long ago i realised talent, even when seated have a tendancy to move around quite a bit, especially so in a longer type of interview. it was then i realised that shallow dof though nice wasn't much use if the talent was out of focus.

unfortunately though musicvid adage is good advice, i find for everyday situations it's always safer to give your talent the maximum 'lean room' as possible......

I think what Laurence is referring to is this:

Does the field of focus extend further behind than in front of the actual point of focus. My understanding is yes, it does. If I'm correct then also it would be better to focus on the front of the interviewees chair than the back, the tricky part is knowing which limit of the field of focus is the front of the chair at.

For stage shoots I try to find something mid stage and rack focus back and forth to find the limit and then set the dial to a point between those points.

I'd also agree with Laurence, more DOF is good, especially for faces and especially if the camera is applying any edge enhancement.

Bob.

Yes, that's a good lesson. If you do not have control over the back and forth movements of the performer, avoid too shallow a depth of field.

Pretty much just adding to the posts already, its all about DOF control. And you (and all of us) being in control of it.

We just went through a 'style' with DSLRs going shallow DOF crazy, mainly from internet 'gurus'. Shooting at 1.2 etc.. But its not always appropriate. Most people are now realizing its about capturing the feeling of the moment accurately , not just 'get it as shallow as possible'. Im personally glad this is going out of fashion!

Avoid f1.8 if you need head movement back and fourth. Its true to always aim for the eyes - thats the most crucial facial feature, same as in still photography. But controlling the DOF is more important for video/film because our subjects do move!

Giving yourself a working range, front to back distance where the talent can move without effecting focus too much is the trick. This can mean f2.8 f4 f5.6 or even higher. And of course its effected by the camera to subject distance too.
If you are getting grain at 2.8, presumably having to ramp up the ISO, then its a sign you are hitting the limit of the camera sensor. You would need more light to compensate. Or run your grainy footage through Neat de-noiser.
Something has to give, stop down, more light, reduce grain.. Its a mixture of those things and experience to nail it every time.
Another trick is to focus 1/3 into the subject range, not quite half way. That's the optimum point for maximum range within the field. 1/3 in from front, 2/3 back.
Some simple experiments at home can illustrate the working range you get at different f stops.

Paul.

"Does the field of focus extend further behind than in front of the actual point of focus. My understanding is yes, it does."

That's correct, but not by a lot. I guess one sense is that as a portrait shoot or video interview progresses, the subject tends to relax and the neck and shoulders lean back. If the head and upper body come forward, the interview may be becoming uncomfortable for the subject.

You'll see in a lot of vintage portraits, the nose is actually slightly out of focus, and this was considered a valid photographer's technique during the mid-century (airbrushing notwithstanding).


Here's a quick reference for still photographers. We used to use complicated dial calculators that never quite worked ;?)
http://www.bobatkins.com/photography/technical/dofcalc.html

Determining actual depth of focus, especially in motion picture and videography is complicated; much more than aperture and lens focal length. However, reducing focal length or zoom is still the second most effective way of increasing DOF, when the aperture is at or near full open.

For really close/shallow interview pictures, it probably wouldn't hurt to have two cameras, one wider that you could always cut to as needed. The level of closeness you're describing would seem to be a little too close to me for a long interview. Give the person some breathing room. :)

I know I saw a chart somewhere showing that there is a lot more d.o.f. focus behind the focal point than in front (maybe 2x? 3x?) but that was probably in a book from long ago...

Musicvid said:
"We used to use complicated dial calculators that never quite worked "

Like the markings on the lens barrel?
Cameras such as the EX1 provide the same kind of information in the viewfinder and it can indeed not quite work. No doubt the calculations the camera is using are valid based on what the lens is set at. What I'm now 99% certain confounds things is the Detail / Edge enhancement logic. Turning Detail Off seems to avoid the problem, even turning Detail way down can still leave some artifacts going on.

