AC160 Ripple in AC160A footage

ullanta

ullanta

Veteran
Greetings! I have just reviewed some footage from today and have found a very strange problem... Strange rippling across the image, which looks like waves in the floor, or clothing. This seems to happen most on the long end of the lens, and in brighter areas. Happened while camera was on a tripod, and still (though I was holding the pan bars, so may have been transmitting some vibration). Anyone familiar with anything like this?
 
Vibration with OIS off would be my first guess. Outside chance it could be rolling bands caused by a mismatch between the light frequency and the shutter speed.
 
Hey Barry, long time!

i will do some testing later today before shooting again, but time will be tight, so I'm trying to maximize.

I've used this same camera of course many times, with basically the same settings (though of course need to check that nothing changed accidentally), without seeing anything like this. Really, if the floor is bright and white, it looks like it's slowly boiling. People look like they ace Aliens ready to burst out from all over their bodies as their clothing ripples and boils.

The he one change I've definitely made is that, where I usually use a focus controller (the Manfrotto focus/iris one), for this show I have disabled it (via a switch on the controller) and am just using it for iris control. Manual focus numbers in the viewfinder are steady at 72 feet. Could the manual focus assist be doing something weird? I've never used it

Something with OIS sure seems like a possibility, and I'll try it both ways, but the camera's pretty steady on the tripod.

and yes, it does seem somewhat related to brightly lit objects, so I wonder if maybe there's one particular spot with a timing issue? Ugh, I have no control of lighting for this live show, so can't test with better granularity than "whole 3-hour show".

Or perhaps my zoom controller is going and doing something subtle and weird with small fluctuations that leave the VF numbers steady?

ugh!

I have two cameras, so will switch them if I don't figure this out, and see if the camera does the same thing when locked down in a wide shot.

i suppose the hanging, motorized, LED spot itself could be doing something... But it doesn't seem like it. The local distortions, at least, seem like lens system or encoding issues.

And of course, none of this is apparent in the viewfinder, just on the monitor in post. I don't have a full monitor to use live... Ugh!
 
I haven't seen that type of distortion since the days of tube-based cameras, when timebases were much less stable than they are today.

Two ideas that come to mind are to check that you're in the highest-quality AVCHD mode for the resolution and frame rate you're using, and see if the shutter is turned on and set to a speed slower than twice the frame rate. (Shutter off is probably the safest starting point.)

Are those tungsten lights? If not, maybe you should experiment with synchro-scan a bit.

OIS off, of course.

To my eye, this looks more electronic and processing related than optical. But that's just a first impression.

- Greg
 
There's quite a mix of lights, from tungsten to LED, and there's also a projector creating the backdrops. But this seems to be happening in a variety of lighting setups... More obvious with the bright spot on the white, but there even in darker scenes. Shutter is 1/36 for 30p - in the best quality the camera can shoot (PH).

Unfortunately, this is a live show and ice no chance to test with the actual lights involved outside of the performances.

but I can at least start testing without the lights now!
 
OK, I kid you not, I switched cameras - put the 160A on the tripod for the wide shot, using all the accessories that were there (power, Zoom/Focus/Iris remote, audio); and put the 130 on the main tripod with all the accessories there. Now the 160A did fine, and the 130 had the exact same image issues that the 160 did yesterday. I could post footage - but it looks pretty much the same.

OK, I've looked at a few years worth of semiannual recordings of the same show in the same space with similar lighting, with the same exact equipment, and see nothing similar.

So - I don't think it's a problem that's developed in the (either) camera. Which leaves us with either:

1) Problem with the zoom and/or iris controller? That doesn't seem to change zoom or iris settings in any perceptible way... but could there be tiny fluctuations?

2) Some design/encoding defect that didn't pop up in using these same settings in the same situation for years?

3) Extremely localized seismic activity affecting one tripod but not the other?

4) My Qi Gong training is finally paying off and the powerful qi emanating from my hands enters the tripod and distorts the image?

5) (Your idea here)

Tomorrow maybe I'll try without the controllers, though I'm not smooth that way, and it's very uncomfortable the way the tripod is set up in the theater.

Could DRS do this? I think not, and I've used DRS 1 as I have this year in all previous years without anything like this. But it does seem to be much more pronounced in and near very bright areas. As Greg Smith said, almost a kind of distortion we used to see in analog equipment.

Help!
 
OK, tonight I revived an old HMC150 to use as the main camera, and thus had a little bit of a chance to play with settings on the 160A (set up as a third, extra camera) in the same lighting conditions. So, the camera was on a tripod that was not being touched, on a solid concrete floor. There were no accessories attached. I tried a few shutter speeds and "shutter off". Tried OIS on and off. Tried a few Iris settings.

Got the same exact ripple effect.

Hmmm......
 
OK.... I'm going to have to guess we're down to a CODEC issue - because the same ripple is there - even worse it seems! - in the HMC150 footage. Too bad I no longer have a DVCProHD cam to try....

So why haven't I seen this before? I'll have to dig in the archives again, but it seems to me that costumes in this footage have much more detailed patterns than other things I've looked at.

So maybe the CODEC is overwhelmed with detail and this is the result?

I don't know - these are pretty still shots.

Maybe it really is the qi. I have yet to compare the same footage between the HMC and AC160 to see if ripples are happening to the same degree in the same places!

But... if it's a CODEC issue present since the HMC150, has none seen this before?
 
Hmmm... I'm starting to believe more and more that the Green Barry's suggestion of actual heat shimmers in the air may be the culprit. I WAS shooting over a projector. Might explain why the wide camera (at the same level as the projector) didn't exhibit ripples while the main camera, shooting over the projector, did.... I am trying to grasp the physics of this, why the ripples would be drastic when zoomed fully in and nonexistent when about halfway through the zoom range...
 
Hi

Looks like heat shimmer to me too. I've seen it myself on footage outside o sunny days. It's so subtile that it only shows up more or less fully zoomed in.

Regards,

/Bo
 
I have observed this same effect at telephoto on several AC90s, even (outdoors) on cloudy/overcast days... which would tend to eliminate atmospheric density turbulence. Tier II Panasonic support was able to create the waviness on their unit and said it was due to the inability of the AVCHD codec at 1080p60 with a sampling rate of 4:2:0 8 bit depth rate and 28Mbps rate to faithfully reproduce fine detail at extreme telephoto lengths—and that this type of artifact is reported with professional cameras across all manufacturers that use the codec.
 
Seems easy enough to test, record internal and external and see if the internal is "wavier" than the external.

Why would it have trouble at full telephoto? If it is a bitrate issue, I would expect it to be more troubled by wide angle footage, where everything is in focus and thus the codec should theoretically be taxed more... but like I say, easy enough to test.

Anyone who can make this happen repeatedly should try with an external recorder too.
 
Perhaps this phenomenon is happening regardless the focal length, but that it is magnified at telephoto—which is why it becomes increasingly apparent as you zoom in.
 
Can you describe the circumstances that it will happen under? I presume you can't see it on the LCD, right?

If you can see the effect on the LCD, then that means it isn't related to the codec.

If you can't see it on the LCD, then I presume you would have no way of knowing you were being affected by it until viewing on a full-size monitor?

I can try to do the external recorder test but I don't want to chase various red herrings so if you can let me know an exact scenario that I can replicate that will show the effect then I can try to duplicate it.
 
I ~can~ see it on the LCD. I didn't notice it on the LCD until I saw it in the MTS files. Then I started looking for it and can see it on the LCD.
 
You must log in or register to reply here.
Back
Top