Making video look more like film, the options

[Guest Daniel J. Ashley-Smith]

The only way's I have used it to set the shutter speed to 1/25th and deinterlace the footage.

 

Any other ways out there?

[DanStewart]

Use a 24p camera if you can, it's less hassle and looks more 'film-like'. Also if you can afford it use a mini-35 adapter. The PSTechnik and the dvx-100a together are generally considered the best on a budget right now (other forum members have shot this combo, I haven't).

 

Or you could just shoot film.... what's your budget?

[David Mullen ASC]

Shoot progressive scan, period. Most solutions taking interlaced-scan video and turning it into a film-look are mediocre at best. You're basically compromising the video signal just to make it behave like something it isn't -- i.e. progressive-scan capture. You end up with bad video that sort of looks (barely) like poor film photography. The point isn't to achieve a film-look no matter how bad it makes the image!

 

Beyond that, it's the typical advice: light and compose like a professional production, reduce the depth of field, lower the edge enhancement, minimize clipping, etc. In other words, any behavior or artifact that is typical to interlaced-scan video and atypical for film, reduce or eliminate.

 

It helps to write a list of all the things that remind you that something was shot in video, not film, and then work on each item one by one (if you can.) But I'd say that the motion characteristics of sampling reality 50 or 60 times a second as fields is a BIG part of the video look.

[Alvin Pingol]

But I'd say that the motion characteristics of sampling reality 50 or 60 times a second as fields is a BIG part of the video look.

But most importantly not the only. I'd say that progressive 24fps capture accounts for only half of the "film-look."

 

The rest,

light and compose like a professional production, reduce the depth of field, lower the edge enhancement, minimize clipping, etc. In other words, any behavior or artifact that is typical to interlaced-scan video and atypical for film, reduce or eliminate

 

like David said, would be the other half.

 

I just get a bit upset when someone thinks that as soon as the progressive capture switch is flipped, the production (and end result) somehow becomes a lot more professional.

 

There was a wonderful short film on television once that I swore was shot on film. The shot was locked, of an old woman sitting next to a window, with subtle fill light, and a strong edge light, all warm CT, set was hazed a bit. It looked wonderful.

 

And then the woman tilted her head up and that's when I noticed interlaced (video) motion rendering.

 

Had this been shot on 24P HD or 24P SD (with a camera like the SDX), there's no way I would've ever known that it wasn't shot on film.

[Guest Daniel J. Ashley-Smith]

I would love to shoot progressive but my camera only shoots 50fps interlaced.

(It's one of those crapy consumer standard MiniDV cameras...)

 

The Mini35 adapter gives it a depth of field, which makes it looks more "film-like". But again, won?t fit on my camera.

 

If I could afford it, I'd buy the XL1s (Or the XL2 if it was out) with a Mini35 adapter. That?s probably the closest budget link to film.

 

Shame they didn't use the Mini35 in "28 Days Later".

 

Just wondering, people say that with digital you can't really get a depth of focus, because of the CCD size. But you obviously do get A depth of field.

What would happen then if you were to adjust the iris on the lens? Would it affect the depth of field a little but nothing really noticeable?

[Carlos M. Icaza]

I think that in "28 Days Later" they did use the PS Techniq Mini 35 adapter.

 

You could always go the post route and use Magic Bullet and some other effects suite to make your video look like film. But as some of the previous post states, you should light it like film, and think everything film like and be aware of all the videos intricacies.

 

I have shot two music videos using an XL-1 and lit as if I were lighting for film ( I come from a film photography background ), then in post, I used Magic Bullet with assorted other filter plugins from After Effects and you can actually see a noticable difference between the post and the raw video. But, it is no panacea, film will be film, but the digital tools are better than nothing. Does it make video look like film? No. But from my little experience and my exposure in doing post processing work, some folks have had to second guess what camera I have used. When I tell them XL-1 they were impressed at the results. But it is not because of the camera, most non film people don't know or don't understand what goes in lighting accordingly but they can certainly see a big difference in the fact that the footage does not look like video.

