FS7: Fujinon announce MK 18-55 T2.9 and MK 50-135 T2.9

[JPNola]

What you see on a screen is something that already happened. Since you are pulling focus based on that then indeed you are certainly late. Yes, it does not always matter, but you are late.

Then every video camera operator who is operating using an eyepiece is "late". Only with an optical eyepiece on a film camera ( or digital reflex camera shooting stills ) is the operator seeing "real time".

Delay on an EVF is not more than 1/120 or 0.008333 second. The human delay- lag time between observation and muscle movement in response, is far great than that. The human lag is on the order of hundreds of milliseconds. So if the delay using an OVF is .1 seconds the delay using an EVF would be .108 seconds

Remote monitors fed via camera outputs can have more of a delay. ( which is why often times wireless video is used instead of a hard-line, wireless video having less latency ). It depends on the camera. One of the early RED camera models was known for an intolerable delay on its SDI out.

Jib operators for sporting events and concerts are pulling focus from the monitor. Those shots are typically wides, but there are times where they have to pull focus on long-lens tights.

The technology has greatly reduced latency. It used to be more of an issue and thus the "never pull from a monitor" maxim used to have more merit than it does today. There is great value in being able to actually see the focus. Not just see the image ( framing, subject ) but SEE whether or not you are in sharp focus. You really cannot find "critical focus" without using a monitor. The more shallow the DoF and the more critical the focus, the more an AC will use the monitor. If you pull focus sans monitor, you are just best-guessing as to the focus.

How did you guys get on this topic, anyway?? :happy:

 

[broughtonfilm]

Most focus pulling at 'good AC' level is pre-emptive or proactive rather than reactive.

Now you are talking about an AC who is also a fortune teller - because that's what he / she would need to be in order to be proactive in a situation where (as JP suggests) they can't even see a set. This means they can't anticipate an actor walking into a frame by seeing outside of what the camera sees by looking at imaginary boundaries of camera's view, they can only react to an actor already entering a frame on their monitor. Hundreds of other examples like that, not being able to understand the dynamic of a scene, not being able to see reactions of actors outside of a frame, all of which help dramatically in being proactive. I remind you, we are talking about "exclusively" by a monitor, AC not even looking at a real set, often remote from it.

 

[broughtonfilm]

Then every video camera operator who is operating using an eyepiece is "late". Only with an optical eyepiece on a film camera is the operator seeing "real time".

Delay on an EVF is not more than 1/120 or 0.008333 second. The human delay- lag time between observation and muscle movement in response, is far great than that. The human lag is on the order of hundreds of milliseconds.

Delay on a video monitor connected through a 25 ft video cable would be slightly greater, but we are still talking about an insignificant amount of time.

Do you really don't understand I'm not talking about video signal latency?

 

[Ben Scott]

Now you are talking about an AC who is also a fortune teller - because that's what he / she would need to be in order to be proactive in a situation where (as JP suggests) they can't even see a set. This means they can't anticipate an actor walking into a frame by seeing outside of what the camera sees by looking at imaginary boundaries of camera's view, they can only react to an actor already entering a frame on their monitor. Hundreds of other examples like that, not being able to understand the dynamic of a scene, not being able to see reactions of actors outside of a frame, all of which help dramatically in being proactive. I remind you, we are talking about "exclusively" by a monitor, AC not even looking at a real set, often remote from it.

This happens all the time.

Clearly you go set marks for actors and rehearse the action beforehand and you've set those marks on the FF but you still need to be able to read people.

ACs pulling with monitors tend to be by the side of the camera or off to one side.

 

[broughtonfilm]

This happens all the time.

Clearly you go set marks for actors and rehearse the action beforehand and you've set those marks on the FF but you still need to be able to read people.

ACs pulling with monitors tend to be by the side of the camera or off to one side.

Well, exactly. You said yourself that it's about reading actors, reading cues. Problem is they often happen outside of frame, for example camera pans into something that happened outside of frame. If you are restricted to only seeing what camera sees and can's see outside of it (JP suggests that these AC's don't even need to see actors, he does not need to worry he is blocking the view) then you can't be proactive.

 

[Ben Scott]

In fairness there are tons of multimillion dollar movies being made with ACs pulling off monitors, whether they are camera side or in a dedicated area not in view of the set. A lot of them won't ever pull a tape out of the bag.

