M6ii Electronic Shutter artifacts

I've noticed a couple of things when using the electronic shutter on my M6ii, and had questions about them.

The first, I noticed back in the spring using the 18-150mm. If you look at the following image, you'll notice blurring/double image about one quarter over from the right-hand side of the image (clearly visible in the sunglasses and the pavement at the very bottom of the image). The shutter speed was 1/250.

Electronic shutter interacting with image stabilization?

My guess is that the IS moved while the electronic shutter was reading out that part of the image? Or is there some other cause? I was going to post this image back when I noticed it, but since then I've use the 22mm and the 32mm way more often, so it's not been something I've worried about.

The second question I have has to do with lens flares.

Just this past week, I was outside taking photos of my daughter using the 22mm with the mechanical shutter. The sun was low and behind her, and so I was trying to get lens flare in the photos for the sake of effect. I could clearly see ghosting flare in the EVF and on the rear screen, but when I took the photo, the brightest part of the flare I saw was entirely gone, with just a more faint part of the flare remaining. I wondered if the camera was somehow intelligently removing it, but turning off everything when reprocessing the RAW in camera didn't change anything. I took a quick video panning back and forth with the sun in the frame, and the video clearly had the flare. Because of this, I switched from mechanical to electronic shutter, and now the ghosting flare was fully present in the image.

Here are two SOOC JPG's taken back to back, first with the mechanical shutter and then with the electronic shutter (despite the slight shift in perspective, the flare was clearly visible in the viewfinder when taking both photos).

Mechanical Shutter

Electronic Shutter

This shows me that if I ever want this kind of flare, to use the electronic shutter (and to avoid or lessen it use the mechanical shutter), but does anyone know why the flare is so dramatically affected by which shutter mode is used?

These are both really interesting examples. I am still contemplating the flare issue as my first thoughts seem to have run into some math problems.

With regard to the first image, clearly you have some sort of motion blur. It could be camera movement, subject movement, IS movement, or some combination. The blurry stripe on the ground tells us that it certainly is primarily something on the camera side. The readout time on the full electronic shutter is about 47 milliseconds. With the camera tipped on its side, the readout progresses from either the right to left or left to right, depending on which way you have the camera tipped. We can definitively say that the event that caused the blur lasted for approximately 5.5 milliseconds and began about either 8 or 33 milliseconds after the start of the exposure. It is an interesting example. Given the incredibly short duration and large magnitude, I would rule out human cause. I think you are right that the IS jumped during the exposure. It is an interesting example illustrating another potential downside to slow readout electronic shutters. I suspect this is a very low probability event, though.

TBIRallySport wrote:

I've noticed a couple of things when using the electronic shutter on my M6ii, and had questions about them.

The first, I noticed back in the spring using the 18-150mm. If you look at the following image, you'll notice blurring/double image about one quarter over from the right-hand side of the image (clearly visible in the sunglasses and the pavement at the very bottom of the image). The shutter speed was 1/250.

Remember that this shutter speed tells us for how long each pixel is "on" to record the light. It still takes about 1/20 sec between the moments the first line and the last line on the sensor are read. This slow readout speed is the source of the distortion in moving subjects.

Electronic shutter interacting with image stabilization?

My guess is that the IS moved while the electronic shutter was reading out that part of the image? Or is there some other cause?

It looks like there was a "knock" of some kind close to the end of the exposure and the OIS tried to recover from it. You can see a few blurred vertical lines of pixels going through the child's left foot, then a slightly sharper area, then a slightly blurred area again towards the edge of the image. It is hard to say if the OIS caused the "knock" in the first place. It should not have. I would keep an eye on this problem in case there is a fault with the lens.

I was going to post this image back when I noticed it, but since then I've use the 22mm and the 32mm way more often, so it's not been something I've worried about.

The second question I have has to do with lens flares.

