Working with 24p & deep DOF -- some postive ideas.

[Steve Mullen]

Originally Posted by Tony Tremble

I am constantly surprised by "filmmakers" shooting video trying to emulate the film medium of Hollywood. Personally I see these cameras a different medium in themselves and see the differences between film and video as positives to be explored rather than negatives to be worked around e.g. with expensive clunky 35mm adaptors. Some of the greatest filmmakers and if not the finest of our time, Abbas Kiarostami, regularly use Deep Focus as a device.

Deep DOF has several consequences:

1) By not throwing background out of focus, when one pans with a moving subject to eliminate Foreground judder, Background judder will not be prevented.

a) We all know that zooms are taboo. In fact with Prime leneses, zooms are impossible. So, why do we continue to pan? We have widescreen and HD -- so we can put a camera near the ground, shoot-wide, and let the subject move through the frame. Or, place the camera high. The subject will have a low motion-vector and so no FG judder.

b) Normally when we pan using widescreen we leave more open space in the direction the subject is moving. This open space gets more of our attention -- since new objects are entering it -- than the space behind. Thus we will see BG judder here more than behind the subject. Can we compose a tighter shot with the subject in the middle? The extra DOF makes tight framing easier with a moving subject.

2) By not throwing background out of focus, when one moves the camera with a moving subject to eliminate Foreground judder, Background judder will not be prevented. This is a tough one because the availability of steadicam makes these shots so EZ. Perhaps too EZ. Everything is being shot with a moving verses a static camera! Frankly, the style is already out of style. Why carry it into HD?

3) By not throwing background out of focus, when one moves the camera using the ever so popular jib, judder may be introduced. What you may not realize is that when shooting interlaced video -- flat-panel HDTVs will "bob" the entire picture causing 50% of the vertical resolution to be lost. So like pans and zooms, these camera moves should be avoided.

In short, HD allows and encourages, a style where life unfolds through a window into "reality." Think of the westerns that were shot very widescreen and composed with very wide angle shots.

4) We are taught that we should throw the background out of focus to better isolate and focus our vision on the subject.

a) I recently saw a 4:3 movie with a wide-shot of an actor with lush trees behind. When the CU came, the trees went out of focus. Frankly, it bothered me. With widecreen, I would let the tree leaves remain in focus and let the viewer do the focusing on the actor. Rather than "distracting" the viewer, the act of choosing where to look may be more powerful to the viewer.

b) In a photo mag., a photographer who uses Leica cameras with wide-lenes commented that he considers deep DOF vital to providing the "context" in which the subject exists.

In short, although there is a lot of talk about using 24p to make "movies" or "increase the quality look of video" -- the concepts that come along with these desires reflect a very limited view of "cinema."

These limited views lead to judgements of what an HD camcorder "must" offer -- such as narrow DOF. As Tony Tremble said in the quote at the top -- we live in a new world and it seems creativity might be the appropiate response to changing technology.

[Floris van Eck]

Steve, my compliments. Great post, and this is exactly how I think about the whole "DOF" and 24p matters. The only thing many people on this board are worried about is emulating the film look. Film and video are different. Like you mentioned, the movies I remember best are films that do not take the usual approach and are shot from a different perspective. I think if you understand how video/HD video works you can really use it to your advantage.

I am happy to see that I am not the only one with this opinion.

[Sean Hsieh]

I agree with a lot of the points you have posted up there Steve, but I think a lot of filmmakers want to make the jump to a shallower DOF because you just can't seem to find that in typical consumer cameras. I have to admit that I would rather see something on the big screen that I can't do with a $300 camera I'm picking up from Best Buy or atleast achieve a look that isn't similar. I guess what I mean is, if I'm paying for to watch the movie, I don't want to feel like I could have made it pretty easily.

I do think that we will start noticing a change in the way movies are being filmed to utilize the deep DOF. I actually noticed some shots in The Grudge 2 this past week and although they really utilized the shallow DOF, they also used deep DOF quite often to provide "hints" on what would happen next. I guess a good balance is what counts.

[Tony Tremble]

What is really important is the concept not whether you could could produce what is being seen on your own camcorder. Have faith in the concept.

If you haven't seen the film Draughtsman's Contract by Peter Greenaway I well recommend it. It's a strange murder mystery set in 17th century England. Greenaway uses an almost entirely static camera and wide DOF and frames his subjects as if framed by the protagonist in the film. The concept being to morph the medium of film into a moving painting (about a painter).

http://greenaway.bfi.org.uk/material...le=draughtsman

Ultimately the concept is free from the constraints of technology but undue emphasis can sometimes be placed on how best to capture those concepts based on established conventions.

If one looks at the history of cinema the periods when conventions change the most are during periods of technological change.

Viva la resolution!