But yes you still need to be on site to set and rehearse and get a feel for how on-point actors are with their marks etc.

Also, to aid this way of working a lot of the monitors - especially if it's on Red or similar will also have lookaround function with an area outside the active frame also visible.

 

[JPNola]

broughtonfilm- You like to argue and you like to be right, don't you?? That's okay, I do as well!

dude, I'm telling you that I have seen many AC's pull exclusively from a monitor for most scenes. They don't pull focus exclusively from a monitor for every scene, but there are scenes where they DO pull exclusively from a monitor and the latency isn't significant enough to be a concern. We know about video latency. But there is always latency, even when pulling by watching the scene directly. There is latency between the movement of the subject and the AC's adjustment of the focus. The AC is almost always REACTING. Even when the AC knows that an actor is going to move on a cue or on a line of script, the AC is still reacting to the actual movement that takes place. So any video monitor latency is in addition to the latency of focus adjustment, a latency that is far greater than the monitor latency. It's like Latency + minute latency. A latency factor of 10 + .1 additional latency.

When the shot is a long lens closeup of an actor, the monitor is the only way for the AC to know if they are hitting critical focus. In most scenes the AC knows from the rehearsal and from the design of the scene what is going to take place in the scene. On that closeup of an actor, the AC knows in advance that the scene is only the closeup of that actor and there isn't going to be a need to rack focus to anything else in the scene. Marks are not of significance because the actor is standing still the entire time, began the scene on the mark, and is not moving from the mark or to a mark. So the ONLY thing the AC has to contend for in the scene is the minute movements of the actor's head and critical focus for that movement can only be determined using a monitor. When all useful information for pulling focus is found on the monitor and only on the monitor, it makes perfect sense to pull exclusively from the monitor. Pana-Tape can't tell the AC if they are in critical focus. Marks don't tell the AC if they are in critical focus. All the measurements in the world made during rehearsal don't tell the AC if critical focus is being achieved.

AC's can anticipate an action, but they are still reacting in pulling focus most of the time. Rarely do they pull focus to a point in space only for something to later move into that space or move into that focal point. The AC knows an actor is going to go from sitting forward in a chair to leaning back, but most often the AC still follows the actor's movement as opposed to pre-racking focus to a focal point and then having the actor "fall" into that focus.

There is a reason they call the device "Follow focus" and not "lead focus". :laugh:

Marks for actors / pre-measurements to those marks or to objects in a scene and the corresponding marks on the focus wheel guide the AC to a general focus area but they don't lead the AC to critical focus. Only a monitor does that.

The statement made by the AC at that workshop was the type of statement made at workshops- a heavy dose of ego and over dramatization there. It sounds to me as if it was one of those arrogant "ah, these punks today..." type of statements. And it is one of those things that used to be more true but for which technological advancement has made less true. Such as saying that no professional shooting video ever used auto focus. But AF improved. Just as electronic latency improved ( diminished ). That's what makes "old guy" stuff "old guy" stuff. Like an old guy who says you should never use LED lights. Or a really old guy who says that any good DP operates camera because that is the only way one can see the frame.

Let me ask you this- when you shoot, do you keep your other eye open at all times? ( I shoot that way sometimes, but not often. ) I am confident that most of the time you are pulling focus as an operator you are pulling focus from a latent image. But you are still finding focus at a tolerable speed.

 

[Ben Scott]

It reminds me a bit of when a makeup artist on a shot I was DPing on Red was telling me how awful Reds are, despite it not being his discipline.

To be honest if he's a steadicam operator moving through a scene it's even harder for someone to be pulling focus just off watching the scene and observing marks and pulling off a monitor is an even better way of doing it.

 

[Ben Scott]

So.... About them Fujis

 

[Donny 123]

Just a last comment.. re the old guy stuff.. auto focus is not pulling focus.. pulling focus is a human.. ok well anyway an AC.. .. deciding where the focus will be.. in a frame.. it could be a tiny % of the frame.. auto focus is a machine deciding where to put focus derived by some sort of parameters .. if you put the term "out of sight ,out of mind " ..into a machine.. it will give you the answer .. invisible idiot... auto focus if it works well enough could be good for some shots.. but most they would be wides anyway ,otherwise there will always be some risk of focus going where you dont want it.. unless you have AI.. and discuss with the machine the early works of Raoul Coutard.. and the true meaning of cinematic.. and if the freeze frame is justified in the end of Les Quatre cents coups

 

[morgan_moore]

In stills I have been AF since 1998 or suchlike after having my ass whipped at an international sporting event by someone who had gone to the dark side.