 

C.-

[David Mullen ASC]

You've got to stop using the phrase "can't get a depth of field" -- it's nonsensical.

 

Anything shot with a lens has a depth of field. The trouble with small target area format is that you have excessively deep focus. What is hard to achieve is a shallow focus look as the target area gets smaller. But an excessively deep focus look like with DV has still got a "depth of field" -- in fact, it has TOO MUCH of it!

 

No, "28 Days Later" did not use the P&S Technik Mini-35. They did use an adaptor to fit 35mm cine lenses on the camera apparently but this did not change the depth of field characteristics.

 

Of course adjusting the iris helps change depth of field -- it's one of the factors that affect it. Trouble is that once you've opened the iris to max, once you've zoomed in all the way to the telephoto end of the lens, once you've focused as close as possible, you can't lower depth of field any more except by using a larger target area and thus longer lenses to get the same field of view.

[Phil Rhodes]

Hi,

 

Even on quite tiny sensors you can get a very appreciable soft focus effect. I used to shoot on crappy little hi8 cameras (as opposed to the crappy little DV camera I now own, but at least it has 1/2" chips) and zooming in normally fixes it up.

 

Phil

[Guest Daniel J. Ashley-Smith]

Ok this depth of field thing is confusing me. Some people are saying that to increase the depth of field, you must close the iris down. But then again, some people are now saying that to get a depth of field, I must open the iris up.

 

It would make sense to say, to get a depth of field, you must close the iris down. Which gives the picture more of a depth and not as flat looking.

So, what?s the clever idea of this, saying it "backwards" thing?

[Phil Rhodes]

Hi,

 

You're not quite getting this - I'm pretty sure I explained it before, somewhere.

 

All images using a lens (pinhole cameras don't, for example) have some kind of depth of field characteristic. The "field" is the area which is in focus. The "depth" of the field is the variation in distance from the camera over which objects will be in focus. For example, if you set the focus to six feet and have a two foot depth of field, everything from five feet to seven feet will be in focus. Anything outside that range will be gradually more and more out of focus as it gets further and further from the area which is in focus.

 

Explaining how this works is slightly complicated. Things which control depth of field include:

 

- The resolution of the recording medium. A lens only actually focuses light to a perfect point at the exact focus distance. However, if the recording medium is made up of big chunky pixels, or film grains, or whatever, so long as the lens is focussing things to a finer point than the recording medium can resolve, the object is in perfect focus - or, as close to perfect focus as the medium can resolve.

 

- The F-stop to which the lens is set. The iris, which I'm sure you've physically seen working in a camera shop, allows light to come through a roughly round hole. This means it can take many paths to the film, from the object and through the lens. The larger this opening is, the more possible paths there are, and the more diffuse the resulting point of light will be. This is a somewhat simplified explanation, but compare the iris in a camera to a pinhole camera. Big pinhole, soft images. Big iris, short depth of field, more focus artifacts.

 

- The focal length of the lens (and THIS is where it gets interesting.) A lens at a given F-stop will always defocus an object at a given distance the same amount compared to the apparent size of the object. However, if it's a zoom lens and you zoom in, you're making that object bigger - so the apparent amount of blur gets larger.

 

- The size of the image sensor. Now this doesn't actually have any direct impact on the depth of field other than that a smaller sensor only "sees" the middle part of the picture from a given lens, so you will tend to use shorter (wider) lenses to create the apparent framing.

 

So it's really kind of circular. You have a smaller sensor than a 35mm frame in a video camera, so you tend to use wider lenses to create the same composition, so you end up with a greater depth of field - more things appear in focus. With small 1/3" domestic video cameras, it's very easy to end up with the depth of field being from a few feet to infinity, so more or less everything appears in focus.

 

A related point is that this is why you don't get really wide-angle lenses for consumer video cameras, and the add-on attachments you can get are often poor quality - you really need a super incredibly wide lens to make a tiny-chip camcorder look at all wide.

