Focus failed totally on aerial shoot with Z1

[Chris Hewitt]

I shot 20 minutes of 1440x1080i from a small top-wing Cessna this morning as I have done previously with other cameras, but this time using a Sony Z1(which is in top condition). This time however, every single frame is unusable. While in the air, I used a pillow under the cam over the open window to absorb any vibration and this has always worked well. The flight was normal, 13kts of wind and clear skies.

The focus was giving me big problems on auto and manual focus set to infinity, it would not stay at infinity but once on the ground, no problems. If I frame jump the footage in Vegas, I can see the focus changing by the frame, always very soft focused and never sharp. Not one single frame is sharp and yes, I know how to operate this camera. Every other setting was at auto except the focus, the OIS was on at times and off at others as I tried to fix the problem. The tape is good as I have shot 15 minutes on the ground either side of the aerial footage without issue and heads are clean. Is it possible the aircraft can interfere electronically with the camera? Again, I haven't previously experienced that.

I don't know if it's an issue, but I had a Sony Circular Polarizer on at all times.

Has anyone experienced this problem because if I go up again, the same thing might happen. It's a very expensive problem for me and I'm facing a deadline.

[Chris Barcellos]

Was the camera focus ring being manipulated by the pillows pushing against it ?

[Chris Hewitt]

No, I made sure about that. I even held the camera with one hand for a while to see if that was the cause. I'm beginning to suspect the polarizer as I've read a moment ago that can cause problems with SLR cameras.

[Bob Hart]

Did you ever have anything plugged into the (forgotten what it is??? ) remote control or LANC socket?? The camera is out on loan so I can't check the book or the camera.

I had the autofocus and manual focus switch positions change places when I had a Mini35 remote control hooked up to the camera. It was totally ungovernable until I took the plug back out and then not until I had switched the camera off and on again. I had used the remote previously with no problems but have not used it since. I don't like to poke sleeping dogs with sticks.

It has played up occasionally before. The timecode went weird after I recorded on a few used tapes which had been previously recorded with MiniDV. I had received them at no cost as they were one-pass event logging tapes which were never re-used.

Most small video cameras have plastic, not metal cases and there have been anecdotal accounts of radio-frequency interference getting into them.

There are two magnetos churning away on the engine and an alternator, maybe a transponder which is a constant duty device, plus several voice transceivers, maybe a GPS though I don't know if these cause problems. If they did they would likely not be allowed on an aircraft. The ignition systems on aircraft are shielded to limit RFI and care is taken that the various radio-frequency devices do not interfere with each other.

You would have been wearing a headset and lip mike?? These are fed by two unbalanced leads which might propogate RF interference from the outer shield to the camera as they are about the right quarter-wavelength for VHF.

My bet would be on some radio-frequency interference coming back off the headset lead, yours or the pilot's if you were not wearing a headset. The sensors under the camera focus ring or the conductor from them to the detector and amplifier which powers the lens focus servo motor may be picking this up, interpreting it as a focus signal and generating false pulses to the focus servo.

I'm talking guesswork and generically here. I only know just enough to endanger myself so don't pay much heed to my comments.

[Chris Hewitt]

Bob, thanks very much for your reply. Incidentally, although I'm living in the UK, I bought the Z1 down at Osborne Park two years ago when it was released. Perth is my home town but I've been over here for four years now but only two more to go.

The electronic interference you mention...yes, I was wearing a headset and mic and it was intermittent as I overheard the pilot commenting on that fact to the ground. Could be something in that.
I would be surprised if the polarizer was the culprit having used this same set up in the past with a Canon G2. The focus set to manual was totally out of my control, no matter what I did with the camera, it had a mind of it's own. And yes, the shots would have been perfect, everything was right....except where it mattered most.

No. I never had any LANC device on this camera and don't use the remote although the menu setting is set to 'on' because I had been thinking about using the remote.

