Focus on the person's eyes. The eyes--the reflections in them--are very sharp!

Great hint for the studio. For those of us that must "run 'n gun," where dozens of exposure, composition, lighting, and focus decisions must be made very minute, it unfortunately is a luxury we just don't have. I still rely on what I learned back with my first 35mm camera back in 1959: small aperture and then focus in the middle of the range of "in-focus." I know some have said that you have to keep HDV camera nearly wide open to get the best focus and, again, in studio conditions that may be something you can argue about. However, f16 is your friend when everything is happening at once, and all hell is likely to break loose at a moment's notice.

I'm not sure why people would say that wide open is the way to get the best focus.

Happily, in sunlight you can run with a tight aperture and get more depth of field, which is helpful for doc-style shooting. Shooting wide open, though, gives you less depth, and a bit more cinematic feel because backgrounds can go soft.

Generally, with depth of field, you get about one third of your depth in front of the focus point and two thirds of it behind.

Usually, it's thought that a lens gets it's best performance around the middle of the aperture range, say f5.6. This is because you are using a portion of the glass that is more likely to have the lens's best resolving power. Shooting wide open, you use more of the glass and are affected by more variations in the lens.

With HD, you may find that you have to revert to the practices of real cinematographers and use a tape measure to get your focus marks. But printing up a target for the subject to hold sound's like a good way to go too. Assuming your lens stays in focus when you zoom out, of course. I was pretty shocked that all three lenses for our Nikon D70 still camera didn't hold focus when you zoomed in and out.

Of course, color viewfinders suck for focusing anyway, but if they're SD and you're shooting HD you may as well just give up on the viewfinder altogether.

Rob Mack

John, I have to disagree with you. My 35 years of experience in shooting stills, film and video have proven otherwise.

You all must have better vision than I do. I put the Z1 in auto so I don't have to worry about the gobs of sweat blurring up the eyepiece.

JJK

Great tip however Z1s seem to have a known back focus issue. That means the focus shifts as you zoom!
If you find you do have this problem get Sony to fix it.

Bob.

Ha!

I used to see eye cups with little tubes running off them to let the vapor out or air in.

You're far from alone with that problem.

Rob Mack

>>I'm not sure why people would say that wide open is the way to get the best focus.

diffraction

more detail
general chat

>>I'm not sure why people would say that wide open is the way to get the best focus.

I was just quoting what I have heard others say. Here's a link to one such post:

Z1 focus issue

That's really good information. seems like the technical solution would be bigger chips. That'd yield shallower depth of field and less diffraction.

I was always amazed when I saw things shot in 35 vs 16. This was stuff I'd been present for the shooting of, by a good friend. All music videos, so the material was always similar. The 35 always looked so much sharper and saturated, and perhaps diffraction was playing a part in it.

The mid-fstop argument, which I think I probably picked up from the american cinematographer manual, claims that wide open apertures suffer from using light from the less perfect outer areas of the lens. Or that's the way I remember it.

It sounds to me like the argument worked well for 35 with it's large target area but that HD imagers are at a big disadvantage because of their smaller size. You either get diffraction or you get lens degradation from shooting wide open. The question would be which is worse. Maybe an HD imager couldn't resolve the lens difference to begin with?

The other question might be whether you can even tell the difference? The upshot of it is that colors in the lower wavelengths will tend to bleed from line to line of the imager and generally soften focus a little. Maybe you'd end up with SD-like resolution in your reds.

This isn't the same as depth of field, but rather a reduction of the lens's resolving power. And the issue is how to accurately grab focus when your viewfinder can't resolve a full HD image. Given that even a 2/3" target is pretty small, and you get a fairly deep depth of field anyway, I'd think that f16 might be a lot more than you'd need to cover your butt.

Rob Mack

There wasn't much in that thread. Certainly no advice to shoot wide open, but the post could easily go along with Serena's links about diffraction limits-that a len's resolving power is reduced as the aperture tightens, and that it can reduce below the resolving power of the chip recording the image.

