Jump to content


Photo

HD Focus, what's going on here?


  • Please log in to reply
31 replies to this topic

#1 Jamie Metzger

Jamie Metzger
  • Basic Members
  • PipPipPipPip
  • 773 posts
  • Cinematographer
  • San Francisco

Posted 20 June 2006 - 10:24 PM

I'm 1st AC on a digital shoot, on the Sony XDCAM, HDSDI out to an Apple 23" monitor.

I'm having problems when I set focus on the canon lens (6.4-128mm), I measure and it looks legit on the monitor, but zooming in for an eye focus reveals the opposite.

I've been told that you need an HD monitor to tell if you are in focus, I'm wondering how true that is since I was 4 feet out of focus on this one shot and everyone thought it looked fine.

I operated an HVX 200 on a feature a couple weekends ago, and it had a "focus assist" which was an awesome magnifier, since the focus on HD is so sensitive; but this sony camera doesn't have that feature.

I'm looking for any advice or suggestions, what should I trust? I was told that I need not back focus this lens either??

Help asap would be greatly appreciated.

Thanks

Jamie Metzger
  • 0

#2 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 20 June 2006 - 10:46 PM

If your tape focus and your eye focus don't match, then it sounds like the back-focus is off. If the shot has enough depth of field, people might not notice that the focus is off, but you should still get it corrected.
  • 0

#3 Jamie Metzger

Jamie Metzger
  • Basic Members
  • PipPipPipPip
  • 773 posts
  • Cinematographer
  • San Francisco

Posted 20 June 2006 - 10:49 PM

That's what I was wondering... I've never 1st'ed on an HD shoot before, and I've heard horror stories about back focus.

I made sure to tell the DP that I had no experience in Back focus, and if he showed me once how to do it, I'd be fine. He said he didn't think that this particular lens had to be back focused, so I took his word (not wanting to worry about the BF)

What advice would you have? I've been measuring or eye focusing most every shot, and the monitor shows it looks in focus, it's only when zooming into 128mm, do things go very blurry.
  • 0

#4 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 20 June 2006 - 11:06 PM

You're probably being covered by depth of field. I don't know why he'd say one type of lens doesn't need to be back-focused while another would.

Now focus is SO critical at the long end of the zoom that it could simply be that you're off in your focus. In other words, the backfocus is fine, it's just that it's only when you're zoomed in and the depth of field has dropped that you can see that you're off in your focusing. It was always off, it's just that depth of field covered the mistake at the wider focal lengths.

The old-fashioned way of backfocusing would be to focus on something at infinity with fine details, like a building or tree. Turn up the peaking in the viewfinder (so you can tell when the focus pops into place), zoom all the way in, focus, and then zoom all the way out. Adjust the backfocus until the image is sharp at the wide-angle end. Lock it off, zoom in and out to make sure that the focus is correct at both ends.

The more accurate way would be to have a focus chart like a Seimens Star Chart (a circle with black & white triangle slices, the points meeting at the center), or a "Sharp Max" device.

With the Seimens Star chart on a wall or c-stand, you could set the lens at 4', let's say (so that it is big enough in frame even at wide-angle), and measure out to the chart so that the camera is exactly 4' from the focus plane. Zoomed in, it should be in focus. Zoom out and adjust the backfocus until it is also sharp, then lock it off. With the peaking up in the viewfinder, the center of the Seimens Star easily shows when the fine points are sharp or not.

The lens should be wide-open, so switch in some ND if necessary.

Now I seem to recall that some ENG zoom lenses have the measurement marks set so that you tape measure from the front of the lens, not the focal plane. You should ask your rental house which is it.

If something looks as sharp as possible on the big monitor, it should be sharp, but the problem if the backfocus is off is that you can't hold focus as you zoom.
  • 0

#5 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 20 June 2006 - 11:38 PM

Agreed, it sounds like a backfocus problem. And David's suggestions are right on target.
  • 0

#6 Jamie Metzger

Jamie Metzger
  • Basic Members
  • PipPipPipPip
  • 773 posts
  • Cinematographer
  • San Francisco

Posted 22 June 2006 - 03:28 AM

Agreed, it sounds like a backfocus problem. And David's suggestions are right on target.


I checked into this today, and spoke with the DP about it, and it seems that Back Focus was never off. Hasn't slipped either. I guess acceptable focus in HD is a lot bigger than in 16 or 35?

Anyone know what 1/2 inch chip for the XDCAM 350 relates to film size wise?