[Greg Laves]

I had a similar experience when shooting out of a helicopter. Even though I had the camera set on manual focus and at infinity, it still looked like it was refocusing occasionally. But I think it did have the image stabilization turned on, if I remember correctly. I assumed that OIS was the culprit.

[Bob Hart]

Chris.


I have never bothered to go looking for grounding points on the the camera body as my aircraft stuff has all been done with a PD150 which behaved itself.

Being as how it is mostly plastic finding a convenient grounding point is a bit difficult. There will be a sheetmetal or diecast substrate in there somewhere.

I dont think the little tripod baseplate would be any good as it screws to the plastic casework. The shield around the XLR socket is the most robust and accessable. The little latch on the XLR socket is a likely handy place to clip onto.

A decent radio tech or the camera service people themselves would be able to trace from the XLR outer shield to the body screws to see if the screws fasten to any internal metal structure. You could then use a longer screw and use it to attach a small tag for clipping an earth lead onto.

The carry handle itself appears to be metal so might be grounded to the camera internals. Via this path, a cushioned (insulated from the ear skin) headset plus a bit of sweat on a human and human hand on the handle might be creating the capacitative and resistive components of a crude tuned circuit.

After advice from the techs, it might be feasable to ground the camera body to the airframe. This will eliminate the human body and the airspace between the headset cable shield and the camera as a part in an unwanted tuned circuit.

Another way to go about it would be to plug a splitter into the headset socket on the instrument panel and feed the headset audio either direct or via an inline impedance transformer to one of the XLR line channels depending on how good you would want the audio to be.

It is useful to have the radio traffic on an audio channel. This would establish an connection to the cable shield. You would now have to mount an XLR mike on the camera for cabin ambience as you cannot split between internal mic and XLR. The mike takes up valubable space but I find turning it back to front gets it out of the way somewhat.

I would not monitor the audio off the camera as you should wear the aircraft's headset and lip mike so you can respond to the pilot's commands. You are enough of an addition to his workload already by simply being there with a camera without the need for your pilot to break routine to look at you for shouted replies, hand signals, or touch on the shoulder.

You might have to be careful about the 24VDC power system on the aircraft and ask the maintenance engineer if there is a 24VDC potential across the conductors that might harm the camera.

People routinely plug their Z1s into large mains powered audio mix desks at music gigs and things don't fry.

A quick and dirty and probably safest test which would not involve any actual electronics would be to connect a wire with small alligator clips from the release latch on one of the XLR inputs on the camera to a bare metal part of the airframe. Lace any added wires to the headset cable so you don't entangle the control yoke.

Your comment about the pilot reporting intermittant interference to the ground??

If your camera focus servo was reacting to radio-frequency interference breakthough, there is every likelyhood it may have been feeding back via the same path into the aircraft systems. A grounding link may likely look after that issue as well.

As a footnote, the permanent magnets in a powered-off PD150 will deviate an aircraft compass by about 3 to 5 degrees at about 200mm. At a music gig, a mobile phone worn by an assistant buttoned off a FX1 on a fixed mount when he went near it to check framing.

[Chris Hewitt]

Greg, I thought inflight that OIS might be to blame so I turned it off with no positive effect. Thanks for your idea though.
Bob, I'm going to see the pilot in a couple of hours and hopefully we can arrange a date to test the camera onboard with the engine running while still on the ground so I can eliminate these points while not paying for actual flight time, and then go up. I also have a Panasonic SD5 camera which I will take to see if that is affected.
That would tend to suggest the aircraft is the culprit if both cameras play up.
I will get back to you here as soon as I know something. You've given me some good suggestions that I can put to the pilot and thankyou very much for that.
BTW, I wasn't recording any audio, it was just the video I was after.

[Chris Hewitt]

Has anyone had any focus problems shooting with the Z1 in a situation where there was substantial vibration i.e. engine vibration? If so, how did it affect the recorded video?

