Andrew Somers
Forum Replies Created
-
Shutter speed relates to motion blur. Common cinema cameras use a 180 degree to 220 degree (open) shutter.
At 25 fps, then means 1/50th a second (for 180 emulation) or 1/40th second (for 220 emulation).
Why shoot at 50 fps and then down convert to 25?? Unless you need to go in and out of slow mo, that makes no sense to me.
If you shoot at 50 fps and a 1/100 shutter, but then drop half the frames, it is THE SAME as shooting at 25 fps and a 1/100 shutter. This is equiv. to a 90 degree shutter. If you shoot at 1/200 it is like shooting with a 45 degree shutter.
NOTE: The battle sequences in “Saving Private Ryan” were shot with a 45 degree shutter to give it that snappy staccato look.
1/1000 shutter will essentially eliminate motion blur – if you want that for an artistic reason go for it – but do tests first, you may not like the results.
-
No, prior to the introduction of sound, films were shot and projected at a widely varying number of rates – from 16 fps to over 26 fps.
In the silent era, each studio had their own standard rate system, and in theaters not controlled by the studios, some theater owners increased frame rates of projection from what was shot, in order to fit in more films per day.
In the sound era, 24 fps was chosen because a single standard was needed for sound, as variations in filming and projection would result in objectionable sound issues – among other things EARLY sound was NOT optical, it was on a separate disk (i.e. a record), that had to run in sync with the film.
Hysteresis synchronous motors allow for accurate motor speed based on AC line power frequency. 24 fps is easy to “sync” using a system built with hysteresis synchronous motors, as is 30 fps. But other “off” frequencies are less straight forward. (a HS motor running at 240 Hz needs only a direct 10:1 gearing ratio for one revolution of a shutter in the camera, etc etc).
24 fps is more economical to shoot than 30 fps, and gives acceptable motion smoothness.
Nevertheless, many Todd-AO “roadshow” pictures were shot at 30 fps, such as the musical “Oklahoma”.
Optical sound tracks can work at various different frame rates, including rates lower than 24 fps – the 24fps was chosen more due to the need for a single standard, and a stable speed, which was easiest to achieve using line powered synchronous motors. In 35mm, consider that the film travels at 90 feet per minute. But in 16mm it’s 36 feet per minute. Sound tracks don’t care about fps, but feet per minute, and 16mm typically used optical tracks as well.
24 fps was considered the best trade off for economy (less film shot), stability, smooth motion, ease of engineering projection systems that ran at the rate, etc etc.
-
No, he’s talking about motion picture projectors in the movie theaters that flash light on the same frame of film twice before advancing to the next frame.
No he’s not – he’s talking about the EXPOSURE LENGTH when shooting film at 24 FPS – it is germane to AE and other such software as when you are adjusting the motion blur settings, you are (in effet) adjusting exposure length.
In a camera that uses a 180 degree shutter, the exposure length (at 24 fps) is 1/48. A 1/48 exposure give a certain amount of motion blur – to emulate that, you set your motion blur setting in After Effects (or whatever program is rendering the motion) to that length to emulate the look of film at 24 fps.
This is not related to projecting in a theater. In a theater, the projector uses a two or three blade shutter, so that each frame of film is flashed on screen two or three times between pull downs. i.e. 48 or 72 flashes per second. This is done to reduce flicker and is not related to motion nor motion blur.
-
When I play the two outputs side by side the 24fps is noticeably less smooth than the 30.
This is probably related to your computer’s monitor refresh rate. What is your refresh rate set at?
For most computers/LCD monitors the*refresh* rate is 60 – and this is a problem for 24 frame video since the computer has to interpolate frames at odd increments to make up for the fact that 60 is not evenly divisible by 24.
Your monitor needs a refresh rate that is evenly divisible by 24, for 24 frame video to play smoothly – such as a rate of 72. OR, the refresh rate needs to be high enough that the interpolated frames are of short enough duration that they are not perceived as jutter.
This is partly why better home theater displays are capably of a true 24P refresh rate, or can refresh at the much higher 120 hz rate. 120 Hz is divisible by 24 by exactly 5.
If you monitor is refreshing at 60 (as most are) then 30 fps will look smoother than 24, especially for certain levels of contrast and motion.