Just this past week, I was outside taking photos of my daughter using the 22mm with the mechanical shutter. The sun was low and behind her, and so I was trying to get lens flare in the photos for the sake of effect. I could clearly see ghosting flare in the EVF and on the rear screen, but when I took the photo, the brightest part of the flare I saw was entirely gone, with just a more faint part of the flare remaining. I wondered if the camera was somehow intelligently removing it, but turning off everything when reprocessing the RAW in camera didn't change anything. I took a quick video panning back and forth with the sun in the frame, and the video clearly had the flare. Because of this, I switched from mechanical to electronic shutter, and now the ghosting flare was fully present in the image.

Here are two SOOC JPG's taken back to back, first with the mechanical shutter and then with the electronic shutter (despite the slight shift in perspective, the flare was clearly visible in the viewfinder when taking both photos).

I would expect the ES shot to capture the flare the way it looked in the viewfinder, because in both cases the sensor is not covered by the shutter and is fully exposed to the light. When a mechanical shutter is used at a shutter speed much higher than the flash sync speed, the sensor is exposed through a narrow moving slit (at 1/2000 sec it will be about 10% of the image height) between the front and the rear curtain of the mechanical shutter. A large part of the flare is caused by the reflections from the sensor to the lens and then back to the sensor. Now, if 90% of the sensor are covered by the shutter curtains at any given time, this cuts down these stray reflections and reduces flare a lot. If you were shooting mechanical shutter at 1/200 sec, the flare would have been similar to that of the ES.

Mechanical Shutter

Electronic Shutter

This shows me that if I ever want this kind of flare, to use the electronic shutter (and to avoid or lessen it use the mechanical shutter), but does anyone know why the flare is so dramatically affected by which shutter mode is used?

This topic was educational.

AKRover wrote:

These are both really interesting examples. I am still contemplating the flare issue as my first thoughts seem to have run into some math problems.

With regard to the first image, clearly you have some sort of motion blur. It could be camera movement, subject movement, IS movement, or some combination. The blurry stripe on the ground tells us that it certainly is primarily something on the camera side. The readout time on the full electronic shutter is about 47 milliseconds. With the camera tipped on its side, the readout progresses from either the right to left or left to right, depending on which way you have the camera tipped. We can definitively say that the event that caused the blur lasted for approximately 5.5 milliseconds and began about either 8 or 33 milliseconds after the start of the exposure. It is an interesting example. Given the incredibly short duration and large magnitude, I would rule out human cause. I think you are right that the IS jumped during the exposure. It is an interesting example illustrating another potential downside to slow readout electronic shutters. I suspect this is a very low probability event, though.

Thanks for doing the math!

I hope that, like you say, it was just the IS jumping and is a very low probability event. I haven't noticed this in any other shots, but I use the 22mm and 32mm so much more that I just haven't had as many opportunities for this problem to present itself.

Dem Bell wrote:

I would expect the ES shot to capture the flare the way it looked in the viewfinder, because in both cases the sensor is not covered by the shutter and is fully exposed to the light. When a mechanical shutter is used at a shutter speed much higher than the flash sync speed, the sensor is exposed through a narrow moving slit (at 1/2000 sec it will be about 10% of the image height) between the front and the rear curtain of the mechanical shutter. A large part of the flare is caused by the reflections from the sensor to the lens and then back to the sensor. Now, if 90% of the sensor are covered by the shutter curtains at any given time, this cuts down these stray reflections and reduces flare a lot. If you were shooting mechanical shutter at 1/200 sec, the flare would have been similar to that of the ES.

Thanks! I knew that the the mechanical shutter only exposes a fraction of the sensor at any one instant at fast shutter speeds, but I had not paid close attention to the shutter speed in this example and forgot to consider that factor.

Kind of interesting how it sort of lets you see which portion of the lens flare is due to light reflecting off of the sensor, and which isn't.

And I hope the first picture is not due to any problem with the lens. I don't often shoot with the 18-150mm, but I'll try to keep an eye out if the issue presents itself again.