Wavy lines in video?

[amosnamos]

I was just testing my camera outside during high noon and I noticed weird color wavy lines in my video. What is happening here? Why is this happening?

Test 2 This clip was shot in 1080p 120 fps. Notice the brick wall on the building, it has wavy lines.

Test 1 This clip was shot in 4K 24 fps. Notice the brick wall on the building, it has wavy lines but not as bad as the 1080p 120 fps.

 

[J Peters]

Look to me like Moire patterns. You'll get a better explanation by doing some internet research, but it's basically where two repeating grid patterns overlap and interfere. The brick wall is a grid and so is the sensor in the camera.

TV newsreaders don't wear shirts with a small stripes or check patterns for this reason.

 

[amosnamos]

Look to me like Moire patterns. You'll get a better explanation by doing some internet research, but it's basically where two repeating grid patterns overlap and interfere. The brick wall is a grid and so is the sensor in the camera.

TV newsreaders don't wear shirts with a small stripes or check patterns for this reason.

Oh interesting I didn't know what it was called and I thought it had to do something with the sun but it was weird how the pattern shows only on the brick wall and no where else. I guess I will do some research on this moire patterns. Thanks for the quick explanation.

 

[Roger99]

What ya got there is commonly known at a moiré pattern or moiré effect. It is the result of the grid of the pixel sites on your sensor lineing up with the linear pattern of the brickwork changing the image density along the lines when the sensor and bricks line up. I think in some places in the television industry it is called hashing and is the reason why they don't like to interview individuals wearing clothing with linear patterns.

 

[thebustos]

In future, if you're looking at getting a new body, it's usually one of the subjects for video performance in reviews. Cameras aimed at video specific usage generally try to reduce this effect, though it can also pop up for shooting stills.

 

[Mark S Abeln]

As others mentioned, one way to get around this is to avoid shooting objects that will cause this pattern.

Or you can stand closer or farther away from the subject, or zoom in or out.

Another way is to use a lens which can’t resolve quite as well, or add a diffusion filter.

Basically, avoid aliasing when shooting video.

 

[John Sheehy]

As others mentioned, one way to get around this is to avoid shooting objects that will cause this pattern.

Or you can stand closer or farther away from the subject, or zoom in or out.

Another way is to use a lens which can’t resolve quite as well, or add a diffusion filter.

Basically, avoid aliasing when shooting video.

It depends a lot on the camera, too. It should be as obvious as the MP count how the camera achieves video resolution. Cameras generally do a decent job of downsampling JPEGs when set to a much lower resolution than the sensor, but for video, they often do not even read the entire sensor, to save bandwidth. Even a camera with an AA filter designed for its pixel size only blurs light over very short, tight distances, and it has no effect two pixels away from where a photon would hit, so when lines of pixels are not read or used for the video output, there isn't even any AA filtering, because the filter blurs too little to make enough of a difference.

When FF DSLRs started sporting video, it was hailed as a great, relatively inexpensive way to get video with shallow DOF, but the problem is, if your "fast" lens is actually sharp wide open, you get oodles of aliasing, making you want to stop down into the range of f-stops that make diffraction visible, and losing any benefit from using the larger sensor, both noise-wise and DOF-wise. A "soft" fast lens may actually be better than a sharp one for shallow-DOF video on an aliasing camera.

 

[John Sheehy]

Look to me like Moire patterns. You'll get a better explanation by doing some internet research, but it's basically where two repeating grid patterns overlap and interfere. The brick wall is a grid and so is the sensor in the camera.

TV newsreaders don't wear shirts with a small stripes or check patterns for this reason.

Oh interesting I didn't know what it was called and I thought it had to do something with the sun but it was weird how the pattern shows only on the brick wall and no where else. I guess I will do some research on this moire patterns. Thanks for the quick explanation.

If there's moire on the brick wall, there is aliasing everywhere there is detail and focus. It is just a lot easier to see on the brick wall, because the spacing of brick and mortar cause large-scale artifacts that just about everyone sees. There is aliasing in the grass and trees, too, but it does not form large shapes, and some people notice it, and some don't.

 

[John Sheehy]

In future, if you're looking at getting a new body, it's usually one of the subjects for video performance in reviews. Cameras aimed at video specific usage generally try to reduce this effect, though it can also pop up for shooting stills.

Not as easily, though. In stills, there is no skipping of lines of pixels in the original readout, so each pixel "sees" something right up to the borders of the adjacent pixels, so a thin line doesn't get completely lost in blind spots. With line-skipping video, a thin line can disappear completely in the cracks, re=appear somewhere else, dissapear somewhere else, etc.

By the same token, if a camera has 4K video, it will alias less than 1080 (2k) video, if line-skipping is used, as less lines will be skipped for 4k, and some cameras use a crop factor for 4K, which uses a 100% crop from the sensor, which reduces moire considerably. I'd rather be annoyed by the crop factor than by line-skipping. The ultimate would be an extremely high resolution sensor that reads and downsamples the entire sensor to all video resolutions, but that is a lot of readout bandwidth and processing power for video, by today's standards.