ADDRESSING OTHER COMMENTS:
The integration point of human sensitivity in sight, sound, and touch for “individual impulse stimuli” vs “continuous stimuli” begins at approximately 12 to 16 Hz) though it can be lower, as low as 4 hz for touch). For stimuli with lower frequencies, we perceive individual impulses, vs a continuous stream – but related to this is the differentiation between “on” and “off” of the stimuli – i.e. contrast.https://en.wikipedia.org/wiki/Psychophysics
I.e. in sound, where we transition rom hearing a series of distinct “clicks” vs a “continuous tone”, or in touch, where we feel distinct “taps” vs a vibration. In vision, we begin to relate to motion in 8 to 12 Hs range, with smooth motion appearing above 16 to 20 fps.
FLICKER
The region between 16hz and 75 Hz is very dependent on contrast, that is, the difference between the absolute value of ON vs OFF. But also critical is the length of the ON vs the OFF signal.
But flicker can be perceived as high as 100 Hz and higher.
https://en.wikipedia.org/wiki/Flicker_fusion_threshold
The phosphors in old CRT type PAL monitors were formulated with a longer persistence than those for NTSC monitors because PAL refreshed at a 50Hz rate, and NTSC refreshed at 60Hz.
In your LCD monitor, the backlight is typically going to be operating at a very high rate (NOT related to the refresh rate), so it won’t flicker – but the REFRESH rate will have an impact on the perception of *motion*.
MOTION
Motion is processed by a different area of the brain than still images. Things that help trick the brain into perceiving smooth motion include motion blur, number of images per second, and importantly the *equal distancing* of the stream of images.
While the region between 16 fps and 24 fps may have a varied perception of smoothness of motion, at 24 fps, smooth motion is “normally accepted” – but interestingly studies have shown that human’s have a different emotional response to higher frame rates – between 24 fps and 72 fps, higher frame rates result in the perception of being “more real”. There is no improvement in the “reality” sensation above 72 fps. A hypothesis is that as frame rates increase, the area of the brain that processes that experience shifts with higher frame rates.
In the case of creating a sequence of images intended to trick the brain into perceiving motion, the length of time eace image is exposed will affect the amount of motion blur.
Note that in the film “Saving Private Ryan”, in the battle sequences, they set the camera shutter to a very short 45 degrees, to get a “staccato” sharp effect. While you still perceived motion, it had a “snappiness” to it.
Some film cameras allow the shutter to be opened as much as to 220 degrees, allow for increased motion blur – and some digital HD cameras allow for an “equivalent” 360 degree shutter – but in that case, the motion blur is SO long that it “feels like video”.
Cinema 4D: If at 24 fps, the equivalent shutter angle for a 1/48th exposure is 180 degrees. At 220 degrees, the equivalent exposure is close to 1/40th
24 FPS films in theaters: Most modern projectors use two or three bladed shutter to project the image on screen two or three time between pull downs – this eliminates flicker that would occur from a single bladed shutter cutting light off at 24 fps.
ShowScan: Showscan is not about making motion seem “more smooth” it’s about making the total experience “more real”. Running the film at 60fps allows for a small single bladed shutter, resulting in much more light hitting the screen. Also, Showscan is shot on 65mm (70mm projected) so the image area is nearly 4 times that of 35mm. The large image area and high frame rate together reduce the appearance of film grain (and dust), and greatly increase the perception of “being real”. Showscan has a nearly 3D feel to it (without the need for glasses, LOL).
However, there are reasons that this is not necessarily desireable in a narrative film – Stu has this to say on the subject:
-
I wanted to do this effect with a handheld camera. Basically I want to film someone teleport then move the camera to find them as they reappear. Problem I’m not sure how to get this because my understanding is the movement would have to match exactly with the other shots of filming the same teleport spots with the actor not in them.
I can think of a couple ways to handle this shot.
The “standard” way is to use a motion control rig (i.e. not hand held), and do the move with the actor(s), and then repeat the move for the second portion of the shot. With motion control, each pass will match and it is then trivial to blend from one to the other.
A “second” standard way would be to shoot the actor entirely on a green screen stage (hand held) with tracking marks on the green screen set. Then extract the 3D camera, build a virtual set around the actor, etc etc.
— BUT
Assuming you don’t have a mocon rig at your disposal, or a large green screen stage, here is a LOW BUDGET method for the hand-held version:
You need three elements for the shot:
1) A clean plate with the hand held camera motion and the pan from point A to point B and NOT the actor that transports.
2) The actor walking up to point A before he disappears (shot locked down, and shot with a wider lens from the same relative camera position.)
