HELP! Close-up 24P Motion Problems

[Greg Schlueter]

Greetings All-
Continuing to navigate the V1U learning curve. Any help or validation on this one would be great.

I recently shot a procession in 24p [not A], 1/48 - framed area between 10 and 3 feet, as one by one they passed through the framed area. Horrible motion blurr... or fragmenting... or ghosting... or whatever one might call it. It's obviously less recognizable viewing from the camera vis-a-vis HDMI on an HD screen. I captured through HDLINK, selecting 3:2 pulldown, large file, progressive mode-- to AVI.

I'm looking for someone to give me some good settings to eliminate this (acquisition), or to suggest that 24p is not the rate to shoot for close-ups with motion, or any settings in capture, or NLE where this can be eliminated... or anything else! I use the latest Premiere Pro 2 with AspectHD.

Thanks for any feedback!

[Marcus Marchesseault]

"Horrible motion blurr... or fragmenting... or ghosting... or"

Describe what you are seeing or post a clip. Ghosting, fragmentation, and motion blur are three completely different things. I suspect you are noticing "stuttering" at 24p since the framerate is so low, but I can't tell from your description.

[Steve Mullen]

Originally Posted by Greg Schlueter

... or to suggest that 24p is not the rate to shoot for close-ups with motion

I "suggest" you not use 24p in any situation where objects pass in front of the camera.

Yes, you can buy my book which does gives you the "dos" and "don'ts" -- but you still need tons of practice to "see" situations and setup a working shooting position for 24p.

And, even if you do have the experience -- the nature of low-cost video cameras and video lenses -- makes judder, even with a 180-degree shutter, NOT look like film.

TIP 1: If you are serious about 24p, you should buy a JVC camcorder and spend even more on prime adaptors. Then, with a whole lot of experience, you can get video that looks as good as this sample:

http://www.cineverapictures.com/video/CineVeraV1-HD.mov


TIP 2: Shoot 30p, not 24p. The slightly higher frame-rate helps and there are no editing issues.

[Bob Grant]

Originally Posted by Steve Mullen

I "suggest" you not use 24p in any situation where objects pass in front of the camera.

Yes, you can buy my book which does gives you the "dos" and "don'ts" -- but you still need tons of practice to "see" situations and setup a working shooting position for 24p.

And, even if you do have the experience -- the nature of low-cost video cameras and video lenses -- makes judder, even with a 180-degree shutter, NOT look like film.

TIP 1: If you are serious about 24p, you should buy a JVC camcorder and spend even more on prime adaptors. Then, with a whole lot of experience, you can get video that looks as good as this sample:

http://www.cineverapictures.com/video/CineVeraV1-HD.mov


TIP 2: Shoot 30p, not 24p. The slightly higher frame-rate helps and there are no editing issues.

I'm curious as to how a lens affects motion rendition, all else being equal.
Certainly 35mm lenses and their impact on resolution and DOF are part of the look of film but not the motion rendition.

[Steve Mullen]

I never used the word "motion." I used the word "judder."

If you understand judder -- and also lens MTF and the frequency response of video cameras verses that of film -- then the connection should be clear.

There's a real reason why CineAlta cameras with 1920x1080 chips and very expensive prime lenses, don't show the nasty judder than inexpensive video cameras do.

Some folks keep insisting that if the shutter-speed is the same -- then the judder "must" look the same. Others of us have long known this simply isn't true. Judder from the V1 is far more objectionable than that of film.

And, why the 24p of different low-cost cameras may have different amounts of "visible" judder.

[Douglas Spotted Eagle]

Greg, if you're going to shoot 24p, you'll find it helpful to learn some cinematography techniques vs shooting angles that you'd normally shoot with 60i. For example, have you ever noticed that long and mid shots in most commercial works are rarely ever perpendicular to the action? Whether shooting Varicam, HDCAM, XDCAM, 16mm, or 35mm, the technique is about the same. There are a lot of great books on the subject, or you might search out posts here from industry gurus like Chris Leong, Charles Papert, Nate Weaver, Greg Boston, or others that work heavily in 24p. We use it for projects wherein clients demand it, and if you know even the basics, you can do a good job. But if you shoot 24p with the same attitude used for 60i, you'll likely be disappointed most every time.
24p on it's own is magical and troublesome at the same time. Angle/composition is critical with 24p.
There are some sweet advantages to 24p, but if you're looking for a film cadence, bear in mind you cannot pull 24p from 30op with even the best tools available. 60i to 24p works well, however.
Bear in mind that a lot of what surrounds 24p is hype vs reality, and if you're coming from a video background vs a film background, it's can be very challenging.

