artist left in Egypt who have done what I describe? Find one, and he/she will be an interesting person to meet. Banging off single frames on a Bolex etc is an unmissable experience for any artist working in film.

Edited November 29, 2016 by Gregg MacPherson

David Mullen ASC · November 29, 2016

The switch to using digital still cameras for stop motion and time lapse happened some time ago because you aren't limited to recording video codecs, you can shoot full-frame high resolution stills like at 4K, 6K or 8K and all in raw mode, keeping all the dynamic range of the camera's sensor with no compression added. Then you can batch convert the stills to whatever format you want.

Sarah Thompson · September 16, 2017

I was obsessed with this kind of stop motion when I was a young teenager in the 1970s. I used my dad's Super 8 film camera with a cable release so I could fire it one frame at a time. These days, you can do this with basically any DSLR or even a compact camera so long as you can set it to manual exposure and put it on a tripod.

What I am pretty sure was the technique used for most of the film above was to set up the camera on a tripod, frame the subject then lock the tripod down. Shoot a frame, then move the whole tripod back a few inches and have the actor move the same distance. Shoot another frame, move, shoot, move, etc. For some of the later shots where the camera was stationary, the same technique can still be used -- move the subject, shoot, move, shoot, etc. It takes a lot of patience!

The reason for manual exposure is to prevent flickering from frame to frame as the camera adjusts itself to the scene.

Finally, you can usually just import all the frames into your editing software as an image sequence and then set a frame rate. You might find that setting a slower than 24fps rate will work best -- my guess is the video did that too. Back in my Super 8 days I would often click twice for each movement, so 12fps would probably work well.

Give it a try -- one thing is certain: you will have a lot of fun doing it! :)

David Mawson · September 16, 2017

What I am pretty sure was the technique used for most of the film above was to set up the camera on a tripod,

The reason for manual exposure is to prevent flickering from frame to frame as the camera adjusts itself to the scene.

That's how I'd have done it with film. But with digital, give they were shooting in quite crowded areas - and on some surfaces where tripod alignment may have been difficult, maybe not.

Instead shoot with room to crop, import all the frames into a stills editor one after the other, align and crop. If you use the grid and digital level you won't have to crop much, and it's better than try to defend a tripod. Oh - and if you want to main constant subject distance, lock the focus and then use focus peaking and magnified view to position the subject with the lens wide open. Then close down to whatever aperture you want.

..If camera subject distance is always the same, and the subject position on the grid is, and the camera is always level, then I'd say a tripod is redundant.

(But I may be reacting too much from my experience of shooting street fashion stills - which taught me to do almost anything to avoid deploying tripods and light stands, right down to building softboxes I could hold with one hand while shooting the camera with the other.)

Edited September 16, 2017 by David Mawson

Brian Drysdale · September 16, 2017

The tripod isn't redundant, shooting fashion stills is very different process, although ideas can be taken from many places, as a standard procedure you're just creating unnecessary work for stop motion, unless you're after a stylistic effect. In this particular case they may have used software that motion tracked the face and allowed some variation in the camera/subject distance.

Edited September 16, 2017 by Brian Drysdale

David Mawson · September 17, 2017

Brian Drysdale, on 16 Sept 2017 - 11:51 PM, said:

The tripod isn't redundant, shooting fashion stills is very different process,

I don't find that a meaningful statement. There is no one process for shooting fashion stills and each frame here is a still. That's it. If you think the tripod is redundant you need to explain why - because speaking as someone with a degree in theoretical physics, highschool maths should tell you that it if distance, orientation and position in the frame are constant, the images will match.

In this particular case they may have used software that motion tracked the face and allowed some variation in the camera/subject distance.

They may have done all sorts of things - like using a gold covered helicopter to make a journey they could have made in a van. But that has nothing to do with whether using peaking, the grid and scale will work. Generally, people do the simplest and cheapest thing that will work rather than the most complicated, which is why you see more vans than gold helicopters. Certainly that's the smart thing to do if you want to repeat the result.

If the subject is in the same apparent position in the frame and at the same distance each time, the images will merge. Getting position right is easy using the grid - or you can mark the positions for the eyes on a screen protector. Then narrow peaking and a wide aperture will let you keep distance the same within a limit of better than 2% with a fast lens. That's it; you're done - no software required. It's basic geometry. You could even do without the focus peaking if you match positions for a decent size triangle of points - say the shoulders and belt buckle.

As for "allowing variation", that's even easier - you just align using the above method then move the camera a tiny amount, using the position of the pupils on the screen relative to the dots on the screen protector. Probably you'll draw a line between those two marks and add a perpendicular. You could another marker for a position on the torso if you wanted and draw a line to that.

Or if you want a very controlled shoot, you could have an animator draw the key frames in a tracing paper flick book - just using stick figures. You stick them over a tablet connected to the camera with wifi and the model just matches the pose.

...I write software, and I used to work in a videogames company that specialized in motion capture. And honestly, there is no way I can imagine the software giving a better result - and it certainly wouldn't be easier to use. (I mean, you track the image, input the data into a 3D animation package where an artist has configured a model to exactly match the subject, then the artist adjusts the model to get the next frame and you try to match it? Why? Drawing keyframes on tracing paper would give the same result at a fraction of the cost and pain.) It isn't meant for applications remotely like this.

Edited September 17, 2017 by David Mawson

Tim Smyth · December 29, 2017

It is a nicely done piece, and this form of stop motion is called pixilation, if anyone is interested.

My professional guess is they used no software to do this, as that would take a bit longer, as David above mentioned, and quite frankly it just does not look like they used any frame-grabbers, in some ways it is just to fun looking. I think they winged it blind, which also makes me think they used a tripod, as taking so many frames that line up is not the same as taking hand held photos then having to match them in the computer. Certainly they used tripods for the steady shots, so they probably did for all, as again, it would have saved them a lot of work.

Anyone who has ever tried stop motion with a hand held camera will understand this.