3) The actor walking away from point B after he reappears (shot locked down, and shot with a wider lens from the same relative camera position.)
Now, 3D track the hand held shot using PFtrack, Syntheyes, BouJou or whatever. Set manual markers at the actor’s location for point A and point B.
Import that 3D camera into After effects, using the manual markers for Point A and B as null objects for reference.
Now take the static elements, and add them at the locations of the null markers. Adjust as needed so that the static shots are locked solid to the motion track.
Roto around the actor to get a good blend into the clean plate, and add whatever FX you want for the transporter energize bit.
Note that for this method, you can be hand held, but you’ll pretty much need to stick to a pan, or at the very least, limit your Z axis movement for the portions where the actor is going to be in shot.
That is, you could plan the shot that you you were handheld, but essentially in one spot for the disappear, and then you walk around and are in another spot for the reappear.
MARK both of these spots, and use *that mark* for the lock down tripod locations for the two elements with the actor. Overall this will work best with wider compositions, a decent distance from subject to reduce perspective errors – if you want tighter, then shoot and work at a higher res, and crop/resize later.
-
A google search for the term “Stock Footage HD” will bring you a wealth of information. You don’t mention *what* kind of footage you need, so it’s rather hard to give you an answer.
You most likely want “royalty free”, as rights managed footage will be *very* expensive, even for a “one off” event.
Some royalty free sources:
Pond5.com
gettyimages.com
thoughtequity.com
And this site has a free 7 day trial where you can download 20 clips a day:
and they are a creative cow sponsor.
-
I want to know how to make the fake cigarette look real by having fake smoke come from this fake cigarette
You will also need to make the end “glow” as it burns, and get brighter when he inhales.
It is not terribly difficult to get realistic smoke, and the glowing end – but it is time consuming and potentially labor intensive.
In addition to After Effects, you will need a 3D tracking software like PFTrack, A good 3D particle system such as Trapcode Particular, and some images of smoke wisps against black to use as a sprite.
STEPS:
1) Do a 3D motion track of the shot (unless the camera is locked down for the duration).
2) As part of the above 3D track, do a motion capture track of the cigarette end.
3) Export the track, selecting the tracker of the cigarette end so it comes into AE as a null, and also select tracker points that define the position of objects that the smoke may need to move around.
4) Import the 3D comp into AE, and add your 2D footage at the bottom of the comp.
5) Create a new comp sized solid, and add Particular to it.
6) Write an expression that takes the position data from the NULL for the cigarette tip, and use it to drive the emitter of Particular.
7) Add a smoke image, and use it as a sprite in Particular – for more on using Particular, watch some of the tutorials at VideoCoPilot. In short, particular will use a sprite of real smoke and emit from the cigarette tip.
8) You can use a second instance of particular, with the same expression but different emitter parameters to create the “burning tip”, animating the particles per second to vary the intensity during “inhales” etc.
Those are the basics. But you will need to get the software and climb the learning curve for 3D tracking and Particular to get the job done.
To ask, in a sarcastic way, did you guys just run out of matches or something?
-
The Gen Arts Sapphire suite (about 240 visual effects in the collection of plug ins) and the monsters suite, among others, are CUDA accelerated in After Effects. A CUDA card can be very useful as the CUDA gpu is used for accelerating rendering of these plug ins.
And nVidia’s CUDA cards are also OpenGL capable, so there is still benefit in After Effects for the non-CUDA capable operations.
CUDA and OpenGL are separate functions, and I have not heard or personally found any incompatibilities between CUDA and OpenGL.
CUDA is a general purpose parallel computing architecture. OpenGL is a cross platform API for graphics.
-
Go under the “WORLD TRANSFORM” section, near the bottom of the plug in, above “Visibility”.
ALTERNATELY, create a camera in the comp, and animate the camera – Particular will obey the camera movements.
-
I don’t know how it would be neater to use just one, because you’d have to designate which part of the mask is sharper than the other. It sounds like just as much work to me as creating two masks.
Among other things, variable edge feathering allows you to have gradual transitions from sharp to fully feathered, which is particularly useful for objects that are partly in focus and partly out of focus. Gradual transitions in feather-width are difficult to do using multiple masks, but trivially simple in Mocha.
Using multiple masks is good for multiple portions of an object that move independently (i.e. arms/body/head), but being able to adjust edge feathering should require multiple masks, especially when feathering is a feature that needs to be animated to match blur/focus issues.
An additional advantage of Mocha is that masks can auto-track object movement.