DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   Techniques for Independent Production (https://www.dvinfo.net/forum/techniques-independent-production/)
-   -   What effects Depth of Field (https://www.dvinfo.net/forum/techniques-independent-production/2452-what-effects-depth-field.html)

Rob Lohman June 20th, 2002 06:48 AM

What effects Depth of Field
 
I have some questions regarding Depth of Field. I have been
experimenting with my XL1S and the standard 16x lens to shorten
the DOF (ie, forground subject in focus, background out of focus,
a bit blurry). I haven't had much luck.

Which parameters change the DOF? I know that the more the
lens is open the shorter the DOF (I was at f1.6 while testing).
Does shutter speed and where I am in the zoom range affect
any of this as well? Are there any other things?

Thanks!

Al Osmond June 20th, 2002 07:33 AM

Hi, Rob

I have always understood that depth of field is a function of the size of the lens aperture: NOT the aperture expressed as its relation to the focal length (f/2.8, f/16) but the physical size of the hole.

Thus, in the case of a 300mm f/2.8 lens, the aperture will be a certain size in millimetres. With a 10mm f/2.8 lens, the aperture is the same as far as exposure is concerned, but the actual hole will be considerably smaller.

A lens designed for use with a small target such as a 1/3" chip must have a much shorter focal length than a lens intended for 16mm or 35mm film if it is expected to cover the same field of view. The 16x Canon is 5.5mm focal length (?) at the wide end so the physical size of the aperture is miniscule compared to a lens covering a similar field of view on a 35mm camera.

The same problem (if problem it is) applies to 16mm and 35mm cameras, where lenses covering the same field of view on the two formats don't give you the same depth of field.

The shutter speed does come into the equation, but only because the shutter speed needs to be shortened to keep the exposure constant as the aperture is opened up. It is the aperture which affects the depth, though, not the shutter speed.

To reduce the depth of field to achieve a 'look' similar to that of a 35mm film camera, you're going to have to operate at the widest possible aperture (using ND as necessary) and the longest possible focal length. And the focal length is usually dictated by the size of the subject and the dimensions of the studio!

Realistically, it probably can't be done, although the film industry is populated by folks who believe they can change the laws of physics! :-)

Al

Margus Kivilaan June 20th, 2002 07:47 AM

>and the longest possible focal length.

is'nt it shortest?
Or maybe i'm messed up with terms (i don't feel very confident in English)

Margus

Al Osmond June 20th, 2002 07:52 AM

...better than I am in Estonian. :-)

If we are trying to reduce the depth of field we need to increase the focal length. Therefore, the longest possible focal length...

Al

Margus Kivilaan June 20th, 2002 08:20 AM

yeah Al you are probably right
i''ve used wide-angle lens for doing background smaller and thus less significant, but never looked for focus issues. By using wide-angle can happen bad distortion of linearity which is also no good

Estonian is a weird language and difficult to learn to anybody so 'dont try this at home'
hope to meet you at IBC in Netherlands this autumn (are'nt you coming?)

Margus

Al Osmond June 20th, 2002 08:22 AM

Hey, why not? I will take any excuse to visit the Netherlands...

Al

Margus Kivilaan June 20th, 2002 08:41 AM

done-deal
let's contact somewhere in the beginning of September
Amsterdam is a really good place to visit!

Margus

Jeff Donald June 20th, 2002 02:44 PM

Hi,

Speaking of Laws of Physics and depth of field, one is, that 1/3 of DOF is in front of the focused subject and 2/3 is behind the subject. Once you have focused on your subject, and established the DOF limits, lock your focus and move the subject to the near or far limit without refocusing. Sometimes I control depth of field by not focusing on my subject, but rather a very near object or very far object. Then the subject is placed at the near or far edge of the DOF. This doesn't work for all scenes, but occasionaly gets me out of a jam.

Jeff

Rob Lohman June 21st, 2002 04:43 AM

So Al, with the equipment that I've got, what would be the
most ideal settings to achieve this? Fully zoom in? Zoom out?
f-stop settings?

Since I live in The Netherlands I'll definitely be attending IBC this
year. Would be great to hook up with some people overthere.

Al Osmond June 21st, 2002 07:15 AM

To reduce the depth of field, you need:

the longest focal length you can (zoom in)
the widest aperture (f/1.6).

(Note that you won't get f/1.6 when you are zoomed right in.)