If you like your cinema to flex the old grey matter, check out Greenaway and Kiarostami's work to see concept heavy work free from the constraints of conventions.

TT

[Steve Mullen]

Originally Posted by Tony Tremble

What is really important is the concept not whether you could could produce what is being seen on your own camcorder. Have faith in the concept.

If you haven't seen the film Draughtsman's Contract by Peter Greenaway I well recommend it.TT

Thank you -- and Greenaway is just about my favorite director.

"I think a lot of filmmakers want to make the jump to a shallower DOF because you just can't seem to find that in typical consumer cameras."

I understand the comment, but the idea that one would choose a look so that it doesn't look like something else -- seems to be reversing the whole process of "creativity."

Video Salon (Japan) has a story on shooting Super 8 and using an adaptor to pull segments of film through a scanner. They list the amazing selection of film stock available. Perhaps, those who want the look of film should just shoot film!

I remember one definition of artistic creativity is the ability to work well within the constraints of the technology being used.

[Bogdan Tyburczy]

Shallow DOF is only one of film look aspects that so many of us are after. There are many other aspects making it kinda difficult to precisely define what "film look" is. How much of what we see on big screen is the characteristics of the negative, how much depends on developing, how much is post and CGI, especially these days?

There are many factors at play, but shallow (or maybe I should say "not too deep") DOF is closer to characteristics of human eye. The way we see the world in terms of DOF is closer to characteristics of 35mm frame than small video sensor. Often we don't even notice that because of natural attention focus on what we look at. It works naturally very well in 3D world, but in 2D space it's different story.

Of course, human faces is what we recognize the best and focus majority of our CPUs on. That's why we see "faces" in so many objects like clouds, tiny drops of water, pollen or dust caught in the flash (sometimes called "orbs") and many others. Human faces is where, imho, shallow DOF is most important because it brings flat screen experience closer to our natural perception.

[Brett Sherman]

While the post here deals with the aesthetics of deep DOF. The other practical consideration is that it is a lot easier to maintain focus on a subject with deep DOF. Since with Cameras like the V1 you're trying to focus on a really low-res display this is a huge blessing. Until the viewfinders get a lot better deep DOF is your friend.

I think the reason that people prefer shallow DOF is that it is less like reality. It's the same reason people prefer 24p over 60i. It creates a sheen of vagueness that extends the suspension of disbelief for fictional films. Shallow DOF also allows the DP or director to control what the viewer focuses on. I think both shallow and deep DOF are both useful tools in our disposal. Personally I prefer the shallow DOF look.

[Greg Boston]

According to an article I read sometime ago regarding the history of film production, shallow DOF was used so that they didn't have to worry about distracting elements in the background because they didn't have the budgets to carefully arrange everything in view. It was easier to throw the background out of focus. That may not be the only reason and it has certainly become a sought after principle even to this day.

I am personally a fan of deep DOF. Not to make focusing easier, but I feel that it more closely mimics the way I see the world. When we look near and far, we do a rack focus, but it happens so fast that it just appears that everything we look at is in focus. With HD, I think we should be working towards deep DOF.

However, shallow DOF is used to force the viewer to concentrate on what the storyteller wants to emphasize at that point in time be it still or video. It's like someone grabbing your head and saying, "Here, look at this."

-gb-

[Steve Mullen]

Originally Posted by Greg Boston

According to an article I read sometime ago regarding the history of film production, shallow DOF was used so that they didn't have to worry about distracting elements in the background because they didn't have the budgets to carefully arrange everything in view.

The history books are probably correct because much of the background was painted to look 3D -- just like on the stage. Even staircases were cut-outs with the "posts" painted to look like they were "real."

Also, with VERY low-sensitivity film there was no choice to shoot with the lens wide-open.

So shallow DOF was dictated by the technology of the day and has become a "convention."

The convention will changea as technology changes -- at least as it does for those with less than $5000 to spend. There is always 24p XDCAM HD with 1/2-inch chips. And, let's be fair -- the JVC HD-series provides both 24p and you choice of lens. So you have $5000, $10000, and $15000 options.

If you really, really want 24p AND shallow DOF you certainly have multiple choices, but you will have to spend money for what you want -- just like it when you buy many things you "want."

[Charles Papert]

I would guess that a lot of the shallow DOF craze that has permeated the small-format world has been partially flavor-of-the-month, and also because it allows users to further simulate the look and feel of 35mm that began with the introduction of 24p/frame mode etc. I do feel like much of the time I see "forced" shallow focus in many people's work these days. I have owned a Mini35 for a few years and have had a few people here and there comment on some of my projects, "it doesn't seem all that shallow". The reason for that is that I shoot exactly what I want to shoot, which may be about isolating someone from the background or it may not. The nice thing about this having the flexibility to make the choice for a given shot rather than being limited into the endless depth of field of 1/3" video. I might shoot a day exterior at T11, or knock it down to T2.8 if I want, with a significant visual difference between the two.