[Steve Mullen]

Don't get confused by the misuse of the word "cadence." There is no difference in cadence between 24fps film and 30p. The only difference is their RATE. A difference that really won't be noticed by an audience. (Both are doubled for presentation by a factor of 2.)

When 24p video and 24fps film are converted to 60i video, the cadence IS changed by the 2-3 pulldown. Now one has a pattern that does look different. It looks different than film.

So you have to ask yourself -- do you want the cadence of film OR the cadence of film shown on video. If the former, then shoot 30p.

If you want to go to film -- I expect a very unlikely event -- then you should shoot 24p and edit 24p. (Or, shoot 60i and deinterlace to 24p and edit at 24p.)

[Bob Grant]

This article here:
http://www.bbc.co.uk/rd/pubs/whp/whp...les/WHP053.pdf

Covers the impact of Modulation Transfer Function, AC and DE on judder quite nicely.

It's far from a simply matter though, increased high frequency detail from lenses with high MTF could well serve to increase judder, at least by my reading of the article. I've worked with 16mm and 8mm film, a lot of it shot through very old low MTF uncoated lenses and seen no significant impact on judder, rather it's less not more noticeable, probably explains why 8mm at 18fps didn't look half as bad as it should.

I suspect the problem the original post is referring to is where the entire moving object appears to judder or strobe and somehow I doubt MTF has much to do with it. In my crude tests with a V1 directly connected to a monitor and walking in front of the lens being in focus or not has no impact on the judder or strobing.
My guess is it's the viewing conditions such as ambient light levels, what the display is doing and shutter angle that'll be having way more impact than MTF in this case.

[Steve Mullen]

The MTF of the lens determines the "frequency" -- size of detail -- that has the greatest contrast. It acts as a lo-pass filter of what gets to the CCD or film. The more it truncates the upper response (which reduces DETAIL), the more mid-frequencies (EDGES) appear relatively stronger. Thus, a cheaper lens removes more DETAIL than does a really expensive prime lens used for CineAlta and 35mm film. The better lens allows a wider frequency range to pass. So it has it's role.

Of course, the frequency response from the CCD (and EIP if one) play the biggest role. In order to get an adequate signal level at the highest frequency (FINE DETAIL) -- the more the mid-range frequencies (EDGES) must be emphasied (boosted). Think of it as boosting an audio signal at 4kHz in order to boost a sagging upper-end at 15kHz. "Presence" is thereby boosted.

Film's frequency response is inherently flatter -- with no mid-range "bump." It has no Nyquist filter because it is an analog capture. (It's the Nyquist low-pass filter that kills the high-end. That's its job.)

Why is all this important?

Judder appears on edges that are moving. Not on fine details that are in motion. The sharper the edge, the more visible the judder. Inexpensive video camera inherently have a huge bump because their CCDs have so little real resolution. They also don't have the expensive amplifiers and filters that can flatten the response curve from the CCDs.

These cheaper cameras thrive on "edges" rather than "fine detail." Their boosted edge response creates tons of judder. You can set the shutter-angle to be the equal of film -- it won't help because the judder is due to their CCDs' frequency response.

You can set the shutter-angle to 216 to increase motion blur. But, this won't help because the blur is applied equally to edges and fine detail. So all you get is an overall blurry pix. Not at all what film gives you.

If you want a film look, forget the V1 and other inexpensive "24p" video cameras.

All you'll get is the temporal rate of film with far more judder than film. Worse, if you view 24p carried within 60i -- you are even further away from film. Now you have "temporal judder" plus "cadence judder" from the 2-3 pulldown.

An ungodly mess IMHO.

PS 1: that's why it won't help all that much to follow the rules of film shooting. You aren't shooting film. :)

PS 2: See the current thread on using a ground glass adapter for 35mm lens. Consider what the glass may do as a diffuser. Perhaps this is why the JVC looked so great. It used an adapter.

[Douglas Spotted Eagle]

Originally Posted by Steve Mullen

PS 1: that's why it won't help all that much to follow the rules of film shooting. You aren't shooting film.

This statement is the one that most folks turn to when they don't understand the production process. No, you're not shooting film. And very little of the film look involves latitude.

Of *course* the "film look" involves cadence; cadence is the measure of beat, movement, or increments of time. 24p is 24 progressive frames per second. 30p is 30 progressive frames per second. 60i is 30 half-frames per second. Wouldn't you agree that these are all very specific measurements of movement or increments of time?

You seem to know a lot about characteristics of 24p and film; I for one, would appreciate seeing any video you've shot that incorporates any filmic value or characteristics. I'm sure we could all learn something from viewing it.