You can shorten the shutter speed to force the aperture wider, or you can add more ND using filters. A polarising filter may help by reducing the amount of light getting to the chips. You're possibly already using -5dB gain.

All of the above assumes your subject matter will allow it. If you're shooting in a small space under the blazing Dutch mid-summer sunshine, you may not have much control over things. :-)

I've marked the dates of IBC on my wall-planner...

Al

Rob Lohman June 21st, 2002 07:44 AM

Thanks Al! I think I understand it now. If time permits I'll run
some more tests this weekend. It is all about testing first. I shot
some beautiful snow shot last december when I just received
my XL1S. It had some gorgeous depth of field in it. Thinking back
I probably had zoomed in quite a bit and ofcourse the lens
was pretty open with an ND filter on it due to heavy sun with
snow light.....

Perhaps things fall into place now. The most times you want
some short DOF is when having a character in the frame. This
might make it a bit ackward with a heavy zoom. Guess I'll
have to see how it all goes.

How about using a wide angle adapter with my standard 16x
lens? Would this help too?

Thanks again.

Al Osmond June 21st, 2002 07:48 AM

The wide angle adapter will increase the depth of field by shortening the focal length of the lens.

You're right about testing. You never have enough time to do all the tests you need.

Al

Margus Kivilaan June 21st, 2002 08:25 AM

to be honest i've never tried to achieve film-like look with video camera by myself, i will try it for sure
i don't know maybe you've already have read it..
http://www.dv.com/features/features_item.jhtml?LookupId=/dv/xml/feature/2001/billups0901
Scott Billups writes how he used handycam camera for making playstation commercial (directed by David Lynch btw.) He used Sony PD150, but i think it is'nt big difference between PD150 and XL1. Not much about technical details, but maybe this helps

Margus

Margus Kivilaan June 26th, 2002 05:29 AM

hey
i found accidentaly from deepest internet corner a description of add-on for DV cameras, which makes filmlike depth of focus. Costs a hell of lot, but very clever solution
http://www.pstechnik.de/pstechnik.htm products>digital film>mini35digital

Margus

Rob Lohman June 26th, 2002 07:17 AM

Margus, that product is widely known (check www.mini35.com).
We even have one user here on this forum using this rig!

Al Osmond June 26th, 2002 07:45 AM

Costs rather more than twice as much as a complete XL1s outfit, and once you've paid for it you can start saving up for a lens.

But when you've got it all up and running, every single one of your shots can then have an out of focus foreground, or background, or both, just like the Mini 35 demo video.

Why, one asks, does a dog lick his private parts?

I just wonder how often film camera users find themselves wishing for a little more depth of field...

Al

Adrian Douglas June 26th, 2002 07:53 AM

Rob,

to reduce the Depth of Field in video a common method is to shoot long. As Al said this increases the focal length and therefore reduces the depth of field.

This is the same reason that wide angle lenses don't require as much focusing as telephoto zooms, for example the 3x vs the 16x

Charles Papert June 26th, 2002 07:50 PM

Regarding your (maybe rhetorical) question, Al, we prize the shallow depth of field of film in almost all circumstances.

I shot a commercial this weekend on Super 16 that was originally supposed to be on 35mm. Film stocks and telecine have improved to the point where, for a television delivery, the gap between the formats has diminished greatly (especially in regards to grain structure), yet I was admittedly disappointed that we had to go to 16 for budgetary reasons, since it forced me into a different lighting design. When you don't have focus as a means to separate foreground and background, you have to work with other tools such as color and shade to achieve the desired effect. There are some tricks also, using smoke to haze the background of a set will give a shallower feel.

Instances where the minimal depth of field of 35mm is problematic: when working with long lenses, it can be tricky for the focus puller to nail it (usually when there is six inches of depth or less). We help them out by lighting to a higher level or removing ND if applicable to give them a fatter stop and thus more depth to work with.

I sometimes think about the tricks that were employed on "Citizen Kane" to achieve the deep focus shots on that film. I think the filmmakers would have been delighted to work with the massive depth of field available in the DV format!

Rob Lohman June 26th, 2002 11:45 PM

Thanks for all the responses, invaluable! I haven't had time to
test anything yet with this new data. And since I have some
party's to attend this weekend it probably won't happen for
another week. Charles, can you talk some more about the
Citizen Kane thing? What was their problem and how did they
solve it? Always interested in hearing such stories. Thanks.