So in one media I have used AF for a while.. but it is still human, selecting a zone, going manual, knowing that backlight messes it up, knowing when a locked focus is safer than an unexpended focus dive by the camera

The human still uses skill to drive the machine.

 

[Ben Scott]

Sorry but stills and moving images are simply not comparable in this regard. Focus in narrative work is an intricate ballet, sometimes leading and sometimes being led.

 

[morgan_moore]

Sorry but stills and moving images are simply not comparable in this regard. Focus in narrative work is an intricate ballet, sometimes leading and sometimes being led.

Both are built using technology to get the result you want. Do you use a gimbal? Or only a 'proper' human controlled steadicam. A light meter? Or only proper manual eyeballing?

I freely admit to using 99% manual focus. And agreeing that the current technical offerings are in the main useless.. as they cannot lead or be led.

Hence a desire for a nice manual lens in the 18-70 range!

 

[Ben Scott]

Until you can use autofocus to stay on one object, change the object you want to track, then go back to the first object before pulling to something else entirely at the end of the scene all in a graceful way that is not too fast and not too slow and is just right and matches the pacing of the action on screen and doesn't miss a single focus mark, it has no place in narrative cinema.

Before someone chimes in to say you could programme this - that's like programming a cnc router to slice a carrot and there still no guarantee it will be perfect. Just slice the bloody thing yourself and it's done.

 

[stokstad]

You would only use AF when it made sense. To program it in complicated ways feels like the wrong idea. To be able to track a face running/walking towards the camera for example, would be a real benefit of AF.

 

[Ben Scott]

So it's a tech you could only use on occasion, which means you'd have a focus puller anyway, so why not just use a focus puller?

Do you know of any theatrically released movies that have used AF on a shot as you describe?

 

[morgan_moore]

Ben - this thread is sort of about a pair of lenses that to my eye are for the solo OMB operator - so maybe stay in that realm of discussion.

On a feature one could see approaching race-cars and the like being done with AF on your 600s and 400s. (Pretty much the same situations in which one uses AF in stills as it happens!)

 

[Donny 123]

In stills I have been AF since 1998 or suchlike after having my ass whipped at an international sporting event by someone who had gone to the dark side.

So in one media I have used AF for a while.. but it is still human, selecting a zone, going manual, knowing that backlight messes it up, knowing when a locked focus is safer than an unexpended focus dive by the camera

The human still uses skill to drive the machine.

How many times did you pull focus on those stills shoots..?.. this the expose to right for every shot .. change ISO every shot thing.. that can trip up people with a stills background.. . its totally a different ball game.. picking one frame in dozens or hundreds .. to shooting 25 or 250 a second in a 5 minute take..adjusting focus maybe every 3 or 4 seconds.. not the same universe.. .. there is just no way auto focus will become part of real film making .. I,d rather be stuck with auto aperture than auto focus .. if a gun was to my head..

 

[Ben Scott]

Ben - this thread is sort of about a pair of lenses that to my eye are for the solo OMB operator - so maybe stay in that realm of discussion.

On a feature one could see approaching race-cars and the like being done with AF on your 600s and 400s. (Pretty much the same situations in which one uses AF in stills as it happens!)

It doesn't matter either way if it's an AC or the operator with a follow focus.

Pretty certain for all the crazy techniques they employed in production that Rush didn't use AF.

 

[morgan_moore]

How many times did you pull focus on those stills shoots..?.. this the expose to right for every shot .. change ISO every shot thing.. that can trip up people with a stills background.. . its totally a different ball game.. picking one frame in dozens or hundreds .. to shooting 25 or 250 a second in a 5 minute take..adjusting focus maybe every 3 or 4 seconds.. not the same universe.. .. there is just no way auto focus will become part of real film making .. I,d rather be stuck with auto aperture than auto focus .. if a gun was to my head..

Whatever. The human knows when to turn the machine off. which I said above was currently 99% of shots for motion.

As it happens I see no shame in the stills heritage and learning to frame shots. Find out about this guy and who he trained. http://www.jamesravilious.com