Best shutter speed for filmic look?

someday

Well-known member
Hello.
I have noted that HD footage motions (pans, tilts and fast moving elements) are not fluid as film; even at 1:50 shutter.
Why? After all it's progressive and i don't understand the field scattering of HD compared to film.
Is it safe to film at 1/35 shutter? (i don't like the shutter at 1/25 because it look washed out)

.
 
35mm film cameras generally use a 180 shutter, which results in a 1/48 shutter speed at 24fps - the shutter revolves once per frame, open for 180 degrees (or half) of that time, 1/24 over 1/2.


Part of getting the smooth movement is from the operating, but i don't know enough to explain the specifics of it, bu someone else undoubtedly will. (please? i want to know as well)
 
One reason why it is difficult to get smooth pans and zooms with the stock HVX is that it has more DOF. A panning or zooming shot on a film camera is not so objectionable because the background is usually blurred.

Add a 35mm adapter to the HVX with some nice primes and blur the background.
 
The slower the shutter the more motion blur.

Are you sure this isn't just a playback device issue? Lots of older systems can cause 'tearing' when HD video pans.
 
this is an interesting question...i have had this discussion with a lot of my friends.....I have found that even though a 1/48 shutter is equivelant to 180 degrees film shutter it doesn't have the film look completely down when the camera tilts or pans or if there is faster motion...you sort of see the video blur.....

I find a higher shutter works better and gives that film look....the extreme of what I am saying is if you shoot at say a 1/250 shutter or like a 45 degree shutter.....this renders more of a filmic look but at the same time it is more stroby...

so I have found if you shoot closer to a 1/60 shutter or 160 degree or 142.5 degree shutter that it helps with the motion staying more filmic

this is my opinion and you should do tests to see what you like

hope that helps.....

Ronster ;)
 
I have no basys to explain this but trough my own eyes. I'm no engineer. But you can achieve better movement in post. How? I don't know, but a post house in Hong-Kong did it, and it looked great.

I did a tv spot in 108025p, with some shots at 25 shutter. On my FCP edit, the pans looked strobby.

They edited the footage in an Avid DS, then Telecined it, and did some more post in smoke. FInal output looked 100 percent like film movement. I was impressed when I saw it broadcasted. I asked them what they did, and they said "nothing".

My theories:

1- Conversion from DVCPRO HD to HDCAM trough HD-SDI into an HDCAM SR Deck, interpreted the footage as 1080 25p, and removed the interlace flicker (I don't know if this makes sense). Made some sort of pull down?

2- Avid DS did something to the footage/ or the telecine/davinci, or Smoke, and removed the flicker.

3- Miracle.

Suffice to say, the final edit looked motion wise a lot better than my original edit.
 
There are a few posts on this subject and a section in Barry's book about it; its actually a pretty complicated topic.

As a quick breakdown there is a no-mans-land for panning, the pan has to be very slow (and is a function of zoom level / lens size) or fast enough to smear the background (which is a function of exposure).

The trick is how to slow down the fast pan to get the pan speed you want. Understand what you are trying to achieve is a relatively static subject with a background that screams motion. Think Formula One race car photography: you've seen those beautyful stills where the car looks like its moving due to the background blur? You are trying to capture frames that look like that. If it is done properly it defeats the strobing by smearing the background in a specific direction. A 35mm adapter can create background blur (bokeh), but it would actually lessen the blur effect. If you really want clean motion you should probably remove the 35mm adapter and set it up to have the background in focus with your subject. Trust that the motion will 'obliterate' your background.

Many NLEs can do this in post. I know Avid Express had two different functions for adding blur to moving objects which would include background. (I have not used it yet, but I will be soon.) If I understand it correctly the NLE must analyse two frames to see where an object started and where it finished. If that is the case then it would be wise to keep the background simple, with large, chunky objects instead of small specular details.

Bob
 
SPZ, could it be your monitoring? Was your FCP session output to the same monitor as the broadcast you observed? Are you using something like a Kona2 or BM card and monitoring HD-SDI to a broadcast monitor (CRT/LCD?) or from FCP via computer DVI to a 2nd computer monitor?

Was the broadcast in HD or SD, and if so, you observed on a CRT or LCD broadcast or television monitor?

Computer LCD monitors tend to run at 60Hz, and 25p may not be as smooth on playback. Perhaps someone can confirm this.

Interesting topic, many views.
 
One other take on this... since the CCDs scan at 60fps native and then get converted into 24fps, is it possible that we're talking about a non-interlaced cadence issue?
Example:
An object moves across screen in one second
Camera senses this as 60 stages of motion
but outputs the following frame numbers
1,4,6,9,11,14,16,19,21,24,26,29,31 etc
so the object appears to be rapidly changing speed

Simple test to check if this is it. Overcrank to 30fps and see what your pans look like.
 
I find a higher shutter works better and gives that film look....the extreme of what I am saying is if you shoot at say a 1/250 shutter or like a 45 degree shutter.....this renders more of a filmic look but at the same time it is more stroby...

I agree totally
 
Thanks very much.
I found a speed to avoid that effect (it's slooooow). then i looked to many films with faster motions (i don't only mean pans and tilts, but even moving objects) and there is nothing like the HD scattering effect.
 
One other take on this... since the CCDs scan at 60fps native and then get converted into 24fps, is it possible that we're talking about a non-interlaced cadence issue?
Doesn't work that way. On the Panasonic cameras, the CCDs are actually reclocked to the frame rate you're scanning (i.e., they don't always run at 60fps, they run at whatever the appropriate speed is, up to 60fps. So if you put the camera in 2fps mode, it reclocks its chips to scan at 2 frames per second.
 
One other take on this... since the CCDs scan at 60fps native and then get converted into 24fps, is it possible that we're talking about a non-interlaced cadence issue?
Example:
An object moves across screen in one second
Camera senses this as 60 stages of motion
but outputs the following frame numbers
1,4,6,9,11,14,16,19,21,24,26,29,31 etc
so the object appears to be rapidly changing speed

Simple test to check if this is it. Overcrank to 30fps and see what your pans look like.

No.

When in 24p, 12p, 2p, 48p, the cmaera clockrate slows to that exact number. its not stripping the frames out of a constant 60p stream.

The motion is the same. When its projected in a theatre in a cinema Digital projector, or telecined to film, the motion is IDENTICAL.

as far as SPZ's experience, most likely the time seperation, between when you submitted the footage, and saw it on TV, the context changed and you forgot what you were "looking" for. After i shoot footage i forget the mistakes after some time, and it suddenyl looks TONS better than i remebered.


( you beat me to it barry)
 
I assume adding to the shutter (adding 20 degrees to the generally accepted standard of 180) add brightness without adding noise like gain would. Any idea what portion of a stop 20 degrees would be?

Bob
 
Well, I sincerelly do not know why. Its not my eyes, since there really was a significant difference. I need a place that hosts footage. I would show you guys a "before" treatment file vs an after treatment file.
 
I assume adding to the shutter (adding 20 degrees to the generally accepted standard of 180) add brightness without adding noise like gain would. Any idea what portion of a stop 20 degrees would be?

To go from 180 to 360 would be a one-stop increase, so from 180 to 200 would be a 1/9th stop increase.
 
Back
Top