Charles Papert June 27th, 2002 12:08 AM

Rob:

On "Citizen Kane", Orson Welles and his cameraman Gregg Tolland sought to achieve frames that had actors and elements staged at multiple distances but with all remaining in focus. Generally they achieved this by doing the opposite of the things described here to make the image more shallow; they used a lot of light (minimizing the size of the aperture), wide lenses, and in a few instances a split diopter, which is sort of like a large glass contact lens chopped in half. This lens (mounted like a filter) will allow for multiple planes of focus within a given frame. The catch is that the edge of the diopter will be seen in the shot as an abrupt transition and it needs to be buried by being placed over an existing line within the frame so it can't be noticed. There's a banquet scene in the movie that at one point features some ice sculptures which appear to be inches from the lens; and they are as sharp as the rest of the foreground and background, due to split diopters being brought in from either side.

A more elgant approach to hiding the transition is to use slant-focus lenses, also known as swing-and-tilt or swing-and-shift, which pivot the lens relative to the film plane. These were popular in commercials a few years back in achieving that look where one side of the frame was oddly out of focus. Also seen recently in the opening scene of "Swordfish" with Travolta in the coffee shop.

OK, my all-time favorite split diopter story...a few years back I worked on the movie "Office Space" which was directed by Mike Judge of Beavis & Butthead/King of the Hill fame. This was Mike's first live action movie, and he was always curious about the technology, so he would occasionally ask us things like "now, what do you guys mean when you say 'split diopter'? or "what are sticks, exactly?" We'd tell him, he'd listen and nod, and that was that. And then days later, an Entertainment Tonight camera crew would show up, and he'd ask them "OK? Are you guys rolling?". Then he'd lean over to our camera department , clear his throat and announce authoritatively "Alright guys, now I want you to go ahead and put the split diopter on the sticks for this setup..."

Funny guy.

Margus Kivilaan June 27th, 2002 01:26 AM

ooops!
 
<Margus, that product is widely known (check www.mini35.com).
We even have one user here on this forum using this rig!>

yeah now i noticed this stuff is been mentioned in this forum earlier also. It was new for me
sorry

Josh Bass June 27th, 2002 03:08 AM

Here's my two cents, for what it's worth (keep in mind I'm a moron). I bought the manual 16x lens recently, and while playing around with it, I found you get excellent shallow depth of field (that's what you want, right?) If you use the macro function on there. It's not very orthodox, but you don't have to get real far away and zoom in that way. Other than that, the other guys said everything. Not a cheap solution either.

Rob Lohman June 27th, 2002 08:09 AM

Hehe... thanks Charles! Fun to read. Margus, no problem. I was
just pointing it out to you, no need to appologise.

Bill Ravens June 27th, 2002 09:00 AM

Since DOF is intimately affected by the maximum aperture of the lens, that is to say the design diameter of the primary optic, I would suggest you consider the following. 35mm lenses are designed for an aperture consistent with 35mm film format. As such, the depth of field is LESS than for a 1/3 inch format(as in XL1s CCD)at the same iris setting(f/stop).

Therefore, if you're looking for a film look, one important maneuver is to use 35mm lenses with your XL1s. Use an EOS or FD lens adapter and a Canon 35mm lens. Unfortunately, you'll have to back away from your subject by a factor of 7, but, oh well.

Al Osmond June 27th, 2002 09:06 AM

I'll hire you a pair of handy-talkie radios, Rob, if you're backing off that far!

The builders say they can come along and knock a hole in the studio wall but they won't be here until next Wednesday morning...

Al

Charles Papert June 27th, 2002 11:02 AM

Hmmm...Bill...at an equivalent field of view, won't the depth of field characteristic remain the same? i.e. once you've taken the 7x magnifcation into account, and thus zoomed in your 1/3" lens to match the framing, you should end up with the same depth of field--hmmm?

Bill Ravens June 27th, 2002 11:06 AM

nope....it's a function of the overall light gathering ability of the lens...i.e. max diameter, even if you're only using a small central portion of the image plane.

Justin Chin June 27th, 2002 12:43 PM

My head's going to explode.

Charles, I think you're right.

If you put a 35mm lens on your XL1 with an EF adaptor you're basically only seeing the center part of the image, which effectively "increases" the circle of confusion. In order to match framing by backing up your camera, you're effectively defeating the purpose of the 35mm lens.

In general the limiting factor is the size of the film plane (ccd).