Bob, the pilot reckons that RFI interference is 'most unlikely' but we shall see when I get a chance to reshoot.

[Bob Hart]

Chris.


Your pilot is probably right. A great deal of R&D goes into aircraft systems so that they work right every time.

The issue for the radio transmission is that the aerial which transmits it is a perfect part of a tuned circuit. If the aerial has to be cut shorter than is ideal for a given radiofrequency, then compensation is built in to the design to allow that aerial to work. Radios in aircraft have presented some challenges over the years.

Modern design and components have come along way and maybe now for each VHF channel, there may be now a perfectly tuned load for the transmitter through some sort of automatic antenna tuner for each and every VHF channel it can switch to.

In earlier days, a fixed antenna was trimmed or an antenna tuner adjusted to the wavelength of the centre of the several frequencies (band) used. Those towards the extremities of the band the antenna was intended to send would not quite be as efficient. It was possible to use an antenna tuner for fine tuning the antenna to each frequency or radio channel however this task is an added burden for a pilot and the facility is not included. For practical purposes it does not matter

When a fixed length antenna and a tuning load is not quite right, then interesting things happen and some of the RF energy comes back down the feedline to the transmitter. You may find some of the reflected energy present on the outer shield of cables attached to that transmitter.

Don't discount RF interference entirely. It can be an elusive beast.


Another consideration comes to mind. You mention window-open ops. It may just be that air-pressure fluctuations might be "pumping" the moving group inside the lens if the group is a snug fit inside the tube it moves in.

I would have expected that if at all, this might be seen as a trembling effect on the zoom rather than a focus shift, however focus and the zoom movement are linked so it may be manifesting differently. When next you fly, try shooting through the the closed window then with the window open.

There is not much you can do about that except to put the camera in a dive case which is impractical for air work.


As for vibration affecting a camera braced against the airframe, this might be observed as a blurring of the image at shutter speeds of 1/50thsec or thereabouts and an interlace artifact with higher shutter speeds which will present as a sort of comb effect when the motion is vertical, ie., every second line appearing darker or maybe every fourth or eighth or other combinations. It will be more apparent with the lens zoomed-in.

Other past posts in this subject here have suggested it is more or less accepted practice not to brace against airframes.

The power pulses of a 4 cylinder engine from each cylinder firing, may introduce a harmonic of 50i (PAL) at 2250 engine rpm. Propellor airflow buffet upon the airframe is the same for a two bladed propellor. If your engine power setting is very close to 2250rpm, then you may observe a periodic effect with the image going softer and sharper or interlace artifacts appearing and disappearing, probably fairly slowly.

A "pumping" effect on the lens might be similarlly effected if the air-pressure fluctations at the camera are caused by propellor buffet. This and vibration could also be working in combination against you.

Rather than brace via a cushion against the airframe, you might be better served, when the aiming direction of your shot permits, by handholding with the heel of your right hand under the battery and fingers up the back of the battery and the heel of your left hand under the zoom ring ( rocker switch selected ) and fingers under the lens hood if you still have it fitted for air work. You end up using your thumb for manual focus trims. I'm left-handed so the opposite grip might suit you better.

This feels awkward but forces your arms and body posture into something like the iso-elastic arms of a steadycam. I also favour using the LCD not the rear viewfinder for air work as the viewfinder cup anchors the rear of the camera and aggravates movement as well as poking you eye out in a bump.

[Chris Hewitt]

Bob, your quote as follows "....As for vibration affecting a camera braced against the airframe, this might be observed as a blurring of the image at shutter speeds of 1/50thsec or thereabouts and an interlace artifact with higher shutter speeds which will present as a sort of comb effect when the motion is vertical, ie., every second line appearing darker or maybe every fourth or eighth or other combinations. It will be more apparent with the lens zoomed-in."

I think you have just described how this footage looks. Of course, with the shutter speed being on auto, it was exactly 1/50. And the pillow resting on the lower frame of the open window, it's an open path for vibration which I am starting to suspect is the cause.