The whole "detail" thing in video cameras also seems to me to cause other visual artifacts. If the subject moves the natural motion blur can also stop the detail logic from adding sharpness to the image. Maybe my noticing this brands me as a "pixel peeper", certainly once you start noticing these things it's hard to stop looking for them. I now generally turn Detail Off.


One other thing worth a mention is that all but the most expensive lenses don't perform to their best wide open. There's some exceptions to this such as primes specifically designed to be used wide open. I think it pays to "know" your lens.

Bob.

can't keep the subject from moving, or most subjects.. I just think maybe a bit too tight, you can always zoom some in post ..why HD is in some ways forgiving (to me) give some leeway and tighten a bit in post

"Does the field of focus extend further behind than in front of the actual point of focus. My understanding is yes, it does."

Twice as much is not a lot?

As others have said .. depth of field extends 1/3 in front of the focus point and 2/3 behind. Exacty HOW much depends on the f-stop but the ratio is always the same.

There are on-line DOF calculators that you can study before a shoot, for example:

http://www.dofmaster.com/dofjs.html

"As others have said .. depth of field extends 1/3 in front of the focus point and 2/3 behind. Exacty HOW much depends on the f-stop but the ratio is always the same."

Sorry, but this is factitious myth, and is 100% incorrect. The front / rear ratio is always different, and ranges all the way from ~1:1 to 1:inf. You are easily not the first person I have heard repeat 1:2 as if it was a monolith (there are none in photography).

At the distances, focal lengths, and apertures commonly used for portraits and video interviews (the only thing we're talking about here), the front/rear ratio very nearly approaches 1:1 (front and rear depths of focus are essentially the same). People with real experience in either discipline already know this to be true. Thus the statement, "That's correct, but not by a lot" is an entirely fair and responsible representation of fact in the context of this discussion.

The one thing that is not reasonable is the presumption that this is a linear relationship. It is anything but linear. In all fairness, I was taught at age 11 that a normal focal length lens (relative to 6x9 film format) at 10 feet and f16 is "about" 1/3 front, 2/3 rear. That rule of thumb remains true in this century, to the best of my memory.

Here is what I have believed to be true for the last fifty years, as explained by Wikipedia:

Near:far distribution

One area where the lens markings "are" useful (except with some poorly made telephoto lenses that may never quite focus at infinity), is in determining the hyperfocal distance. Setting the focal point to where it just includes the infinity mark in the dof range relative to a particular f-stop, is a reasonably good indicator of the near hyperfocal distance. Sorry for allowing myself to be taken so far off-topic in an otherwise good discussion.
;?)

OK, trying to get back on topic. A traditional still photographer would use a 75-90 mm at maybe f/2.8-f/4 and back away from the subject. I don't know how this relates to a prime (85?) on a Mark III 5D but it would give you a different starting point. The slightly longer lens will give a 1:1 DOF ratio for all practical purposes, and a more pleasing rendition of facial features, by most accounts.

For an APS-C DSLR the crop factor is typically 1.6, so to get the equivalent of 85 mm focal length on a full frame or 35 mm camera you would need to use 85/1.6 = 53 mm.

A lot has to do with the distance between the subject & the lens. If the subject is close, use 2.8. If the subject is far, use 1.8. Shallow DOF increases when the subject is near to the lens & decreases when the subject is further away. That is why I select my aperture based on this variable.

Using the DOF calculator I referenced above, then a full frame DSLR or 35 mm camera with a 85 mm f4 lens focused on a subject 2 m away would have a depth of field of 206.6-193.8=12.9 cm.

An APS-C camera with a 53 mm f4 lens would give a similar field of view, but the depth of field would be 211.7-189.5=22.2 cm.

Thus the smaller the sensor, the greater the depth of field, with all other factors equivalent.

In working with head and shoulders or seated subjects, it is the subject that dictates the shooting distance relative to focal length, all else being equal. Just a different way of looking at it. One wants to fill the frame, but an 85 at 2 meters would be a little tight in some situations, esp. an interview, where the first rule still is that the subject is going to move around.