This is how I understand it. Generally by moving the camera and zooming in you flatten the image. This is not always desirable since it separates the viewer from the action.

Justin Chin June 27th, 2002 12:51 PM

If anyone on this thread hasn't seen this already. Here are some test images I shot with my Mini35 rig.

I used a 35mm cine (35mm focal length) lens at T1.4. The subjects at their nearest are just less than 3 feet.

As you can see I have less then 2 inches of focus.

http://www.monsterrocket.com/mini35/testimages/

Bill Ravens June 27th, 2002 12:58 PM

...and that is why 35mm lenses are so great. Even at that, 35mmx7=245mm. Great for portraits.....kind of compresses everything, tho'. Not to mention the film/sound stage needs to be REAL big....LOL.

Charles Papert June 27th, 2002 01:36 PM

Bill, this is making my head explode a little bit also.

I'm going to chat with an associate in the lens division at Panavision, because you do suggest an interesting theory, that the aperture has a different physical size on a 35mm format lens than a 1/3" format lens, to see if (given identical field of view) the circle of confusion is affected. Speaking of affected--this thread should properly read "What affects Depth of Field", eh?

Justin, your choice of a 35mm lens in your example has probably exploded a few other heads, as there may be a confusion between a 35mm focal length and the 35mm format...ever take a glance at your Zeiss speeds to check which lens you have up and catch your eye on the engraved "35mm" and get fooled into thinking that's the focal length when actually they all say that on the barrel? Hate that...!

Anyway, Bill, I'm not sure if we are all talking about the same thing--Justin's use of a 35mm focal length on his Mini-35 is yielded a true 35mm field of view and depth of field characteristic, which is very different than the same lens on an adaptor which turns it into a 245mm as you say. I would be interested in seeing a, say, 10mm lens on the Canon adaptor and comparing that to the 10mm setting on an the XL1 manual lens to see what the difference would be visually in terms of resolution, color, and depth of field etc.

My personal feeling about all this is that while it would be very nice to have the ability to control the depth of field in the 1/3" format, that characteristic is secondary to the choice of focal length. As you guys have pointed out it is not necessarily desirable or possible to be working at telephoto as it distances one from the action. There is a tremendous difference between a head and shoulders shot on a medium wide lens and a long lens. For an intimate scene between two people, shooting the closeups from 20 feet away may create more of a voyeuristic feel than being inside and privy to the conversation. Many filmmakers feel that comedy plays best on wider lenses, also. I would recommend thinking about blocking your action so that scenes play as far as possible from walls which will create that depth for you as well as make it easier to light the actors apart from the set.

Bill Ravens June 27th, 2002 01:43 PM

Charles...

LOL....yeah, i understand...boggles the mind. Most of my info comes from KODAK. If I can find the site, I'll post the URL.
Let me know what you find out. I work with optical physicists and they have supported what I have said...that depth of field goes inversely with the max design aperture.

Agreed that 35mm FL lenses and lenses designed for 35mm image plane are VERY confusing terminology.

Justin Chin June 27th, 2002 02:21 PM

Yes, the 35mm FL example is really confusing. But I wanted to illustrate the actual lens I used for the shot.

My lenses actually don't have their format etched in them. But so their is no confusion whatsoever I put yellow gaffers tape on the cap with the FL written on it. No need to read the small writing wrapped around the front of the lens AFTER you pull off the lens cap.

Charles, cool. I'd love to hear what the Panavision people have to say. The video lens numbers really start to confuse me. Math is not a strong asset of mine.

I always thought that the F stop (or the more accurate T stop) is the Focal Length divided by the size of the aperture. i.e. 100mm lens with a 50mm focal length is at f2. So a 25mm focal length (on a cinema 35mm format lens) is equivalent to 12mm focal length (on a 16mm format lens) the lens itself is smaller, hence the aperture is smaller... ahh... brain hurting...

Bill Ravens June 27th, 2002 02:33 PM

Justin...

You're right about the f-stop definition. And, it's pretty common knowledge that the DOF goes directly with the numerical T-stop(f-stop)...ie the smaller the f-number the less the DOF.

The discussion above refers to the MAXIMUM aperture(iris open all the way). On a lens designed for 35mm film, the diameter of the glass is larger than the glass diameter on a lens designed for a 1/3 inch ccd. My contention is that this also effects DOF.As Charles points out, a larger diameter lens has a larger circle of confusion. Therefore it has less depth of field.