It's a subjective thing though. For hand-held work one might feel that it's better to shoot with a lot of depth of field and have a good chance of being in focus even if you lose a little resolution. After all, soft focus is a hell of a lot worse than what diffraction is going to do to you.

Rob Mack

There are many compromises made in the manufacture of lower end video cameras (which includes ENG). There are many design compromises to get our zoom lenses down to affordable prices and all we can do is to be aware and to work with them to get best results under our particular circumstances. Everyone should undertake quantitative testing of their own gear rather than assuming that generic data is representative.

Some interesting stuff I read some time ago over at Luminous Landscapes.
The apparent DOF is affected by resolution. In other words the DOF will look different at low res (say SD TV) than it will on say the silver screen, all else being equal i.e. from the same film or video. Thing is there's only one point that's in focus. Anywhere before or after that is out of focus.

Except we don't see that if the circle of confusion is less than the resolving power of the system including our eyes.
So using a small low res monitor to judge focus when using a hi res imaging system with good glass really seems to be asking for trouble as the DOF may appear much greater than it is, for example the whole face may appear 'in focus' whereas the DOF is much shallower and the point of focus is on the ears.

I guess there's also a reverse side to this. What appears as a subtle shallow DOF on the big screen might be totally lost on the average TV audience.

Bob.

That's my biggest pet peeve about either DV or HDV: The lame viewfinders. If you've ever looked into the viewfinder of a Betacam (or similar camera), the clarity is astounding. A full 2 inch BW CRT. Not cheap at $2000-$3000. Yet all MiniDV cams, and most low-end HDV's use a LCD viewfinder and/or screen. I'm almost never sure I'm in perfect focus because the viewfinder/flip-out screen is subpar for optimum focus.

Bob, you have mentioned circle of confusion and described it accurately. As you say, the diameter assumed determines theoretical DOF. How is that diameter determined? By viewing conditions. The old rule involved viewing an 8 x 10 inch print at normal reading distance. Put the print further away and the CoC is bigger (because of the resolution limits of normal eyesight) and bring the print closer (eg using a magnifying glass and the CoC needs to be smaller. So the angle the image subtends at your eye is the basis of the rule. Just to complete your description (from a much earlier thread):

Depth of field is a function of lens focal length, aperture and diameter of the Circle of Confusion (Cc). The Cc represents the diameter of an out-of-focus image of a point of light that we perceive as being "in focus". The smaller the viewed image the larger Cc needs to be before we observe "not focussed". Similarly the larger the "negative" the larger can be the Cc for the viewed image size. So the Cc is related to the magnification of the viewed image. For cinema projection of 35mm film the generally accepted Cc diameter is 0.025mm, and for 16mm it is 0.015mm. It will be smaller for a 1/3 inch CCD and will depend on the viewing setup (monitor or projection).
So for all video cameras with 1/3 inch CCDs with lens of same set focal length and same aperture, the depth of field is the same. Whether zoom or prime, doesn't matter.
DOF calculations are based on the hyperfocal distance H (when focussed at the H everything is perceived as in focus from half the H to infinity),
H= F*F/(f*Cc), where F = focal length, f=f/stop

so then for any focussed distance S:
DN= H*S/(H+(S-F)) -- near limit
DF= H*S/(H-(S-F)) -- far limit

DOF = DF-DN

earlier thread

So for all video cameras with 1/3 inch CCDs with lens of same set focal length and same aperture, the depth of field is the same. Whether zoom or prime, doesn't matter.

=====================================================

Well yes, that holds true for Cc size as presented to the imager however the effect of that depends on the resolution of the imager surely. If the Cc is smaller than the resolution of the imager then it's irrelevant, all points with a Cc smaller than the resolution of the imager will appear equally in focus.