It doesn't matter if you're shooting a cell phone camera or a Viper, managing the camera as though it were film is the only viable option for getting the "film" look. Angles, lighting, and particularly managing movement, whether the subject is moving or the camera is moving is hyper-critical.

[Bob Grant]

Originally Posted by Steve Mullen

The MTF of the lens determines the "frequency" -- size of detail -- that has the greatest contrast. It acts as a lo-pass filter of what gets to the CCD or film. The more it truncates the upper response (which reduces DETAIL), the more mid-frequencies (EDGES) appear relatively stronger. Thus, a cheaper lens removes more DETAIL than does a really expensive prime lens used for CineAlta and 35mm film. The better lens allows a wider frequency range to pass. So it has it's role.

Of course, the frequency response from the CCD (and EIP if one) play the biggest role. In order to get an adequate signal level at the highest frequency (FINE DETAIL) -- the more the mid-range frequencies (EDGES) must be emphasied (boosted). Think of it as boosting an audio signal at 4kHz in order to boost a sagging upper-end at 15kHz. "Presence" is thereby boosted.

Film's frequency response is inherently flatter -- with no mid-range "bump." It has no Nyquist filter because it is an analog capture. (It's the Nyquist low-pass filter that kills the high-end. That's its job.)

Why is all this important?

Judder appears on edges that are moving. Not on fine details that are in motion. The sharper the edge, the more visible the judder. Inexpensive video camera inherently have a huge bump because their CCDs have so little real resolution. They also don't have the expensive amplifiers and filters that can flatten the response curve from the CCDs.

These cheaper cameras thrive on "edges" rather than "fine detail." Their boosted edge response creates tons of judder. You can set the shutter-angle to be the equal of film -- it won't help because the judder is due to their CCDs' frequency response.

You can set the shutter-angle to 216 to increase motion blur. But, this won't help because the blur is applied equally to edges and fine detail. So all you get is an overall blurry pix. Not at all what film gives you.

If you want a film look, forget the V1 and other inexpensive "24p" video cameras.

All you'll get is the temporal rate of film with far more judder than film. Worse, if you view 24p carried within 60i -- you are even further away from film. Now you have "temporal judder" plus "cadence judder" from the 2-3 pulldown.

An ungodly mess IMHO.

PS 1: that's why it won't help all that much to follow the rules of film shooting. You aren't shooting film. :)

PS 2: See the current thread on using a ground glass adapter for 35mm lens. Consider what the glass may do as a diffuser. Perhaps this is why the JVC looked so great. It used an adapter.

I assume you've plotted MTF curves of the V1 and other HDV progessive scan cameras to backup your conclusions? Such plots and a bit of maths would determine if the camera is producing negative MTF at any point in it's dynamic MTF curve.
I also have to ask what this has to do with the original post, the issue you've brought up will affect all objects regardless of how much they fill the field of view. I'd hazard a guess that it's less likely to affect large objects (less edges relatively) than small objects.
It'd also be reasonable to point out that film is more capable of producing sharp edges than any video camera and the problem of judder on sharp edges is just as apparent on film as it is on HD cameras. That's not to say that the use of AK might not be contributing to the problem but I guess that by now the designers of HDV cameras are well aware of the potential traps.

Judder also doesn't appear just on edges, I can quite readily create judder on the whole of a large object and one with very soft edges. This is simple to do, throw it out of focus. Sure this doesn't look 'filmic' but that's not the point, the point is that AK and it's impact on edges isn't the only cause of judder. Come to think of it shooting through a piece of GG that harms resolution is equally not very 'filmic' either, certainly not when seen on a big screen.

[John McManimie]

Why not offer some practical advice instead of telling everyone to buy a specific book and a JVC camera? Throwing money at every problem or not using the camera for what it was intended is not practical advice.

Douglas Spotted Eagle provided some very practical advice --- learn and apply proven techniques.

There is no reason that the equipment available (such as the V1U) can't produce acceptable images.

Originally Posted by Bob Grant

My guess is it's the viewing conditions such as ambient light levels, what the display is doing and shutter angle that'll be having way more impact than MTF in this case.

Absolutely! Part of the problem stems from the fact that many people are still thinking about film and video as they were, not as they are. The viewing situations have changed; we now have more people watching on computer displays in bright environments and huge televisions at close range – these play a huge part in the look of the final product. The refresh rate of the display and the ambient light play a large part in avoiding flicker. The brightness of the video itself plays a part as well. It really depends on the intended target.

-----
From “Digital Video Quality Vision Models and Metrics” by Stefan Winkler
(p 37):

“As analog television was developed, it was noted that flicker could be perceived at certain frame rates, and that the magnitude of the flicker was a function of screen brightness and surrounding lighting conditions. A motion picture displayed in the theater at relatively low light levels can be displayed at a frame rate of 24 Hz. A bright CRT display requires a refresh rate of more than 50 Hz for flicker to disappear.”