Thanks Mr. Mullen again, I printed out your suggestions and follwed them right when I got to set.
  • 0

#7 Laurent Andrieux

Laurent Andrieux
  • Sustaining Members
  • 1527 posts
  • Cinematographer
  • France

Posted 22 June 2006 - 05:47 AM

Weird... The fact that the focus gets off when you zoom in and the fact that it only occures on one lens let's think 99% it is a the back focus of that lens.

I mean it's FFD, actually.

Check as David said, and if that test is good on all the lenses but that one, for sure it's its FFD
  • 0

#8 Hunter Sandison

Hunter Sandison
  • Basic Members
  • PipPip
  • 29 posts
  • 1st Assistant Camera
  • Los Angeles, CA

Posted 22 June 2006 - 10:26 PM

The Seimans Star can give false focus in HD(about 18" off, I think). I don't know why but thats what I've been told at the rental houses. For this reason they use an HD focus chart. Its made of concentric rings instead of the triangles and is lower contrast. Its usually black and grey instead of black and white. If someone could enlighten me as to why this is, or if its true at all, I would appreciate it.
  • 0

#9 Laurent Andrieux

Laurent Andrieux
  • Sustaining Members
  • 1527 posts
  • Cinematographer
  • France

Posted 23 June 2006 - 04:34 AM

I didn't know that but I noticed that focusing on a siemens star is a bit tricky even in SD because of the contrast of the chart. There is this border effect - that is even stronger in the vf when you use a strong peaking - that may puzzle one. May be this is why they like better using this chart you're talking about... but if you encounter a problem with one specific lens... I'd say it's that very lens...
  • 0

#10 Sam Wells

Sam Wells
  • Basic Members
  • PipPipPipPip
  • 1751 posts
  • Cinematographer

Posted 23 June 2006 - 08:06 AM

If the back focus is off on a zoom lens, too far from image plane, then strictly speaking it will only focus properly at one point in its zoom range.

-Sam
  • 0

#11 Jamie Metzger

Jamie Metzger
  • Basic Members
  • PipPipPipPip
  • 773 posts
  • Cinematographer
  • San Francisco

Posted 23 June 2006 - 09:59 AM

If the back focus is off on a zoom lens, too far from image plane, then strictly speaking it will only focus properly at one point in its zoom range.

-Sam



Makes complete sense.
  • 0

#12 Brian Drysdale

Brian Drysdale
  • Basic Members
  • PipPipPipPip
  • 5069 posts
  • Cinematographer

Posted 23 June 2006 - 11:51 AM

The Seimans Star can give false focus in HD(about 18" off, I think). I don't know why but thats what I've been told at the rental houses. For this reason they use an HD focus chart. Its made of concentric rings instead of the triangles and is lower contrast. Its usually black and grey instead of black and white. If someone could enlighten me as to why this is, or if its true at all, I would appreciate it.


Paul Wheeler discovered this while they were testing how accurately people were setting up the back focus on HD cameras. His chart (It's the one that Panavision sells in the UK) has both the Siemens Star and the Annular Focus Target.

He seems to conclude that the apparent sharpness was being influenced by high contrast of the star, so giving false results - this is his best guess for the reason, it's not scientific.

After the tests Paul Wheeler came up with this new chart.
  • 0

#13 Sam Wells

Sam Wells
  • Basic Members
  • PipPipPipPip
  • 1751 posts
  • Cinematographer

Posted 23 June 2006 - 01:06 PM

Sometimes I used to check back focus on ENG cameras using standard charts - and a waveform monitor. More sharpness = more contrast at the edge of the line pairs = more modulation.

Best to turn all detail OFF.

-Sam
  • 0

#14 Michael Nash

Michael Nash
  • Sustaining Members
  • 3330 posts
  • Cinematographer
  • Pasadena, CA

Posted 23 June 2006 - 04:26 PM

Anyone know what 1/2 inch chip for the XDCAM 350 relates to film size wise?


None. B)

A 2/3" chip is pretty close to the size of a Super 16 frame, close enough to use similar focal lengths for any given framing (a skosh shorter with 2/3"). A 1/3" chip would be closer to 8mm film. A 1/2" chip is exactly between those sizes.

So on a practical level you would use shorter focal lengths with a 1/2" system than you would use for Super 16mm to attain the same angle of view. Maybe about 20% or so.
  • 0

#15 Adam Frisch FSF

Adam Frisch FSF
  • Sustaining Members
  • 2009 posts
  • Cinematographer
  • Los Angeles, USA

Posted 25 June 2006 - 05:43 AM

Isn't it funny to know that if Sony didn't build their cameras so crappily and out of papier mache, then that problem would never occur. Clairmont rebuilt all their F950's and voila! - no need to ever check back focus again.

I recently did a DigiBeta-job where the images we shot on the wide zoom were soft, despite dual back focus checks. These cameras are a joke.