On close inspection, the focus blur manifests itself as a vertical jump, no more than an interlace line or two about a dozen times a second, possibly more as no single frame is sharp, continiously moving up and down, and this is enough to make the footage appear out of focus.
For arguments sake, if I was dumb enough to shoot out of focus, the frames would at least be somewhat constant but the case is that there is vertical shake.And that manual focus was most definitely set to infinity (while it lasted.)

Thanks for bearing with me on this, I will get to the bottom of it and post the results so other Z1 owners might benefit. I am weighing each point you put across, Bob, and will figure it out. Any further thoughts you may have are very welcome.
If I can think of a way to simulate aircraft vibration, I could then test the camera without paying for expensive airtime while I investigate. Do you think that with the nature of HD resolution, vibration is more apparent that the normal 720?

[Bob Hart]

Chris.

To replicate the aero-engine vibration, you may find that strapping a mobile (cell) phone to the front of your hand-held camera on bottom or on the left side of your lens hood and selecting the vibrator if you have that option may be enough to set the thing off with 70% to 100% zoom-in or maybe wider if the phone is particularly violent.

You may have to reposition the phone face-on or side-on until you get the plane of the eccentric weight moving the right way to provoke any artifact.

My experience of this effect was with a groundglass disk running out of balance on a home-made AGUS35 35mm adaptor. It was enough to show up with a 50mm prime lens which is about the equivalent of about 20mm zoom-in on your Z1.

Phone vibrators are just on the bottom end of audible at about (guessing) 40 to 50Hz. If you loosen the lens hood screw so that the lens hood itself shakes about a bit, you may get a harmonic period close to the power pulse and prop buffet of an aeroengine with a two-blade prop.

The reason for choosing the left side or bottom of the hood is that the hood won't come off and dump your phone on the floor when you loosen that thumbscrew.

If you dont get a harmonic from the hood vibrating, a marble sized lead sinker or a small automotive wheel-balance weight loose inside the hood, tagged away from the lens glass with a bit of sticky-tape but otherwise free to move should do it for you. If the cushioning of your hands dampens the vibration, find the weakest flimsiest tripod and fix the camera onto that. If it has an extendable neck, set that at its highest. Something will have to shake then.

If you capture some of the footage and de-interlace and render out, you may observe the artifact changes to a vertical weaving movement and may sharpen just a fraction.

SD versus HDV, the effect on sharpness I expect would be the same, but less movement would be required to provoke an interlace artifact because of the finer pitch of the lines or horizontal pixel rows. With the newer Sony CMOS cams and their canted pixel rows I expect that with any rolling shutter effect discounted, a diagonal artifact might occur.

[Aric Mannion]

I haven't read this entire thread but I would:
Make sure nothing is on auto, if the shutter speed changes from 60 (for example) it can make the picture shift.
Don't have steady shot on, that can soften the picture too.
And I wouldn't even set the camera on anything, but my own hands. It depends on what type of shot you are getting -but does it still vibrate in your hands?

[Chris Hewitt]

Further to this problem....I have tested the camera while resting it on the top of my diesel car engine(!!) to simulate vibration and sure enough, the same symptoms are seen, lots of vibration, so I guess that as you say, hand holding the camera is the best way.
I haven't had the opportunity to go up in the plane for a reshoot yet due solely to the lousy British weather but will as soon as the light is right.
I will definitely ensure that everything is set to manual. The reason that I was confused by all of this is that I had shot the same conditions last year with a Canon GM2 (0n auto) and the vision was perfect. Maybe HD cams are more sensitive!!
I'll post back when I've done the reshoot. Now to get all the diesel out of the tape transport!!

[Dave Blackhurst]

Originally Posted by Chris Hewitt

Maybe HD cams are more sensitive!!

Only about 4x more data running about to get scrambled <wink>! There's definitely a learning curve when you do something HD vs. SD!