-----
From “Digital Video and HDTV Algorithms and Interfaces” by Charles Poynton
("Flicker, refresh rate, and frame rate", p51):

"Many displays for moving images emit light for just a fraction of the frame time: The display is black for a certain duty cycle. If the flash rate – or refresh rate – is too low, flicker is perceived. The flicker sensitivity of vision is dependent upon the viewing environment: The brighter the environment and the larger the angle subtended by the picture, the higher the flash rate must be to avoid flicker. Because picture angle influences flicker, flicker depends upon viewing distance.

The brightness of the reproduced image itself influences the flicker threshold to some extent, so the brighter the image, the higher the refresh rate must be. In a totally dark environment, such as the cinema, flicker sensitivity is completely determined by the luminance of the image itself. Peripheral vision has higher temporal sensitivity than central (foveal) vision, so the flicker threshold increases to some extent with wider viewing angles.

In the darkness of a cinema, a flash rate of 48 Hz is sufficient to overcome flicker. In the early days of motion pictures, a frame rate of 48 Hz was thought to involve excessive expenditure for film stock, and 24 frames per second were found to be sufficient for good motion portrayal. So, a conventional film projector uses a dual-bladed shutter... to flash each frame twice. Higher realism can be obtained with single-bladed shutters at 60 frames per second or higher.

In the dim viewing environment typical of television, such as a living room, a flash rate of 60 Hz suffices. The interlace technique, to be described on page 56, provides for video a function comparable to the dualbladed shutter of a film projector: Each frame is flashed as two fields. Refresh is established by the field rate (twice the frame rate). For a given data rate, interlace doubles the apparent flash rate, and provides improved motion portrayal by doubling the temporal sampling rate. Scanning without interlace is called progressive.

A computer display used in a bright environment such as an office may require a refresh rate above 70 Hz to overcome flicker. (See Farrell.)

Farrell, Joyce E., et al., “Predicting
Flicker Thresholds for Video
Display Terminals,” in Proc.
Society for Information Display
28 (4): 449–453 (1987).
-----



As for MTF, it varies with film cameras as well (see: http://bssc.sel.sony.com/Professiona...lSeminar2.pdf).

[Greg Schlueter]

After not getting any response for a few days, I am VERY grateful for the input of all-- and even banter-- on this subject. It is all very instructive... 1) Need to shoot differently than video... angles, etc.; 2) Need to consider acquisition settings... (I probably should shoot 60i HDV... leaving open the possibility of eventual 24p in editing... I use Premiere Pro 2/ AspectHD).

I appreciate the challenge to distinguish what my film did-- "judder" is probably accurate.

For what it's worth, the temperature variance in this place was horrible-- probably about 20 different temps in the building (what do you do in this situation?); I adjusted the critical area in the full (available) light, set the white balance (etc.)... then went to capture b-roll (can't easily adjust WB when action is happening)... just before the procession, setting up in the critical area, someone turned off the front lights.

To what degree that mattered, I don't know-- but judder is distinct from outside 24p shots.... Yes, the angle advice from DS Eagle is very helpful.... Unfortunately, a realm of my shooting is live, and I have little control.

I'm going to keep reviewing what you all have said... want to excel at this... and appreciate your willingness to help....

[Greg Schlueter]

John McManimie-- your post is worth it's weight in gold.... Thoughtful, documented response that hits the bullseye ... a basis for me to concretely approach the next situation with the "right" criteria and considerations....
THANK YOU!

[Douglas Spotted Eagle]

OK John, if you're gonna put my name in a post along with Poynton's (His book sits next to my mastering system with the 2007 Broadcast Engineering guide) then I need to add "Buy my book." Heading into the third edition, it doesn't delve at all into MTF, but it does cover what you need to know, including a discussion of 24p and suggested compositional methodology for small format camcorders.
We've reached a stage where we've supplied HDV-acquired media to most every major network, and it was most recently used in an upcoming Steve Terrell film (theatrical release) as well. A great deal of it (including the Terrell piece) is 24p. The aerial Superbowl commercial was done with 24p. The goofy "Snickers" commercial that received all the post-Superbowl flak....24p HDV.
There is a sweet spot in which these cams work very nicely, you just can't push them to certain extremes. This is where XDCAM/low end CineAlta start to creep in. You can spend 4.5k on a V1 or 45K on an F350 with great glass. What you choose usually depends first on budget and second on needs.
Back to the closeup question, you might experiment with a faster shutter.
Also, are you sure your HD display isn't interlacing the image?