So Sony, how hard would it be for you to build a lens mount and a CCD block that sit together as one unit?
  • 0

#16 Hal Smith

Hal Smith
  • Sustaining Members
  • 2280 posts
  • Cinematographer
  • OKC area

Posted 25 June 2006 - 07:50 AM

Isn't it funny to know that if Sony didn't build their cameras so crappily and out of papier mache, then that problem would never occur. Clairmont rebuilt all their F950's and voila! - no need to ever check back focus again.

Almost every single Sony product I've ever had any experience with has had something wrong with its mechanical design and manufacture. When a Sony dies, it's invariably a mechanical problem. I've had to junk expensive professional CD players, cameras, monitors, radios, TV's, computer gear, etc. because Sony didn't put enough effort into mechanical design. It's as if they have a corporate culture that recoils from the notion that spending a little more time and money at the design and manufacturing level will give the customer added value and service from their equipment.

I've got old Toshiba gear that keeps humming along, I still have my first CD player, a Toshiba XR270K (circa 1985), it still works, has needed one small repair over the years (a tray belt), and is in my den playing a CD as I write this. I know of no Sony CD of that first generation that still works.

Designing and building gear to last isn't cheap - but it's possible. Agilent (Hewlett-Packard test equipment) sales people bemoan the fact that their worst competitor is often themselves. They try to sell someone a new piece of gear only to hear from the customer that the twenty year old HP piece of gear they own is still doing a great job - and is still better than much of the non-HP gear built today.
  • 0

#17 Matias Nicolas

Matias Nicolas
  • Basic Members
  • PipPipPip
  • 157 posts
  • Cinematographer
  • Buenos Aires, Argentina

Posted 26 June 2006 - 10:55 AM

I worked on 2 films with HD as 1AC, and I did have the problem of the back focus. But I also had another problem. Sometimes, we set the lens scale in the morning , and at midday the scale changed . We reescaled it, and the in the afternoon , was off again . I had this problem in both films .
The HD techs, argumented that its a ccd problem : changes position inside te camera with the differences in temperature during the day. So imagine ! I measured the actor's position, and while shooting, the DoP or the Director had to stop, because of out of focus ! So my question is... this problem with te ccd is true? did you had this problem?
  • 0

#18 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 26 June 2006 - 11:05 AM

That's why the back-focus has to be adjusted -- for one thing, temperature changes cause the aluminum block holding the prism & CCD's to expand or contract. This is a reason why Clairmont Cameras replaced the block with a stainless steel one, and why Sony is going to do the same for their future replacement to the F950.

Also, these cameras are basically heat generators, so all sorts of little problems can happen as they overheat, including more noise in the image.
  • 0

#19 Dan Goulder

Dan Goulder
  • Sustaining Members
  • 1259 posts
  • Cinematographer

Posted 26 June 2006 - 11:27 AM

Also, these cameras are basically heat generators, so all sorts of little problems can happen as they overheat, including more noise in the image.

Have you noticed a degradation in performance with these cameras over the course of a day's shoot? Thanks.
  • 0

#20 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 19759 posts
  • Cinematographer
  • Los Angeles

Posted 26 June 2006 - 12:13 PM

Have you noticed a degradation in performance with these cameras over the course of a day's shoot? Thanks.


No, but I just did a film-out of some F900 material of a dark, moody movie shot a lot in fake pale blue moonlight, and some scenes have blue noise problems and others don't, and I can't figure out what I might have done differently. But as I was shooting in Texas in August and the camera once wouldn't start up because it was too hot & humid one day, heat may have played a factor in the extra noisiness of a few shots. Or maybe I just crossed some magic line in minor color-correction that caused the blue noise to kick in. Oddly enough, it wasn't really visible in the color-correction suite, only the film-out.
  • 0


FJS International, LLC

Abel Cine

Opal

Wooden Camera

Willys Widgets

Broadcast Solutions Inc

Rig Wheels Passport

The Slider

Paralinx LLC

Metropolis Post

Tai Audio

Media Blackout - Custom Cables and AKS

Aerial Filmworks

rebotnix Technologies

Visual Products

Ritter Battery

CineLab

Glidecam

Technodolly

Gamma Ray Digital Inc

CineTape

Gamma Ray Digital Inc

Glidecam

Tai Audio

The Slider

Willys Widgets

Technodolly

Wooden Camera

rebotnix Technologies

Broadcast Solutions Inc

Rig Wheels Passport

Media Blackout - Custom Cables and AKS

Visual Products

CineLab

FJS International, LLC

Aerial Filmworks

Ritter Battery

Paralinx LLC

CineTape

Abel Cine

Opal

Metropolis Post