I was watching Kramer vs. Kramer on Blu-ray the other day. There is a pivotal scene in a coffee shop where Meryl Streep says she wants her kid back. For almost the whole scene, Dustim Hoffman's close-up is out of focus. When he briefly leans forward a couple of times, his face is back in focus.

My guess is that in '79, they didn't have video assist for the focus puller to monitor. They probably did it the old-fashioned tape measure way, but somehow Hoffman moved his chair. It's just weird to see that on a major motion picture. At least with us and video, there shouldn't be any excuse to monitor the focus and adjust accordingly.

Still no excuse for such a movie. They should have shot the scene again.

This is my attempt at shooting interviews w/ SDOF. My workaround for when the interviewee moved out of focus was cutting away to B-Roll (which may not work in your case) You'll also notice that some of the shots are 'More Shallow' than others. This was because I was attempting to compensate for the interviewees movement and trying to get a larger range to suite their movement. It's a lot easier when you have a loooong background environment behind them that will stay out of focus as you adjust to find your sweet spot.

]

I'm not as Technical as this discussion. Instead of going by numbers & settings, I tend to go by 'The Look'. Once I have 'The Look' I want, I lock it down. I'm not even concerned with what the settings are... unless I'm trying to match multiple cameras.

Some very basic research here, back to school:

def 1) The hyperfocal distance is the closest distance at which a lens can be focused while keeping objects at infinity acceptably sharp. When the lens is focused at this distance, all objects at distances from half of the hyperfocal distance out to infinity will be acceptably sharp

def 2)The hyperfocal distance is the distance beyond which all objects are acceptably sharp, for a lens focused at infinity.

def1 and def2 pretty much give the same result so choose any it doesn't matter.

wiki:
"The DOF beyond the subject is always greater than the DOF in front of the subject. When the subject is at the hyperfocal distance or beyond, the far DOF is infinite, so the ratio is 1:1 as the subject distance decreases, near:far DOF ratio increases, approaching unity at high magnification. For large apertures at typical portrait distances, the ratio is still close to 1:1. The oft-cited rule that 1/3 of the DOF is in front of the subject and 2/3 is beyond (a 1:2 ratio) is true only when the subject distance is 1/3 the hyperfocal distance."

"true only when the subject distance is 1/3 the hyperfocal distance", which is pretty much most of the time if your shooting an interview in a room. Although it may well be something closer to 1:1 as in the example of a wide aperture portrait.

"true only when the subject distance is 1/3 the hyperfocal distance" - that hardly makes it 100% incorrect now does it!

And in the practical (real) photography world its an accepted and reliable method of setting DOF range within a shot - without the need for calculators, charts or anything else.
Too many numbers!!! this really isnt as complicated as this thread is starting to show. The 1/3 rule is excellent in most practical cases, I, and thousands of other photographers are not interested in theoretical situations that quite honestly don't even matter in the real world. We need to get the job done - not procrastinate over the values!

These days with video: use an EVF - see the range with your own eyes. Now how complicated is that?

Paul.

Nice video by the way Evo, i like the moving shots in particular and pace of edit. And the shallow DOF is not 'too' shallow, which was a point i trying to make earlier. Too many ultra shallow shots these days - because we can doesn't mean we should. I am personally guilty of this in my own videos, but i have now made a conscious decision to stop this nasty habit and just get the shot right - with the right amount of DOF to suit the story.

Paul.

"

Depending on your camera, quite complicated. How many cameras come with 1920x1080 EVFs, I think none, certainly none that I'll ever afford. That's why if I know focussing is going to be a big issue I take a monitor with pan / scan pixel to pixel and get my eyeballs very close to it.

Anyway, I'm glad the whole shallow DOF fad is on the wane. Working for a rental company from time to time one gets to see footage left on cards. We went though all the shots taken at one wedding and not one was focussed on the subject. Maybe the couple got lucky and it was the B or C camera.

The next big thing will be aerial shots, now that I can go along with.

Bob.