 

However, this also works in your favour. If you work towards the long end of the zoom with a tiny-chip camera, you're so narrow that you can achieve a short depth of field and lots of blur. It's still good to keep the iris open, and it can be a pain to work a long way from the subject zoomed all the way in. However, you can achieve a short depth of field and lots of blur with a consumer video camera.

 

Phil

[David Mullen ASC]

YOU HAVE TO STOP USING THE PHRASE "GET A DEPTH OF FIELD" WHEN YOU MEAN "CREATE DEPTH"!!!

 

It still hasn't sunk in but I'll repeat again: everything you shoot with any camera has a depth of field. It merely means the range of distances where objects are acceptably in-focus. DV cameras tend to have a deeper range (deep focus) whereas 35mm tends to have less of a range where everything is in-focus (shallower focus).

 

What you really should be asking is how you can reduce the depth of field with a DV camera, i.e. get a more shallow-focus effect.

 

You're confusing the technical term "depth of field" with the aesthetic term "creating depth". Often the illusion of depth and three-dimensional space in a two-dimensional medium can be accomplished by having a shallow focus effect where only the subject is in-focus.

[Guest colour1]

If you can't get your hands on a camera that does 24p motion ... I would recommend shooting 60i at 1/60th shutter and post processing (I use a program called dvfilm) This is a much better option then using the slow shutter effect on the camera. The shutter effect is created by line doubling which softens the image a little too much for my taste.

 

Of course, other variables like lighting, camera movement and depth apply ...

[Aaron Tan]

To get the effect of a shallow depth of field in a video camera, (When I was using an XL-1s, not sure if it will work with standard consumer cameras) I used the built in ND filter and opened up the iris (lowering the f stops). This will only work in the bright outdoors and well lit indoors. But the effect is sufficient.

 

While doing some test shots, I shot a girl walking down a bridge and zoomed in to her using this method and had her in focus and everything behind her a blur.

[Filip Plesha]

My god, someone has his terminology all messed up.

 

More deph of field produces more flat-looking images, less deph of field

produces "deeper-looking" images. Not the other way around.

So when you open the iris you get less deph of field, and when you close it down,

you get more deph of field.

 

And If you want "deeper" images that look more like film, then use larger apertures. But I don't think this is the cruicial part of making video look like film because it all depends on the aperture used. If you shoot video wide open

it will give less deph of field than if you shot 35mm at f/11. It is just that

if you shoot both film and video at the same aperture, film will have less deph of field. A daylight shot , shot at f/16 will have a lot of deph of field, and can look quite flat, but it still has the film look. I think there are other more important things for getting the film look.

 

Partially it is the 24 fps motion, but then again, even if you freeze a frame, or

print it on a photo paper, film will still look like film, and video like video.

Surely, if you light video like you light film it will contribude to film look, but

a long shot of a ship on sea will still have the film look if it was shot on film,

and video look if it was shot on video.

 

I think that the actual way film captures light holds 50% "shares" or more of the film look. And it's not so much about the image quality (even though there is a huge difference today), because it it were, then the film look would be lost in telecine (because the number of color nuances, and resolution is reduced to what a CCD can pull out of film) but it isn't. It's just because of the different way of capturing light.

 

I think this will be possible to simulate once the digital sensors surpase film

in qualitty (I don't see that in the near future, though) because today video has its

own characteristics that are hard to just jump over. Once there are digital cameras that record light almost as human eyes see it, it will be possible to make anything out of that image, even simulate any of the film stocks from past.

 

But now, i think any atempt to manipulate video images (in post) to look like film is just torture (spelling?)..

[Andy O'Neil]

Rays of light from a point in the subject converge to a point in a plane behind the lens. The rays of light form a cone. Depending upon where the film is in relation to the lens, the film would intersect the cone to form a circle (referred to as the circle of confusion.) If the circle is sufficiently small, the circle appears as a point to the viewer, and the image "seems to be" in focus.

 

Therefore, the range at which these circles appear to be points, is the range that images appear to be in focus, depth of field. I.E. if the depth of field extends from 12' (near distance) to 20' (far distance,) the total depth of field in this case would be 8' (20' - 12'.) The longer the range, the more depth of field you have.