Charles Papert June 27th, 2002 02:38 PM

Justin:

The t-stop is an adjusted version of the f-stop that measures the actual transmission (t) to take into account the design of the individual lens, to assure that it is delivering the correct amount of exposure to the imaging system (I was going to say film plane, but I am trying to keep in the spirit of where we are having this discussion!)

I think the more recent vintage of Speeds are the ones with the 35mm etched on them.

Justin Chin June 27th, 2002 02:46 PM

Charles:

As always you are correct on both points, my lenses are older, and the description of the T stop.

Okay, I must stop my circle of confusion.

John Steele July 15th, 2002 08:33 AM

Right OK, this is all hurting my head but I have a question. On an earlier post it was said a wide angle lens would increase the depth of field because it shortens the focal length. It was also said that to give a shallower DOF you would zoom in. So If I wanted to get a shallower DOF at a certain distance from a subject, but couldn't get far enough away to zoom in, could I use a wide angle lens that allowed full zoom, then zoom in and achieve a shallower DOF while being closer to the subject?????

I hope someone understood that question cause I don't think I did :-)

John.

Josh Bass July 15th, 2002 09:27 AM

I'm not sure I did either. Why does how far the camera is from the subject matter to you so much? The main thing is getting what you want in frame, and achieving your desired depth of field, or lack thereof.

Charles Papert July 15th, 2002 10:11 AM

Distance to the subject is one of the factors that affects depth of field. At a given focal length and aperture, focusing on a subject four feet away will result in a shallower image than if the subject was thirty feet away.

There's just one or two variables missing from the question which will help for us to provide an answer. Why exactly can't you "get far enough away to zoom in"? Is it because you are within the minimum distance that the lens you are using will focus (let's ignore the macro setting for the moment)? In other words, when you zoom in, you cannot get the lens to focus because the subject is too close?

If this is the case, it is possible that another lens may have a different close focus characteristic. On the manual lenses, it's easy to see what this is: turn the barrel until it stops, and look at the inscribed number (usually around 3-5 feet). That is the minimum distance that a subject can be from the camera to allow full use of the zoom and retain focus the whole way. I have not seen published stats on the various Canon lenses to see if the wide angle zoom offers a closer minimum focus, or if using the zoom-thru wide angle adaptors alter this range also.

There's a certain amount of academic discussion involved here that needs to be tempered with the fact that DV inherently has massive depth of field, and thus it is tricky to talk about how to minimize it at the wide angle. Suffice to say that if you are inside minimum focus (i.e. the subject is within that minimum distance that the lens can focus), it will appear soft at a telephoto setting but gradually come back into focus when zooming out. This essentially is because a wide angle lens setting has more depth of field than at telephoto, and is able to appear sharp enough even when the subject is too close (the wonderful phrase "circle of confusion" enters into the discussion, but I won't push it here). I would have to say that if you are too close for the standard lens to focus, there aren't any lenses out there that will cheat the minimum focus close enough to give you the effect you are looking for, considering the inherent depth of DV.

OK, so what am I saying? That a telephoto setting has less depth of field than wide angle, so it should produce a shallower image--but also that the closer that the subject gets to the camera, again the less depth of field you have and thus shouldn't the subject be able to get two inches away (where you can't use the telephoto end of the lenses any more) and result in a really shallow focus? Yes, and that's where you can use the macro setting on the lens as was pointed out in a recent thread. It will only work at the wide angle setting, but watch as the background goes out of soft as you focus on the subject's nose or whatever!

This makes all of our heads hurt. If it makes you feel better, I can tell you that in the film world the camera assistants carry around charts or Palm programs that calculate depth of field given all the variables, so that when the DP asks for a specific amount of the depth they can look it up (as well as figure out how much depth they have to play with to maintain focus, sometimes only a few inches).

Hope this helps--somehow!

Rob Lohman July 15th, 2002 03:21 PM

Thanks Charles... you are really giving me a lot of information here
and I seem to understand it better and better. Great! Thanks a
lot.

I did some testing (finally) when partially and fully zoomed and
and I got some lovely shots with a shallow DOF... beautiful!
Only problem is that this will not work in a small space (that
zooming in part)... perhaps I should hunt down my macro
button.... didn't know I had one.

If I can find some time this week I will put up some results
of my DOF test. I'll include settings used.


All times are GMT -6. The time now is 05:33 AM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2021 The Digital Video Information Network