The point about the difficulty of maintaining focus is an excellent one. I believe that we are entering a phase where soft shots in low-budget production will becoming rampant as more alternatives (larger-sensor cameras as well as the adaptors) come on line. Buying the gear is part one of the battle, finding someone to pull focus accurately is a whole different animal. I've already seen many examples of hunting focus in clips on this and other sites. On the big screen, this is much more of an issue than online. Probably this will become the new "look" of cinema--lots of soft shots! (for those who think that they see a lot of this already in movies/TV...that's from seasoned pros who may occasionally miss the mark, and/or directors choosing to use bad takes for performance reasons).

BTW, "Draughtsman's Contract" was shot on Super 16, one of the early features to do so as I recall from the American Cinematographer article at the time, which made it a lot easier to maintain deep focus of course.

[Steve Mullen]

Originally Posted by Charles Papert

I believe that we are entering a phase where soft shots in low-budget production will becoming rampant as more alternatives (larger-sensor cameras as well as the adaptors) come on line.

Amazingly, even Primetime shows have soft shots from time to time.

By the way, for those who love film, follow this link to see Super8 to video. So far NO transfer to HDV, but you can do your own for $1000.

http://www.wrigleyvideo.com/filmxfer/super8sample2.wmv

Picture below of Bolex Super8 camera. I saw it in Germany in the summer of 1966 when Bolex engineers were testing it. I was shooting with a Bolex H8 -- which died. So I bought a Bolex K2. Then a Bolex H16 Rex.

[Steve Mullen]

Really interested in Super8 -- try one of these Canon 1014s, a transfer unit ($1050), and a V1.

OR, here's a transfer to 1080i HDV service link:

http://www.videoconversionexperts.co...se_Process.htm

The lattitude and low contrast of the V1 makes it an ideal transer format.

[Mikko Lopponen]

Originally Posted by Steve Mullen

3) What you may not realize is that when shooting interlaced video -- flat-panel HDTVs will "bob" the entire picture causing 50% of the vertical resolution to be lost.

How can you lose resolution that wasn't there in the first place? Interlaced has half the resolution in moving images. And bob-deinterlacing doesn't remove any resolution compared to the interlaced frame.

[Steve Mullen]

Originally Posted by Mikko Lopponen

How can you lose resolution that wasn't there in the first place? Interlaced has half the resolution in moving images. And bob-deinterlacing doesn't remove any resolution compared to the interlaced frame.

Your post provides the opportunity to provide a link for folks shooting interlace:

http://www.hometheatermag.com/hookmeup/1106hook/

Last year amost 50% of the HDTV tested used "bob" deinterlacing and lost 50% of interlace video's vertical rez.

It's gotten worse -- the number failing is gone up to 55%.

Bottom line, 30p or 60p progressive seems the only safe video to shoot. Everything else will take a huge quality hit when actually viewed in the real-world.

But, don't assume that on the "passing" HDTVs interlace video looks good. Objects that have motion will still lose 50% of their vertical resolution -- and it is painfully obvious when watching on a big screen. You can see the "object" turn into an object with alternating video and black lines. When the object stops moving, full detail returns.

Plus you need to worry when shooting about line-flicker and line-twitter.

Moreover, most HDTVs are unable decode correctly 2:3 pulldown. But, that may be less of a problem if you edit/export so the cadence is never distrurbed.

[Mikko Lopponen]

Originally Posted by Steve Mullen

Last year amost 50% of the HDTV tested used "bob" deinterlacing and lost 50% of interlace video's vertical rez.

That's bull. I can change the interlacing method in my computer at will and bob is excellent for interlaced material. It's absolutely the best way to deinterlace material as it creates 60 progressive frames out of 30. Yes, resolution will go down 50% in moving scenes but that resolution isn't in the interlaced file either. How can you say that bob deinterlacing loses resolutions when it simply displays one field after the other? There's NOTHING lost compared to the original interlaced material.

The only other method that is better is adaptive bob deinterlacing where you "quess" which way objects are moving and use that information to recover lost resolution. But that's actually NOT PRECISE to the interlaced image.

But, don't assume that on the "passing" HDTVs interlace video looks good. Objects that have motion will still lose 50% of their vertical resolution -- and it is painfully obvious when watching on a big screen. You can see the "object" turn into an object with alternating video and black lines. When the object stops moving, full detail returns.

Yes? That's because of interlacing. If you watch that normal footage in an interlaced crt it will do exactly the same thing. Has nothing to do with bob-deinterlacing. In interlaced formats objects flutter. I mean...That's how it works. You can't blame bob for being bad!