I'm assuming we're viewing this on a large enough screen so that the resolution of the system is relevant.

Bob.

Bob, that's an interesting point for consideration. The CoC is specified as the largest circle that a viewer will judge to be a sharp point under the specified viewing conditions. If the system can't achieve that resolution then the viewer will judge that no point is in sharp focus. Obviously you can determine viewing conditions under which any fuzzy point looks sharp just by moving further away from a fixed size image (or making the image smaller for a fixed distance). Any stated dimension of the CoC assumes a specific set of viewing parameters and the standards stated are related to standard conditions.
I might say, looking at DV on a 2 metre screen, that no point is in focus and go on to say that DOF is therefore zero. Other people will disagree because they accept that standard of definition and therefore a certain variation from that is acceptable to them. In other words that the error in focus has to be of certain magnitude before they can see any increase in fuzziness. Of course I wouldn't make any such comment on DV.

BACKFOCUS??
Not having the ability to have an HD monitor on my shoot recently I statrted wondering if I was doing something wrong.
Normaly I do the same Zoom in focus then back out to frame the shot.
Things have seemed really soft lately.
Can some of you who are patient enough to download a 80mb M2t file. Look at this and tell me if my Fx-1 may have a back focus problem.
This shot is a zoom in ffor sharp focus then out(all the way) notice how the lettering and such goes soft fast.
Backfocus.m2t

I'm downloading it but the electrons travel slowly across the Pacific.
However it might not reveal much because as you zoom out things will naturally appear to go a little softer.
I think the test that only you can do is fairly simple though. Set the camera to manual everything. Zoom right in a say a resolution chart or any sort of grid etc. Set the focus. Zoom back out.

Now if you can make the thing you were focussed on sharper by adjusting the focus ring then you have a back focus problem.
You'll probably need a decent large monitor to do this test but if it's a Z1 you might get by with the focus assist.

To fix this in a fixed lens camera does mean a trip to the hospital.
The other thing is that temperature can affect the back focus setting.

Bob.

"Back Focus" refers the distance between the rear element of a lens and the camera focal plane. If this is incorrectly set then a good zoom lens won't hold focus as the lens is zoomed.
a test procedure

Why did I say "good zoom lens". A cheap zoom very likely cannot maintain a constant focus plane. However a further point is that generally lenses on lower end cameras (including ENG) are not truly described as "zoom" but as "varifocal". The difference is in the mechanical complexity of the lens, the varifocal lenses relying on electronic control to maintain focus. To pinch from a CML discussion:

"I think the telling point here is that ZOOM lenses shouldn't need regular back focusing. However prosumer cameras don't have ZOOM lenses but rather VARI-FOCAL lenses, which are ELECTRONICALLY controlled to resemble zoom lenses.

[Short optical primer for those who don't understand the difference: Both types of lens enable the operator to adjust the magnification of the shot without re-locating either the camera or scene. A zoom lens will keep the desired object in focus throughout its range. A vari-focal lens does NOT "track" focus as its magnification is adjusted. Zoom lenses have, typically, twice as many moving parts / lens elements as a vari-focal lens. Until recently, vari-focal lenses were used only on surveillance cameras - in set-and-forget situations.]

So prosumer cameras have electronic circuits and motors to simultaneously adjust the magnification and focus elements as the operator presses the "zoom" control, to simulate a zoom lens. Seems entirely feasible to me that the electronics could be out of adjustment in some examples of a mass-produced item! "

So here the argument is that if you determine that your focus is breathing as you zoom, probably the electronics is out of adjustment rather than the physical position (or back focus) being incorrect. An interesting assertion.

I do the same thing with alot of various cameras I use. I also have a color chart on the card that I've printed. This gives me a good color reference too.

Looking at that clip I'd say something is off. I assume that's in manual focus.

Might be worth getting it checked out by Sony. I recall it's a known problem with Z1 and FX1 that Sony will fix gratis.

Bob.