EOS R6 Mark II Jagged edges in moving hummingbird wings

[Chrisinhouston]

What happened to this hummingbird's wings?

I am curious as to what is going on in this exposure of a hummingbird at the feeder. Taken with my R6 MKII and a Tamron 150-600mm lens. The wings here have jagged unnatural edges. Image exposure was 1/6400 sec at f/6.3 and an ISO of 128000. I think I was shooting in electronic shutter mode to get a fast burst rate and it's only noticeable in this one image. 

[karmlol]

This looks like an artifact of parallel readout from an electronic shutter. Electronic shutter readout happens in bands of a few pixels across the sensor simultaneously, but this readout takes some (VERY short) amount of time. Because the flutter of the hummingbird's wings is so incredibly quick, there is actually a noticeable difference in their position during the readout of each band. Can you confirm the photo was indeed taken in electronic shutter mode?

Below is my (only) example of ever noticing this on my R3, also using the electronic shutter (the bands are vertical, because the image was taken in portrait mode). It was a small CO2 gas explosion blowing the lid off of a test tube, which apparently was quick enough to cause a noticeable difference during the electronic shutter parallel readout. The vertical bands are visible in the gas of the explosion.

[Chrisinhouston]

I'm not sure how to check for shutter settings in the metadata, I don't see anything for that. I know I was shooting at the fastest burst mode on the camera and with the number of images I got I'm pretty sure it would have been in electronic mode. If I count the images taken at 2:42 and 51 seconds I have 11 shots so I know I was in a burst mode, due to the hummingbird moving about I tended to hold down the shutter button for just brief moments. 

[karmlol · Accepted solution]

I think it's safe to assume that what I described is the issue then. It would only present during VERY high speed changes in the frame, and only when using electronic shutter.

[jrhoffman75]

What is Rolling Shutter Effect in Mirrorless Cameras (and What Might Fix it) – Byrder

[karmlol]

I guess it's the same mechanic of rolling shutter, however these (and several other modern) sensors seem to work a bit differently. Rather than reading the image from the top row all the way down to the bottom row one by one, they instead split up the image into bands of a certain number of pixel rows (bands of 12 one pixel rows on my R3 for example), and then read each of those bands out row by row - top to bottom - simultaneously. So instead of reading 1 row at a time, it may read (for example) 333 rows at the same time, spread out across the frame. Through doing this, rolling shutter artifacting is somewhat 'mitigated' because slanted lines 'reset' every 12 pixels; at worst, it would cause bands or jagged lines on the bands' edges - much like in this picture.

So while you're correct that it is a form of rolling shutter, I do think that your article is a bit outdated in terms of what the artifacting looks like on (at least) some modern bodies. At the very least the Canon R3 and R6 Mk II have parallel shutter readout, as showcased in these images. I expect the R1 to be no different, and I've also read that for example the Nikon Z9 also has 12-pixel band parallel readout. That means that traditional rolling shutter as described in your article would not appear on those bodies, at least, not visually in that way. It'd appear more like jagged lines or contrast banding.